TABLE OF CONTENTS

TESOL QUARTERLY

Volume 24, Number 3 u Autumn 1990

A Journal for Teachers of English to Speakers of Other Languages and of Standard English as a Second Dialect

Editor

SANDRA SILBERSTEIN, University of Washington

Review Editor

HEIDI RIGGENBACH, University of Washington

Brief Reports and Summaries Editor Assistant Editor

GAIL WEINSTEIN-SHR, University of Massachusetts at Amherst/ Temple University DEBORAH GREEN, University of Washington

Editorial Assistant

MAUREEN P. PHILLIPS, University of Washington Editorial Advisory Board

Roberta G. Abrabam Iowa State University Margie S. Berns Purdue University Joan Eisterhold Carson Georgia State University Ruth Larimer Cathcart Monterey Institute of International Studies Graham Crookes University of Hawaii at Manoa Miriam Eisenstein New York University Liz Hamp-Lyons University of Colorado at Denver Mary McGroarty Northern Arizona University Thomas Ricento Japan Center for Michigan Universities/ Central Michigan University May Shih San Francisco State University James W. Tollefson University of Washington Vivian Zamel University of Massachusetts at Boston

Catherine Doughty

The University of Sydney Patricia A. Dunkel The Pennsylvania State University

Additional Readers
William R. Acton, Bradford Arthur, Nathalie Bailey, Lyle F. Bachman, Gregory Barnes, Patricia L. Carrell, Marianne Celce-Murcia, Carol Chapelle, Christine Clark, James Coady, Ulla Connor, David E. Eskey, Janet L. Eyring, Christian Faltis, Mary Lee Field, Donald Freeman, Fred Genesee, Christine Uber Grosse, Mary Hammond, Sharon Hillis, Thom Hudson, Barbara Kroll, Ann M. Johns, Robert B. Kaplan, Michael K. Legutke, Ilona Leki, Nora E. Lewis, Patsy M. Lightbown, Peter Lowenberg, Peter Master, Jean McConochie, Sandra Lee McKay, Sharon Myers, Eric S. Nelson, Sonia Nieto, Alastair Pennycook, Martha C. Pennington, Elizabeth Platt, Patricia A. Porter, Ann Raimes, Joy Reid, Patricia L. Rounds, Terry Santos, Robin Scarcella, Thomas Scovel, Tony Silva, Marguerite Ann Snow, Margaret S. Steffensen, Michael Strong, Elaine Tarone, Jean Turner, Carole Urza, Evangeline Varonis, Roberta J. Vann, Elizabeth Whalley, Rita Wong.

Credits
Advertising arranged by Helen Kornblum, TESOL Central Office, Alexandria, Virginia Typesetting, printing, and binding by Pantagraph Printing, Bloomington, Illinois Design by Chuck Thayer Advertising, San Francisco, California
Copies of articles that appear in the TESOL Quarterly are available through The Genuine Article , 3501 Market Street, Philadelphia, Pennsylvania 19104 U.S.A
@

VOLUMES MENU

TESOL QUARTERLY
CONTENTS ARTICLES
Language Minority Education in Great Britain: 385 A Challenge to Current U.S. Policy Sandra Lee McKay and Sarah Warshauer Freedman Preparing ESL and Bilingual Teachers for Changing Roles: 407 Immersion for Teachers of LEP Children Robert D. Milk The TOEFL Test of Written English: Causes for Concern Ann Raimes Student Input and Negotiation of Meaning in ESL Writing Conferences 443 Lynn M. Goldstein and Susan M. Conrad 461 Teaching the English Articles as a Binary System Peter Master Attitudes of Native and Nonnative Speakers Toward 479 Selected Regional Accents of U.S. English Randall L. Alford and Judith B. Strother
To print, select PDF page nos. in parentheses (10-30)

(32-51)

427

(52-67)

(68-85) (86-103) (104-120)

REVIEWS
497 Recent Publications in Sociolinguistics Sociolinguistics and Second Language Acquisition Dennis Preston Perspectives: Sociolinguistics and TESOL Nessa Wolfson Reviewed by Jessica Williams 501 Newbury House TOEFL Preparation Kit Preparing for the TOEFL Daniel B. Kennedy, Dorry Mann Kenyon, and Steven J. Matthiesen Preparing for the Test of Written English Liz Hamp-Lyons Reviewed by Marsha Bensoussan

BOOK NOTICES 507

Making it Happen: Interaction in the Second Language (Dorothy S. Messerschmitt) Classroom, Patricia A. Richard-Amato The Foreign Teaching Assistants Manual, Patricia Byrd, (Wanda Fox) Janet C. Constantinides, and Martha C. Pennington Languages and ChildrenMaking the Match, Helena Anderson Curtain (Noriko Isogawa) and Carol Ann Pesola Conversation Gambits: Real English Conversation Practices, (Teresa Granelli) Eric Keller and Sylvia T. Warner

Volume 24, Number 3 u Autumn 1990

(Eda Ashby) Crazy Idioms: A Conversational Idiom Book, Nina Weinstein Words at Work: Vocabulary Through Reading, Betty Sobel and Susan Bookman (Elliott L. Judd) Techniques and Resources in Teaching Grammar, Marianne Celce-Murcia and Sharon Hilles (Peter Master) A Linguistic Study of American Punctuation, Charles F. Meyer (K. Scott Ferguson) The Longman Preparation Course for the TOEFL, Deborah Phillips (Terese Thonus) Doublespeak, William Lutz (Vincent G. Barnes)

BRIEF REPORTS AND SUMMARIES

Listening Perception Accuracy of ESL Learners as a Variable Function of Speaker L1 519 George Yule, Susan Wetzel, and Laura Kennedy Using Brainstorming and Clustering with LEP Writers to Develop Elaboration Skills 523 Andrea B. Bermudez and Doris L. Prater

THE FORUM
Comments on James W. Tollefsons
Alien Winds: The Reeducation of Americas Indochinese Refugees

529 and Elsa Auerbachs Review Two Readers React . . . Donald A. Ranard and Douglas F. Gilzow The Reviewer Responds . . . Elsa Auerbach Response to Ranard and Gilzow: The Economics and Ideology of Overseas Refugee Education James W. Tollefson Comments on Martha C. Pennington and Aileen L. Young's Approaches to Faculty Evaluation for ESL 555 A Reader Reacts . . . Alastair Pennycook Response to Pennycook: The Political Economy of Information in TESOL Martha C. Pennington 569 Information for Contributors Editorial Policy General Information for Authors 573 Publications Received Publications Available from the TESOL Central Office 599 TESOL Membership Application

577

TESOL QUARTERLY

379

TESOL QUARTERLY

Editors Note
The diversity of interests represented among our readers continues to stimulate innovations within the format of the TESOL Quarterly. The next issue (Winter 1990) will inaugurate a new section: Research Issues. In this forum aspects of qualitative or quantitative research will be addressed, frequently from somewhat different perspectives by two specialists. I am fortunate that Graham Crookes (University of Hawaii at Manoa) has agreed to edit this section. Although contributions will typically be solicited, readers are encouraged to submit topic suggestions and/or make known their availability as contributors by writing directly to Graham Crookes at the following address: Graham Crookes Department of ESL University of Hawaii at Manoa 1980 East-West Road Honolulu, HI 96822 U.S.A. With regret I announce that other responsibilities have taken Linda Stolfi from her post as TESOL Quarterly Assistant Editor. It is impossible to fully acknowledge her contribution during the journals first year at the University of Washington. The editorial staff warmly thanks her for these contributions and for her work on Publications Received in this issue. I am indeed fortunate to introduce an able successor in Assistant Editor, Deborah Green. Deborah comes to the Quarterly with substantial editorial experience, having been manuscript editor of diverse publications ranging from Ramparts magazine to the international neurological journal Epilepsia. Finally, we cannot omit mention of the conflict that now engulfs the Middle East. The TESOL Quarterly staff wishes peace for our readers throughout the world. 381

In this Issue
Articles in this issue of the TESOL Quarterly challenge traditional educational wisdom in several arenas. The lead article examines British educational policies toward minority language education as a challenge to current U.S. policy. The second article describes an innovative teacher preparation course that challenges traditional training for ESL and bilingual teachers. The third article finds causes for concern in the TOEFL Test of Written English. Other articles reexamine claims made for writing conferences, systems for teaching the English articles, and attitude studies focusing on regional accents of English. Each article in this issue suggests approaches to complex issues facing TESOL professionals. Sandra McKay and Sarah Freedman examine the contrasting perspectives on language minority education in Great Britain and the United States. These differing approaches share a common rationale the protection of equality of opportunitybut result in different educational policies. British policies support mainstreaming students. In the U. S., policies favor separating nonnative speakers from their native-speaking peers. Acknowledging differences in educational context and history, the authors suggest that teachers in the U.S. examine British policies with an eye toward (re)evaluating U.S. policy for educating minority language speakers. Robert Milk describes a teacher training course designed to meet the different yet converging needs of bilingual and ESL teachers. ESL teachers receive an immersion experience in Spanish; bilingual teachers are provided an opportunity to enhance their proficiency in academic Spanish; and all receive an intensive simulated classroom experience in small-group, content-based instruction. This innovative program is designed to respond to the changing roles of ESL and bilingual teachers. Ann Raimes raises questions about the new TOEFL Test of Written English (TWE). She describes the institutional and programmatic contexts of the TWE within the Educational Testing Service and its TOEFL. Additionally, an historical overview of ETS composition tests for native speakers suggests a troubling pattern. Raimes presents seven areas of concern with respect to the TWE. These issues include topic type and selection, what the test measures, and even the necessity and utility of the test. She ends with seven recommendations for action by teachers. Lynn Goldstein and Susan Conrad investigate claims that writing conferences ensure student participation. The authors examined student input and negotiation of meaning in writing conferences between one teacher and each of three advanced ESL students. 382 TESOL QUARTERLY

Goldstein and Conrad conclude that there were large differences in the degree to which students participated in the conferences and negotiated meaning. These differences are reflected in subsequent drafts. When students had negotiated meaning, they made revisions that improved the text. Even active participation did not result in improved texts in the absence of negotiation of meaning. Peter Master offers a simplified schema for teaching the English article system. Master argues that English articles can be taught as a binary division between what he terms classification (a and ) and identification (the). His paper details shortcomings of previous approaches and outlines a pedagogical approach to this aspect of English grammar, notorious in its difficulty for the nonnative speaker. Randall Alford and Judith Strother investigate attitudes towards specific regional accents of U.S. English. Through the use of a modification of the matched guise technique, the authors compared reactions of native and nonnative speakers to male and female speakers from three U.S. accent groups: southern, northern, and Midwestern. The results indicated that the nonnative-speaker subjects were able to perceive differences in regional accents but that their reactions differed from those of the native-speaker listeners. Also in this issue: Reviews: Jessica Williams reviews recent publications in sociolinguistics: Dennis Prestons Sociolinguistics and Second Language Acquisition and Nessa Wolfsons Perspectives: Sociolinguistics and TESOL. Marsha Bensoussan reviews two texts from the Newbury House TOEFL Preparation Kit: Preparing for the TOEFL by Daniel Kennedy, Dorry Kenyon, and Steven Matthiesen; and Liz HampLyons Preparing for the Test of Written English. Book Notices Brief Reports and summaries: George Yule, Susan Wetzel, and Laura Kennedy find differences in the ability among different L1 learners to comprehend English language input from learners with different Lls; Andrea Bermudez and Doris Praters findings suggest that the graphic presentation of concepts (in this case through clustering) may help limited English proficient students expand their written discussions of reading material. The Forum: Donald Ranard and Douglas Gilzows commentary on James Tollefsons book Alien Winds and Elsa Auerbachs TESOL Quarterly review is followed by responses from the reviewer and the author; Martha Pennington responds to comments by Alastair Pennycook on her recent article with Aileen Young, Approaches to Faculty Evaluation for ESL. Sandra Silberstein IN THIS ISSUE 383

TESOL QUARTERLY, Vol. 24, No. 3, Autumn 1990

Language Minority Education in Great Britain: A Challenge to Current U.S. Policy

SANDRA LEE McKAY
San Francisco State University

SARAH WARSHAUER FREEDMAN

University of California, Berkeley

British educational policies advocate placing language minority students in mainstream classes where their regular teacher receives ongoing support from a TESOL specialist. By contrast, in the United States, the policies favor placing nonnative speakers in separate programs such as ESL pull-out classes, sheltered English, or bilingual education, where they are taught solely by the TESOL or bilingual education specialist. The same rationaleprotecting equality of opportunityis offered for both approaches. This article compares the events that led to the contrasting solutions and the institutional structures that support those solutions; it gives an example of the British mainstream system at work and shows how the different approaches to educating nonnative speakers reflect different assumptions about language development and definitions of equality of opportunity. The article concludes by asking language teachers three questions about programs for language minorities that are raised by the contrastive examination: (a) What are the consequences of social segregation in educational programs? (b) What are the effects of varied instructional contexts on language learning? (c) What are the most helpful roles ESL teachers can play with respect to teaching subject matter and linguistic competency?

Many parallels exist between the educational issues presented by language minorities in the United States and Great Britain. During the 1960s, both countries experienced a tremendous influx of immigrants with varied countries of origin. In both countries, new immigrants tended to settle in large industrialized urban areas for employment purposes. Because of this fact, the language minority student population in the urban centers increased tremendously. 385

School districts, however, were largely unprepared for this shift in demographics and had no language program in place for the new students. Since 1960 both countries have experimented with various types of educational programs to meet the needs of language minority students. At the present time the two countries seem to be moving toward very different conclusions as to the best model for the education of language minorities; while British policies tend to support mainstreaming (Department of Education and Science, 1965 [The Swann Report]), U.S. educational policies promote separate educational programs such as ESL pull-out programs, sheltered English, or bilingual education (in response to legislative acts such as Title VII and Supreme Court decisions such as Lau v. Nichols). What is ironic is that in both countries the same rationale is being offered for these very different approaches, namely, the rationale of protecting equality of opportunity for language minority students. What follows first is a framework for considering different language minority policies in both Britain and the United States. Then British language minority education policies since the 1960s are described, with the aim of demonstrating how social assumptions impact the making of educational policy. The British decision to place nonnative speakers in mainstream classrooms is discussed in the context of the British educational system, with its provision of language specialists working side-by-side with the subject matter teacher. To show how an ethnically and linguistically integrated classroom works in Britain, we provide a case study of a student learning in such a setting, illustrating the complexities of teaching nonnative speakers, who have come into a new cultural as well as a new linguistic context. We elaborate extensively on British policies for two reasons: First, British language policies are clearly articulated in comprehensive government reports; and second, only by a thorough presentation of British policy can we specify the challenge that these policies present to the United States. With the British context firmly in mind, we review the language minority policies in the United States since the 1960s and discuss the decisions that have resulted in separate programs for nonnative speakers in the United States. In conclusion, we provide a challenge to current U.S. policy as we pose several questions that educators need to examine before implementing any educational policy for language minorities. In this paper, the phrase language minority students will be used to describe immigrants (i.e., foreign-born children who emigrate with their parents), refugees (i.e., foreign-born citizens who enter a country under special conditions), and long-term residents who 386 TESOL QUARTERLY

come from non-English-speaking homes. Language minority students who lack proficiency in English will be referred to as language minority/limited English proficient (LM/LEP). Throughout the paper, we will refer to three social attitudes toward policy planning for language minority groups (Ruiz, 1988): language-as-problem, language-as-right, and language-as-resource. According to Ruizs framework, a society with a language-asproblem perspective views language minority students as having a linguistic deficiency that can best be remedied by replacing the native language with the dominant language, e.g., English. A society that adheres to a language-as-right perspective promotes the rights of language minorities to maintain their native language on legal grounds. Finally, a society with a language-as-resource perspective regards the languages spoken by language minorities as a national resource; and thus, educational policies are designed to maintain and develop native languages. These social orientations toward linguistic diversity have been exhibited in various educational language policies in both Britain and the United States from the 1960s to the present, as will be evident from the historical overview of changing language policies in both countries. EVOLVING LANGUAGE POLICIES IN GREAT BRITAIN While Britain, like the United States, has a long history of immigration, it was only beginning in the early 1950s that speakers of many languages came to settle in Britain in significant numbers all at the same time. These immigrants were mainly refugees from Eastern Europe, East Africa, and Southeast Asia, and labor migrants from Southern and Eastern Europe, and from former British colonies in South and East Asia and the Caribbean (Martin-Jones, 1989). Since these immigrants tended to settle in large urban industrialized areas, there has been, since the 1950s, a steady increase in the number of LM/LEP students in such areas. For example, while in 1978 the inner London area had only 10% LM/LEP students, by 1983 these students comprised 23%. In 1983 LM/LEP students represented 172 different languages with only 14 of these languages spoken by more than 100 students (Martin-Jones, 1989). In spite of tendencies in Great Britain toward a nationally centralized system of education, with a long tradition of national examinations and now the new national curriculum, British school districts, called Local Education Authorities (LEAs), have, according to Martin-Jones (1989), considerable autonomy in policy formulation and curriculum development within their area. Policy guidelines for LM/LEP students issued by the central government LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 387

through the Department of Education and Science (DES) have no mandatory force, although, increasingly, financial controls are centrally imposed and these, in turn, have an impact on local autonomy (p. 9). Early policies viewed LM/LEP students as social problems, and decisions about their education were based on what was perceived as best for the Anglo majority. During the 1960s, one of the first programs local school districts established for LM/LEP students provided separate language centers, termed induction centres, for LM/LEP students. According to Reid (1988), LM/LEP students were separated from their English-speaking peers ostensibly so that they could be taught English to a level which would allow them to join classes in ordinary schools, but also, of course, to satisfy majority parents that their children would not be held back by the presence of large numbers of immigrant children in the same classes. (p. 187) The Department of Education and Science, meanwhile, advocated a policy of dispersal or busing since parents in areas where there were large concentrations of LM/LEP pupils were complaining about the emergence of black majority schools (Martin-Jones, 1989, p. 44). Because of these complaints the Department of Education and Science issued a set of guidelines in 1965 for what they called the dispersal of minority children. The guidelines for this policy presented the following rationale: Experience suggests . . . that, apart from unusual difficulties (such as a high proportion of non-English speakers), up to a fifth of immigrant children in any group fit in with reasonable ease, but that, if the proportion goes over about one third, either in the school as a whole or in any one class, serious strains arise. It is therefore desirable that the catchment areas of schools should, wherever possible be arranged to avoid undue concentrations of immigrant children. Where this proves to be impracticable simply because the school serves an area which is occupied largely by immigrants, every effort should be made to disperse the immigrant children round a number of schools and to meet such problems of transport as may arise. (Department of Education and Science, 1965, pp. 4-5, as cited in Martin-Jones, 1989, pp. 44-45) Of particular significance is the fact that the promotion of this dispersal policy was made purely on the basis of an untested social assumption, namely, that if the immigrant population in a particular school were allowed to exceed one third, serious strains (Department of Education and Science, 1965, as cited in MartinJones, 1989, p. 45) would arise. Determining language policies on the basis of unchallenged social assumptions is, as we shall see, a 388 TESOL QUARTERLY

common pattern throughout United States and British minority education history. Accompanying the view that these children present social problems is the view that their language, too, is a problem. In 1971, the Department of Education and Science issued a national policy document clearly exemplifying a language-as-problem perspective of minority languages: If there is any validity in Bernsteins view that the restricted code of many culturally deprived children may hinder their ability to develop certain kinds of thinking, it is certainly applicable to non-English speaking children who may be suffering, not only from the limitation of a restricted code in their own language, but from the complication of trying to learn a second language. Experiencing language difficulties, they may be suffering handicaps which are not conspicuous because they concern the very structure of thought. (Department of Education and Science, 1971, p. 9, as cited in Martin-Jones, 1989, pp. 45-46) A major and public challenge to the language-as-problem perspective occurred in 1975 with the publication of what is known as the Bullock Report. This central government report was produced by a committee of inquiry whose primary purpose was to investigate native-speaking childrens language development across the school years. However, in the chapter on the language needs of LM/LEP children entitled Children from Families of Overseas Origin, the committee argued that in a linguistically conscious nation in the modern world, we should see it [mother tongue] as an asset, as something to be nurtured, and one of the agencies that should nurture it is the school. Certainly the school should adopt a positive attitude to its pupils bilingualism and whenever possible should help to maintain and deepen their knowledge of their mother tongues. (Department of Education and Science, 1975, p. 294) Ironically, after the publication of the report, few programs were established to promote native language maintenance even though the rhetoric of the report suggested that this should be done, illustrating a discrepancy between policy recommendations and the implementation of these recommendations. DECISION FOR MAINSTREAMING In 1985, a second major educational policy statement regarding LM/LEP students was issued with the publication of the Department of Education and Sciences report, Education for All, commonly known as the Swann Report. This report was prepared by a national committee whose task was solely to examine LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 389

educational policies for language minority students. Whereas the Bullock Committee considered looking at issues such as mainstreaming outside their scope, they did assert, Common sense would suggest that the best arrangement is usually one where the immigrant children are not cut off from the social and educational life of a normal school (Department of Education and Science, 1975, p. 289). The Swann Report went one step further and strongly endorsed the mainstreaming of LM/LEP students: We are wholly in favour of a move away from E2L [English as a second language] provisions being made on a withdrawal basis, whether in language centres or separate units within schools (Department of Education and Science, 1985, p. 392). The Swann Report argued that withdrawal classes establish and confirm social and racial barriers between groups and whilst not originally discriminatory in intent were discriminatory in effect because they deny children access to the full range of educational opportunities available . . . by requiring them to miss a substantial part of the normal school curriculum (p. 389). The report argued strongly that the informal interaction that occurs in schools is as important for language development as the formal context of language development and thus, that it is important for LM/LEP students to be placed in a context where they could interact with native speakers. Mainstreaming was viewed as offering an opportunity for all teachers to consider the language demands of the work they do with all children in the classroom, whatever the language background (Martin-Jones, 1989, p. 52). The Swann Report did not support bilingual education principally on the grounds that to implement it, minority children would have to be segregated. They feared that this might highlight differences and have a detrimental effect on race relations (Edwards, Moorhouse, & Widlake, 1988, p. 81). While the report argued that Local Education Authorities should make school buildings available for native language instruction, the Swann Committee viewed the maintenance and development of LM/LEP students native language as a responsibility of the ethnic community itself rather than the school. The committee argued that by putting LM/LEP children in mainstream classes, schools could provide a framework for promoting a pluralistic society: We also see education as having a major role to play in countering the racism which still persists in Britain today and which we believe constitutes one of the chief obstacles to the realization of a truly pluralistic society. We recognize that some people may feel that it is expecting a great deal of education to take a lead in seeking to remedy 390 TESOL QUARTERLY

what can be seen as a social problem. Nevertheless we believe that the education system and teachers in particular are uniquely placed to influence the attitudes of all young people in a positive manner. (Department of Education and Science, 1985, p. 319) The Swann Report has sparked substantial debate. The major criticisms have come from advocates of instruction in the students first language (see, for example, Khan, 1985; National Council for Mother Tongue Teaching, 1985). First, the critics challenged the reports definition of pluralism, arguing that the report, by not advocating native language instruction in the schools, was promoting a type of linguistic assimilation in which the ability to speak English was equated with being British (National Council for Mother Tongue Teaching, 1985). Advocates of instruction in the native language lamented the fact that the Swann Report offered no support for the earlier recommendation of the Bullock Report for native language instruction in the schools (see, for example, Devall, 1987). In essence, the critics viewed the Swann Report as presenting a language-as-problem perspective. The critics further argued that the Swann Report failed to recognize the important link between first and second language development. Pointing to bilingual programs in the United States and Scandinavian countries and to the work of Cummins (1982, 1984), critics argued that the report ignored the important role that first language maintenance can have in both cognitive development and in the acquisition of a second language. In addition, proponents of instruction in the native language viewed the development of LM/LEP childrens first language as a way of promoting a truly pluralistic society in which government policies actively promoted linguistic pluralism. Finally, proponents of native language instruction criticized the Swann Report for its failure to see the intimate connection between language and culture. Critics argued that in failing to recognise the intrinsic links between language and culture, the Report does not perceive the centrality of language in culture, in the development of ethnicity and of the individuals cultural identity. At the very outset of the Report, ethnic identity is described by stressing a physical attribute of raceskin colorrather than the social attribute of language. (National Council for Mother Tongue Teaching, 1985, p. 501) More recently, support for the Swann Reports negative stance toward bilingual education has come from the Kingman Report (Department of Education and Science, 1989), authored by the conservative forces currently controlling education in Great Britain, who contend that placing language minority students in mainstream LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 391

classes benefits all students awareness of language. This recent report, which outlines a national curriculum in English, maintains that bilingual children should be considered an advantage in the classroom rather than a problem. The evidence shows that such children will make greater progress in English if they know that their knowledge of their mother tongue is valued, if it is recognised that their experience of language is likely to be greater than that of their monoglot peers and, indeed, if their knowledge and experience can be put to good use in the classroom to the benefit of all pupils to provide examples of the structure and syntax of different languages, to provide a focus for discussion about language forms and for contrast and comparison with the structure of the English language. (p. 10.12) While the authors of the Swann and Kingman Reports and advocates of bilingual education disagreed on important issues, all accepted the idea that ethnic pride and cultural respect should be central concerns in the formulation of a language policy. All shared the idea of promoting an ethnically pluralistic society, but for the Swann and Kingman Committees this pluralism meant promoting cultural pluralism in mainstream classrooms, while for proponents of bilingual instruction this pluralism meant developing linguistic pluralism even if it resulted in cultural segregation. What is significant, however, is that in all instances a discussion of the relationship between ethnicity and language programs was considered necessary to the educational decision-making process. The Role of the Language Specialist in the Mainstream Classroom In his summary of linguistic minorities and language education in England, Reid (1988) points out that today separate ESL classes and learning materials are becoming increasingly rare; they are being replaced by English Support for Bilingual Learners, provided in the context of mainstream classes at both primary and secondary school level; or, very recently, by collaborative learning or team teaching. (p. 189) When LM/LEP students are placed in mainstream classes, there is a call for close collaboration between ESL teachers, who are called support teachers, and the subject specialists. In the British educational context, language specialists or support teachers of LM/LEP students play a role unfamiliar in U.S. schools. As regularly certified teachers who have returned to the university or to a teacher training college for a postgraduate degree, the support 392 TESOL QUARTERLY

teachers function as resource teachers; however, instead of pulling students out of the regular classroom, they go into the mainstream classroom to help the subject-matter specialist teacher teach the LM/LEP students. The language specialist helps both in language instruction and in providing LM/LEP students with the support necessary for meeting the normal demands of subject-matter instruction. Riley and Bleach (1985) explain the benefits of the language and subject-matter teachers working together: The development of co-operative teaching looks to be central. It is more stimulating and a good learning situation for both teachers and for children. No matter how gifted the class teachers are, how much language knowledge they have, or how good their initial training has been, the full responsibility for the language learning and total education of developing bilingual pupils should not rest with classroom teachers unsupported. If responsibility is taken away from them, they can never begin to develop their classrooms as places where bilingual pupils have an equal right to learning and being. The same is true of ESL specialists operating in a separatist structure. Co-operative teaching is not the sticking together of two pedagogues, but the development of something new. Co-operative teaching and the taking of responsibility for developing bilingual pupils by the whole school means that from reception stage onwards pupils can be supported over much longer phases of their learning and across all language modes. Literacy can be developed earlier and more consistently, and the students will then have this, as well as spoken means, as an impetus for further language development. (p. 88) Britain maintains well-established postgraduate programs for training language support teachers. For readers interested in a detailed discussion, Levine (1985) describes the program at the University of London, Institute of Education. A British Mainstream Classroom at Work What does the British mainstream classroom look like, and how do LM/LEP students learn in this context? In his essay, Khasrus English Lesson: Ethnocentricity and Response to Student Writing (1990), Alex Moore provides one example. Moore has written sensitively about Khasru, a Bangladeshi boy learning to write in a British mainstream classroom. Moore raises issues about Khasrus needs that transcend the specifics of the teaching context and shows how a teachers ethnocentricity can cause communication problems with an LM/LEP student independent of the classroom model. Khasru has been in England for less than two years. He is in his fourth year of secondary school, the U.S. equivalent of ninth grade, LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 393

and the first of two years during which British students prepare for the national examinations that they must pass in order to graduate from secondary school. His teacher has read the class a love story and has asked the students to write their own love stories for their examination folders. Khasrus first draft begins: once aponar time I fund a grill and I ask har exquiseme wher you going she said? I went to go some way wher you ask me for. I said No I-s Just Ask you you going I am sory about that have you dont mind she said thats OK and anther I fund har on the buse and is was set on the Front and she was set on the back about 4 Five Minuts ago two bay was come And ther

Khasru continues by describing going over to the girl who then asks for his help. They get off the bus together, but she is too afraid to walk home alone, so Khasru agrees to help her. During their walk home she declares her love for him, and he says that he loves her too. They then discuss their siblings at some length, and Khasru concludes, Now we go every day. Moore explains: There is a support teacher in Khasrus class, who sits with Khasru to work with him on this preliminary draft. This support teachers corrections are of two kinds. First, there is a concentration on the production of acceptable Standard English sentences, spellings, punctuation, and paragraphing; on presenting the story so that it makes immediate sense to any reader; and on helping Khasru with obvious confusions. . . . The second set of corrections, made simultaneously with the first, relate to Khasrus storytelling style . . . [e.g.,] Lets get rid of some of these ands. (p. 2) After three sessions with the support teacher, Khasrus second draft shows dramatic improvements in the acceptability and accessibility of the language and in sentence-level grammar: Once upon a time I saw a girl and I asked her, Where are you going? She said Im just going somewhere. What are you asking for? Do you want to know for any special reason? I said No. I was just asking where you were going. Im sorry. I hope you dont mind. She said Thats okay. Afterwards, I saw her on the bus. I was sitting at the front and she was at the back. After about five minutes, two boys got on. They sat at the back near the girl and one of them said to her Hello. Where are you going? 394 TESOL QUARTERLY

She was scared, and the boys tried to do something bad to her. (p. 2) The rest of the story continues in this vein. At this point in Khasrus process, Moore concludes: The omens at this stage are good. Khasru is clearly pleased about his work so far, and his showing it to other Bangladeshi boys in the class seems to have had the effect of encouraging them to take their own stories more seriously. Khasru is fortunate, too, to have one teacher who can work with him on a one-to-one basis, apparently for as long as is necessary to complete each stage of the project: not just any teacher, either, but one committed to a multicultural approach to teaching that, to use his words condemns the Eurocentrism that has afflicted compulsory education in this country since its inception. (p. 25) Moore quotes the support teacher who explains why he thinks bilingual students should be in mainstream classes in mainstream schools (p. 25): Of course they need to be in the mainstream classes: they need to read, listen to, and join in with the languages and behaviors of their English peersand they need that sort of audience and feedback for their work. They need to know, and deserve to know, that were taking them seriously: seriously enough to listen to what theyve got to say, and to give them the sort of space and opportunities we give to every other kid in the school. (p. 25) Khasrus draft again goes to the support teacher who again sits beside him to discuss further possible improvements before Khasru moves on to revise again. Problems surface, however, when the support teacher, in discussing this second draft, questions Khasrus content, asking about the suddenness of the declaration of love and about the talk about siblings. When the support teacher suggests all this stuff about relations . . . This isnt really necessary, is it . . . For the reader . . . What do you think? he is met with silence from Khasru (pp. 25-26). The support teacher then asks Khasru if people would really talk this way: Do people talk that way? In real life? Do they talk about how old their brothers and sisters are? Khasru replies, Yes, Sir. Then the support teacher responds, DO you think so? Im not so . . . (p. 26). At the end of this session the support teacher instructs Khasru: Well, take it home with you, Khasru, think about what weve said, and see if you can make Chapter 1 any better (p. 27). Khasru becomes confused. He has been asked to write a story that is true to life, but when he does so, he is told that what he writes is not really true to life. Khasru stops working on the story and never completes it. LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 395

Moore concludes that the support teacher, at this point, albeit unintentionally, is imposing his reality on Khasrus writing. Further, the support teacher assumes there is a way or set of ways of talking to one another and a way or set of ways of telling a storyin both cases, traditional English ways. . . . This leaves no room for the possibility of linguistic diversity in the broadest sense, that embraces genre, perception, and form, and that is suggested by the whole-school policywhich on one level the teacher supports. (p. 26) Moore warns: There is a very real danger that such children [as Khasru] will grow up not thinking Yes, they do and see things differently here, but Yes, they do and see things properly hereand that consequently, schoollearning will always be that much harder for them: for it is surely easier to learn new ways that are set into a framework where they can coexist with existing ways than it is to learn new ways that must simply replace old ones; psychologically, the problem is very different. . . . schools must clearly work hard to develop and to adopt new styles of pedagogy: styles that will encourage the development of required expertise without promoting the corresponding, and all too prevalent, loss of faith. (p. 27) Khasrus experience shows how a piece of writing evolved in a British mainstream context, with the support teacher helping the students in a regular class achieve regular curricular goals, in this case preparing for the national examination. It also serves as a cautionary tale about the potential effects of unintended ethnocentric response to student writing by teachers of LM/LEP students, whether these students are in a mainstream or a separate classroom context. CHANGING LANGUAGE POLICIES IN THE UNITED STATES DIFFERENT DECISIONS In the United States, educational and government leaders who favor programs that take LM/LEP students out of regular classes argue that these programs are necessary to support students language development. Unlike their British counterparts, they rarely address the potential social effects of these programs cultural isolation, segregation, and racism. In order to understand the different emphases that underlie United States and British education policies for LM/LEP students, we turn now to the United States language minority policies since the 1960s. Like Britain, the United States experienced a large increase in immigrants during the 1960s, largely due to the change in 396 TESOL QUARTERLY

immigration laws of 1965, which abandoned the national origins quota system and gave preference instead to family reunification and occupational skills. As in Britain, these recent immigrants tended to come from varied countries of origin and to settle in large industrialized urban centers. In the sixties, urban schools in the United States, as in Britain, were faced with a large influx of nonnative speakers of English with very diverse language backgrounds. As in Britain, local school districts in the United States have a great deal of autonomy. State and local governments have the primary responsibility for funding and developing policies for public elementary and secondary schools. According to Rotberg (1984), the limited educational funding that comes from the federal government is intended to increase equality of educational opportunity by providing additional resources for areas of the country and for population groups with special needs (p. 134). The United States has little comparable to the Bullock, Swann or Kingman Reports, which set forth national language policies for LM/LEP students. Rather United States policies develop from constitutional, statutory, or judicial sources. As Wong (1988) points out, most LM/LEP programs have arisen from legal issues regarding the entitlement of LM/LEP students to language education services. The primary constitutional basis for LM/LEP services is the first section of the Fourteenth Amendment, which states that No state shall . . . deny to any person within its jurisdiction the equal protection of the laws. The major statutory bases for LM/LEP language education services are Title VI of the Civil Rights Act of 1964, The Equal Educational Opportunities Act of 1974, and Title VII of the 1968 Elementary and Secondary Education Act (also known as Title VII or the Bilingual Education Act). The Civil Rights Act (Section 601), as cited in Wong (1988), states: No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance. (p. 372) The Equal Educational Opportunities Act (Section 170 (f)) states: No state shall deny equal educational opportunity to an individual on account of his race, color, sex, or national origin, by. . . the failure of an educational agency to take appropriate action to overcome language barriers that impede equal participation by its students in its instructional programs. (p. 372) These two acts, along with the Fourteenth Amendment, are used to LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 397

argue for language education programs for LM/LEP students on the basis of equal protection under the law. As we shall see, the issue that has been argued in applying these rights to LM/LEP students is whether equality is to be interpreted as equality of access or equality of outcome. The third significant statutory basis for language programs for LM/LEP students is Title VII. As Hakuta (1986) notes, [Title VII] heralded the official coming of age of the federal role in the education of persons with limited English-speaking ability. Seven and a half million dollars were appropriated for the 1969-1970 fiscal year, to support experimental programs responsive to the special educational needs of children of limited English-speaking ability in schools having a high concentration of such children from families . . . with income below $3,000 per year (Bilingual Education Act, 1968). (p. 198) Rotberg (1984) cites the language of the Title VII Program to note that the original purpose was to encourage the use of bilingual educational practices, techniques and methods (p. 134). However, in 1983, Secretary of Education Terrell Bell proposed amendments that were designed to give school districts greater flexibility in their choice of instructional approaches, so that instruction in LM/LEP students native language would no longer be required for Title VII funds (Rotberg, 1984, p. 135). From the beginning, the majority of programs funded under this piece of legislation have been transitional in nature, with LM/LEP students native languages regarded as a problem rather than a resource. As Ruiz (1988) points out, the Bilingual Education Act (BEA) of 1968 and the state statutes which have followed start with the assumption that non-English language groups have a handicap to overcome; the BEA, after all, was concerned and formulated in conjunction with the War on Poverty. Resolution of this problemteaching English, even at the expense of the first languagebecame the objective of the school programs now generally referred to as transitional bilingual education. (p. 7) The major judicial foundation for LM/LEP language education programs is the 1974 Lau v. Nichols Supreme Court decision. In this case, the parents of 12 LM/LEP Chinese American students filed a class action suit against the San Francisco Unified School District arguing that they had been denied an education because of a lack of language classes with bilingual teachers. Two of the main legal issues dealt with in the case were equality of access versus equality of outcome and discriminatory intent versus discriminatory impact (Wong, 1988). Although in previous Supreme Court decisions regarding equal 398 TESOL QUARTERLY

educational opportunity, such as in the 1954 Brown v. Board of Education Case, the Court found a denial of equal protection only where the state has made different provisions for similarly situated citizens without adequate justification (Grubb, 1974, as cited in Wong, 1988, p. 374), in Lau v. Nichols the Court ruled that, although the LM/LEP students had been given equality of access to the regular classroom, they had been denied equality of outcome because they did not have the necessary language background to benefit from the program. As quoted in Wong (1988), the Court decided: There is no equality of treatment merely by providing students with the same facilities, textbooks, teachers, and curriculum; for students who do not understand English are effectively foreclosed from any meaningful education. Basic English skills are at the very core of what these public schools teach. Imposition of a requirement that, before a child can effectively participate in the educational program, he must already have acquired those basic skills is to make a mockery of public education. We know that those who do not understand English are certain to find their classroom experiences wholly incomprehensible and in no way meaningful. (p. 378) The second issue addressed in the case was the issue of discriminatory intent versus discriminatory impact. The Court argued that placing non-English-speaking students in the regular classroom was discriminatory in effect while not discriminatory in intent because LM/LEP students did not have the basic skills needed to function in the regular classroom. The Court argued that some program must be devised for LM/LEP students other than to leave them in the regular classrooms, but it left the implementation of the remedy to the local school boards (Wong, 1988). According to the decision: No specific remedy is urged upon us. Teaching English to the students of Chinese ancestry who do not speak the language is one choice. Giving instructions to this group in Chinese is another. There may be others. Petitioners ask only that the Board of Education be directed to apply its expertise to the problem and rectify the situation (Teitelbaum & Hiller, 1977, as cited in Hakuta, 1986, p. 201). United States educational policy has tended to interpret this directive to mean that some type of language development must occur before an LM/LEP student is placed in the regular classroom. In fact, according to the decision, the placing of LM/LEP students in regular classrooms without support services would be a violation of fundamental rights (Wong, 1988). In Britain, however, the LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 399

current educational policy of mainstreaming assumes that the development of language skills of LM/LEP students can best occur while they are in regular classes, if some type of language support service is provided. Indeed, the Swann Report argued that any solution that would require withdrawing the students from the regular classroom was discriminatory in effect if not discriminatory in intent. CONFLICTING ASSUMPTIONS What is the basis for such differing perspectives between the two countries? At issue is a definition of what type of equality of opportunity is being considered. In Lau v. Nichols, the issue was the question of equality of opportunity in reference to language skills. Linguistic equality, the Court seemed to suggest, was the primary issue since LM/LEP students would not experience equality of outcome unless they acquired those basic skills referred to in the decision. The fact that special programs dealing with linguistic inequality can result in racial segregation has not been raised as a challenge in the courts even though the basis for the Lau decision was Title VI of the Civil Rights Act. By focusing on equality in terms of linguistic opportunities, the Supreme Court argued that Chinese-American, non-English-speaking students were denied equal educational opportunity under Title VI of the Civil Rights Act when instructed in English, a language they did not understand (Rotberg, 1984, p. 135). One of the few expressions of concern about the matter of racial segregation in LM/LEP language programs came from the 1974 American Institutes for Research evaluation report for Title VII programs. It found that often students were assigned to Title VII Spanish-English classes not on the basis of their proficiency in English, but rather on their ethnic background (Rotberg, 1984). To avert the segregation that could arise from assigning students to classes on the basis of ethnic background, the 1978 Title VII Amendments dealt with the issue in the following manner: In order to prevent the segregation of children on the basis of national origin in programs assisted under this title, and in order to broaden the understanding of children about languages and cultural heritages other than their own, a program of bilingual instruction may include the participation of children whose language is English, but in no event shall the percentage of such children exceed 40 per centum. (U.S. Congress, 1978, as cited in Rotberg, 1984, p. 141) However, striving to minimize segregation by placing students 400 TESOL QUARTERLY

whose native language is English in bilingual classes is quite different from the philosophy underlying the Swann Report. It recommended that racial integration be maintained at all costs in all classrooms even if it results in a lack of support for bilingual maintenance programs. While the Title VII Amendments express concern about the problem of possible segregation caused by special language programs, there are no documents in the United States comparable to the Swann Report, which argues that only language programs adhering to racial integration are acceptable. The contrasting language policies of the United States and Britain rest on very different pedagogical and social assumptions. In the United States, the current policy of removing LM/LEP students from regular classes rests on a definition of equality of opportunity as linguistic opportunity in which the development of English language skills is taken to be primary, even if the language programs result in racial segregation. This view often results in language programs in which LM/LEP students learn English in classes without a large number of native speakers present. In Britain, on the other hand, advocates of the Swann Report equate educational opportunity with the idea of social equality and racial integration, even if this integration results in a lack of support for the native language. Language programs for LM/LEP students are to be undertaken in the mainstream classroom where there are a large number of native speakers. The different definitions of equality of opportunity evident in U.S. and British language minority programs provide a framework for re-examining the social and linguistic assumptions language teachers wish to make regarding language programs for LM/LEP students. The authors support, as does Rex (1988), the idea that the first step in designing any social or educational program is to make value standpoints clear and explicit in order to demonstrate what the system is achieving and failing to achieve (Rex, 1988, p. 219). In his review of British language minority programs, Rex begins by citing the work of Gunnar Myrdal (1944) and his classic study of U.S. race relations. He points out that when Myrdal was asked to undertake a study of race relations in the United States, he argued that social scientists need to state the goals they wish to achieve so that they can then determine what practices are conducive to the attainment of those goals (p. 205). In the tradition of Myrdal, the authors suggest that, as language teachers, we state our goals and value standpoints on language minority programs clearly before we make any recommendations regarding particular programs for LM/LEP students. We urge a careful examination of the following questions: LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN 401

1. What are our views on social segregation in educational programs? How does social segregation rank in our order of priorities in determining language policies for LM/LEP students? Programs that separate LM/LEP students from mainstream classes often result in social segregation. Are the benefits of separate programs greater than any potential negative effects of social segregation? 2. What are our views on language learning? How high on our priority list is interaction with native speakers in promoting linguistic development? Separate language programs minimize the LM/LEP students opportunity to interact with native speakers. Are the benefits of separate programs greater than what might occur if planned interaction with native speakers were to occur in mainstream classrooms? 3. What are our views on the role of language teachers? Do we see our role as primarily one of developing linguistic competency in order to promote content learning, or do we see our role as one of using subject content as a vehicle for developing linguistic competency? If we support the latter role, what benefits exist in developing language and content learning in separate classrooms rather than in mainstream classrooms in collaboration with subject teachers? All of these questions need to be addressed and seriously examined as language teachers evaluate different types of language development programs for LM/LEP students. In the end, policy makers may advocate separate language programs for LM/LEp students or may, like Britain, find that there are benefits to promoting mainstream programs with carefully crafted systems of language support. If, for example, U.S. teachers were to provide language support within the mainstream context, classrooms would likely have to be reorganized to allow for individualized help. Freedman and McLeod (1988) conducted a comparative study of English teaching in the U.S. and the U.K. Through national surveys and classroom observations in both countries, they found that British teachers of English are more likely to individualize instruction while U.S. teachers are more likely to concentrate on whole-group teaching. Classroom contexts that provide for individualized teaching make it possible to handle the diversity of needs within a mainstream class. Our goal with this contrastive examination of national language policies is to raise key issues. Given the differences in educational contexts and educational histories in the two countries, it is not 402 TESOL QUARTERLY

surprising that the approaches vary. What is essential before taking a position either for mainstreaming or for separate programs is to clarify our assumptions and values regarding social integration and language learning so that, as Myrdal suggests, there is a basis for assessing what is or is not being achieved.

ACKNOWLEDGMENTS
We wish to thank the following people for discussing the ideas in the manuscript and helping us to clarify our ideas: Marilyn Martin-Jones, Alex McLeod, and Guadalupe Valds. We also wish to thank the anonymous TESOL Quarterly reviewers for their helpful suggestions. Part of the research reported here was conducted pursuant to a grant from the Office of Educational Research and Improvement/U.S. Department of Education (OERI/ED). The opinions expressed do not necessarily reflect the position or policy of the OERI/ED, and no official endorsement by the OERI/ED should be inferred.

THE AUTHORS
Sandra Lee McKay is Professor of English at San Francisco State University. Her most recent publication, Language Diversity: Problem or Resource?, coauthored with Sau-ling Wong (Newbury House, 1988), presents a social and educational perspective of recent language minority groups in the United States. She recently returned from a teaching exchange at the University of Manchester. Sarah Warshauer Freedman is Professor of Education at the University of California, Berkeley, and is Director of the Center for the Study of Writing. Her latest book is Response to Student Writing (National Council of Teachers of English, 1987). Her research interests include literacy learning for multicultural populations.

REFERENCES
Cummins, J. (1982). Mother tongue maintenance for minority language children: Some common misconceptions. Toronto: OISE Press. Cummins, J. (1984). Bilingualism and special education: Issues in assessment and pedagogy. Clevedon, Avon, England: Multilingual Matters. Department of Education and Science. (1975). A language for life. (The Bullock Report). London: Her Majestys Stationery Office.

LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN

403

Department of Education and Science. (1985). Education for all: Report of the Committee of Inquiry into the Education of Children from Ethnic Minority Groups. (The Swann Report). London: Her Majestys Stationery Office. Department of Education and Science. (1989). English for ages 5 to 16: Proposals of the Secretary of State for Education and Science and the Secretary of State for Wales. (The Kingman Report). London: Her Majestys Stationery Office. Devall, S. (1987). Affirmative action and positive discrimination. In G. Haydon (Ed.), Education for a pluralist society (Bedford Way Paper No. 30, pp. 85-93). London: University of London, Institute of Education. Edwards, C., Moorhouse, J., & Widlake, S. (1988). Language or English? In M. Jones & A. West (Eds.), Learning me your language: Perspectives on the teaching of English (pp. 77-95). London: Mary Glasgow. Freedman, S. W., & McLeod, A. (1988). National surveys of successful teachers of writing and their students in the U.S. and the U.K. (Technical Report No. 14). Berkeley, CA: Center for the Study of Writing. Hakuta, K. (1986). Mirror of language: The debate on bilingualism. N e w York: Basic Books. Khan, V. S. (1985). Language education for all? Chapter 7 of the Swann Report. London: University of London Institute of Education, Centre for Multicultural Education. Levine, J. (1985). On the training of teachers. In C. Brumfit, R. Ellis, & J. Levine, (Eds.), English as a second language in the United Kingdom (pp. 141-148). Oxford: Pergamon press. Martin-Jones, M. (1989). Language education in the context of linguistic diversity: Differing orientations in educational policy. In J. H. Esling (Ed.), Multicultural education and policy: ESL in the 1990s (pp. 36-58). Toronto: OISE Press. Moore, A. (1990). Khasrus English lesson: Ethnocentricity and response to student writing. The Quarterly, 12 (1), 1-3,25-27. Myrdal, G. (1944). An American dilemma: The Negro problem and modern democracy. New York: Harper & Row. National Council for Mother Tongue Teaching. (1985). The Swann Report: Education for all? Journal of Multilingual and Multicultural Development, 6 (6), 497-508. Reid, E. (1988). Linguistic minorities and language educationThe English experience. Journal of Multilingual and Multicultural Development, 9 (l&2), 181-191. Rex, J. (1988). Equality of opportunity and the ethnic minority child in British schools. In S. Modgil, G. Verman, K. Mallick, & C. Modgil (Eds.), Multicultural Education (pp. 205-219). London: Falmer Press. Riley, S., & Bleach, J. (1985). Three moves in the initiating of mainstreaming at secondary level. In C. Brumfit, R. Ellis, & J. Levine (Eds.), English as a second language in the United Kingdom (pp. 77-89). Oxford: Pergamon Press. 404 TESOL QUARTERLY

Rotberg, I. (1984). Bilingual education policy in the United States. Prospectus, 14 (1), 133-147. Ruiz, R. (1988). Orientations in language planning. In S. McKay, & S. Wong (Eds.), Language diversity: Problem or resource? (pp. 3-25). New York: Newbury House. Wong, S. (1988). Educational rights of language minorities. In S. McKay & S. Wong (Eds.), Language diversity: Problem or resource? (pp. 367-386). New York: Newbury House.

LANGUAGE MINORITY EDUCATION IN GREAT BRITAIN

405

TESOL QUARTERLY, VoL 24, No. 3, Autumn 1990

Preparing ESL and Bilingual Teachers for Changing Roles: Immersion for Teachers of LEP Children
ROBERT D. MILK
The University of Texas at San Antonio

With increased emphasis on integration of language and contentarea instruction, the roles of bilingual and ESL teachers are becoming increasingly interrelateda situation that calls for development of common training experiences in the preparation of ESL and bilingual personnel. This article describes a teacher training course designed to meet both the differing language proficiency needs of bilingual and ESL teachers, as well as the common needs of teachers learning to implement content-based strategies for teaching language. Specifically, (a) ESL specialists receive an immersion experience in Spanish, (b) bilingual specialists are provided opportunities to enhance their proficiency in academic Spanish, and (c) both ESL and bilingual specialists receive intensive simulated classroom experiences in small-group, content-based instruction following a cooperative learning approach. A rationale for following an integrated approach in the preparation of language educators for limited English proficient (LEP) children is presented, and data collected from participants in the course are discussed in relation to the potential effectiveness of this type of teacher training format as a vehicle for attaining important teacher preparation goals.

Classroom teachers have often noted the ironic (if not contradictory) mismatch between the kinds of suggestions and directives they commonly receive from experts on pedagogy and the manner in which these suggestions are delivered. Thus, it is not unheard of for elementary school teachers to be lectured on the limitations of the lecture method or to receive information in a large group on the wonders of small-group instruction. Within language education, those responsible for the preparation of teachers have, in recent years, struggled with some of the challenges posed by the need to 407

achieve a greater coherence between the kinds of innovations proposed by methodologists and the means through which these ideas are presented to teachers-in-training. For language educators involved specifically in the teaching of ESL within bilingual education programs, the need to achieve a greater coherence between evolving trends in classroom practice and the procedures typically followed in university coursework is marked. A prominent theme running through much of the recent literature on effective instructional practices for language minority pupils stresses the need to achieve fuller integration between the pupils language development and content-area instruction. Implementation of this approach is just as heavily the responsibility of nonbilingual ESL specialists as it is of the bilingual classroom teachers responsible for subject-matter instruction. Widespread acceptance of this trend has led to altered conceptions of the role of ESL in bilingual education, with greater emphasis on the essential interrelatedness between second language development and content goals in other areas of the curriculum (Milk, 1985). A challenge that remains to be met is how to better prepare both ESL and bilingual teachers for this altered role, given a common tendency to conceptualize ESL and bilingual methodology as essentially independent training activities. This article explores the issue from three separate perspectives: First, a rationale is provided for experientially grounded coursework in the preparation of language teaching professionals, drawing on reports of teacher educators who have attempted innovations in this area; second, research evidence is summarized suggesting the desirability of an integrated language development approach in the education of language minority children; and third, data are presented from an innovative teacher education course that has attempted simultaneously to develop Spanish language proficiency for bilingual teachers while providing a Spanish immersion experience for nonbilingual ESL teachers.

EXPERIENTIALLY GROUNDED TEACHER PREPARATION IN LANGUAGE EDUCATION An important theme in the recent literature on teacher preparation within language education, one that may serve to unify current debates on the preparation of language educators, is the notion that language proficiency is most effectively stimulated when we focus less on language itself and more on its meaningful use in realistic contexts. (For one of the earliest discussions of this 408 TESOL QUARTERLY

insight in the context of the preparation of bilingual educators, see Politzer, 1978.) Enright (1986), in describing the initial learning process of one novice elementary school teacher, emphasizes the profound impact on the teachers development as a language educator when she began to see classroom language as performing both a language teaching as well as a subject-matter teaching function (p. 122). It is this fundamental realization that has motivated bilingual educators to emphasize the tremendous language learning potential for elementary school children of instruction in the content areas. This principle, of course, can be applied not only to elementary pupils, but just as readily to adults needing to strengthen their proficiency in a nonprimary language. The efficacy of a taskbased approach has been reported by a number of bilingual and second language educators who have attempted innovations in the preparation of teachers. Merino and Faltis (1986) developed a course for preservice bilingual teachers based on a task-oriented approach for second language acquisition (p. 46), and found that this experience, in addition to being very motivating, generated a high degree of student-student interaction: The opportunity to use Spanish in problem-solving activities that are intrinsically interesting seems to encourage more students to participate spontaneously in classroom discourse with each other as well as with the instructors. Celce-Murcia (1983) similarly argues that an experiential, problem-solving approach to training language teachers allows prospective teachers to better integrate the content courses with the practical component of the curriculum by making each course an opportunity for dealing with relevant problems (p. 103). In a related vein, Larsen-Freeman (1983) writes that second language teachers need to develop a heightened awareness of choice (p. 264) in order to be able to deal with the tremendous variation in contexts that they are likely to encounter as language teachers. She argues that by preparing people to make choices, we are educating them to be an effective teacher in any situation (p. 264). Each of these four teacher educators reports success in courses that have followed an experiential, problem-solving approach in the preparation of language teachers. The principle is neither new nor is it unique to language education-teacher educators in general have long sought means through which they could ensure greater relevance of methodology courses by providing direct experiences within the course that parallel the kinds of classroom situations that teachers are likely to encounter. (The widespread use of microteaching by teacher education programs for many PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 409

years was motivated by this desire.) Although the positive reports presented above are anecdotal in nature rather than empirically grounded, the basic notion of preparing language teachers through an experiential problem-solving approach has a strong appeal both on logical as well as intuitive grounds. RESEARCH PERSPECTIVES ON LANGUAGE DEVELOPMENT WITHIN BILINGUAL CLASSROOMS In recent years, classroom-based research has provided practical insights into the organization of bilingual instructional settings that can facilitate development in both languages (see, for example, Ramirez, 1985; Ovando & Collier, 1985). Research in three specific areas has particular significance for a contemporary consideration of instructional methods for bilingual classrooms: (a) effective bilingual instructional practices, (b) second language acquisition within classroom settings, and (c) ESL in bilingual education (Milk, 1990). Each of these research areas focuses on a distinctly separate set of issues, and each follows unique approaches to inquiry that reflect different research traditions. The research on effective instructional practices for language minority children draws on a rich tradition of educational research on teaching, and employs a combination of qualitative and quantitative approaches to investigation (see, for example, Tikunoff, 1983; Garcia, 1987). Particularly relevant to dual-language development in classroom settings has been the documentation of the significant role played by two languages (i.e., English and the home language) in mediating learning. Classroom-based research in the area of second language acquisition, by contrast, draws from applied linguistics, and explores the conditions under which acquisition can take place within classrooms (see, for example, Long, 1981; Wong Fillmore, 1982; Johnson, 1983; Long & Porter, 1985; Rigg & Enright, 1986; van Lier, 1988; Chaudron, 1989). One key lesson derived from many of these studies is the critical importance of creating classroom participant structures within which negotiation of meaning in a weaker language can take place. A third area of research, focusing on what constitutes relevant ESL instruction within the context of bilingual education, draws on a tradition of research in the area of second/foreign language teaching that examines which teacher behaviors and what teaching techniques might most effectively contribute to development of the target language (see, for example, Politzer, 1970). Recent work in 410 TESOL QUARTERLY

this area emphasizes the need for ESL instruction for limited English proficient (LEP) children to move beyond effective communication as a primary goal toward a focus on academic competence (Cummins, 1980); this implies a stronger focus on vocabulary enrichment (Saville-Troike, 1984), as well as on critical thinking skills, social skills, and learning strategies (OMalley, Chamot, StewnerManzanares, Russo, & Kpper, 1985). Taken together, developments in these three areas have led to a strong focus on learning outcomes and on cognitive processes for LEP students, as well as on meaningful integration of language and content goals throughout the curriculum. A number of bilingual programs in the U.S. have implemented versions of content-based instruction for ESL (see, for example, Willets, 1986), a practice that has found considerable support from advocates of the sheltered language (Krashen, 1985, p. 70) approach to second language development in bilingual education programs. Another approach (Chamot & OMalley, 1986) following a similar rationale has been termed Cognitive Academic Language Learning Approach (CALLA), in recognition of a need to stress cognitively demanding tasks drawn from academic content once LEP students advance beyond beginning levels of English proficiency. This approach provides for an intermediary bridge between conventional ESL instruction appropriate for the lowest levels of English proficiency, and the regular mainstream instruction received by students who are fully proficient in English. Despite this sharp trend toward more integrated conceptions of bilingual instruction, teacher preparation programs have adapted little to reflect these changing realities. Following conventional modes of dividing up responsibilities for teacher preparation, ESL methods courses typically deal with established methods and techniques for teaching ESL, while bilingual methods courses deal with procedures for teaching content areas in the native language of the student (see, for example, Texas State Board of Education, 1987). It is often not clear within which existing course(s) to include theory and practice related to integrated language teaching. There are, no doubt, institutional factors that may explain this state of affairs. Within universities that provide teacher preparation in bilingual education, it is not unusual for courses related to ESL and courses related to bilingual instruction to be housed in separate departments (and in some instances, different colleges). This separate identity within the university, which often reflects a different discipline base and, in some cases, a different ideological perspective, does not encourage the kind of close collaboration on PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 411

program development that would be needed in order for innovative teacher preparation courses to be initiated. Yet, as more and more research on language minority students points to the importance of an integrated approach to language development and academic achievement, the critical need to reconceptualize the preparation of teachers for LEP populations is evident.

PREPARATION OF ESL AND BILINGUAL LANGUAGE EDUCATORS Bilingual and ESL specialists have distinct roles within instructional programs designed for language minority students. In some programs the bilingual teacher is responsible for all ESL instruction, while in other programs monolingual English teachers work side by side in a complementary fashion with bilingual classroom teachers. (For specific current examples of alternative ways to structure curriculum and instruction in bilingual programs, see Crawford, 1989.) Whenever bilingual and ESL specialists are called on to work together in the same program setting, their roles remain distinct though related. As a consequence, their teacher preparation needs are to some extent distinct. For bilingual teachers, surveys of practitioners reveal that many of them feel a need for stronger development of their Spanish language proficiency (see, for example, Clark & Milk, 1984). For English as a second language teachers, the linguistic demands of their instructional role do not require full proficiency in the non-English language; but there are a number of reasons why ESL teachers working in strictly bilingual settings (i.e., where there are basically two languages in use in the community) might find it useful to have some proficiency in the home language of their pupils. As is the case for ESL teachers who are working in multilingual settings, the experience of learning a new language and of having to deal with academic demands in a second language can provide valuable insights into the world as viewed by an LEP student in an ESL classroom. Despite differing classroom roles, then, there are similarities in the training needs of ESL and bilingual teachers. A common knowledge base related to the understanding of language and to cultural issues underlies effective instructional practices for both ESL and bilingual teachers. Moreover, increased emphasis on the integration of language and content-area instruction has made the roles of bilingual and ESL teachers increasingly interrelated. This newly found interrelatedness, as exemplified by the need for 412 TESOL QUARTERLY

collaboration in designing effective content-based instruction, creates a tremendous potential for common training experiences in the preparation of ESL and bilingual personnel. A SPANISH IMMERSION COURSE FOR TEACHERS OF LEP CHILDREN The experimental teacher training course described here is offered at The University of Texas at San Antonio. The data are from the summers of 1987 and 1988. The course was held for 2 hours daily during the 5-week summer session. For the first offering (1987), 8 graduate students enrolled, and for the second offering (1988), 9 enrolled. Out of the 17 students, 6 were ESL specialists, 10 were bilingual education specialists, and 1 was a bicultural studies major; 10 self-identified as Mexican American, 6 as Anglo, and 1 as African American. With respect to language proficiency, the group was relatively heterogeneous. Students Spanish proficiency, based on formal and informal measures, ranged from advanced beginner (2), to intermediate (7), to advanced (8). The course was specially designed to meet both the differing language proficiency needs of bilingual and ESL teachers, as well as the common language pedagogy needs of teachers who would be implementing content-based instructional strategies for teaching language. Specifically: 1. ESL specialists were provided with an immersion experience in Spanish. 2. Bilingual specialists were provided with opportunities to greatly widen their lexical range in Spanish for the content areas as well as to increase their proficiency in academic Spanish. 3. Both ESL and bilingual specialists were provided with intensive simulated classroom experiences in small-group content-based instruction following a cooperative learning approach. The language goals, therefore, were slightly different for bilingual and ESL participants in the course: For bilingual specialists (most of whom were native speakers of Spanish raised in the U.S. and schooled in English only), the language goal was to increase fluency in academic language and to increase vocabulary range in the content areas of the curriculum; on the other hand, for ESL specialists (most of whom were not native speakers of Spanish, although they may have studied it in high school or college), the goal was a much more modest one of helping them achieve functional proficiency in Spanish at a level that would enable them to communicate at a basic level with children and their parents. PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 413

Perhaps more important than the language goals, however, were the pedagogical goals related to integration of language and content in programs for LEP children. For bilingual teachers, it was expected that insights would be gained with respect to the kinds of ingredients that must be present in order for childrens weaker language to be effectively developed during subject-matter instruction. For ESL teachers, the course was expected to provide an experiential basis for demonstrating the potential effectiveness of content-based language instruction. The research on classroom-based second language acquisition (SLA) cited above suggests that the SLA process is facilitated when students are asked to work within groups that require negotiation of meaning in the second language. For this reason, the course stressed highly interactive small-group learning activities. Given the diverse Spanish language proficiency levels of course participants, it was decided that heterogeneous grouping based on differences in Spanish proficiency (i.e., deliberate mixing of high proficiency with low proficiency students within learning groups) would maximize learning possibilities for the students. Within this heterogeneous grouping, students with lower levels of Spanish would benefit from the input provided by more proficient classmates, and students with higher proficiency levels would benefit from the need to explain cognitively demanding tasks in Spanish to their limited Spanish proficient classmates. In order for heterogeneous groups to function effectively, the course needed to establish a cooperative learning approach that could serve both as a modus operandi for the class as well as a model for the type of instruction that participants should be implementing in their own classrooms (McGroarty, 1989). Hence, a step by step process was introduced from the outset through which cooperative learning procedures (including reinforcement of cooperative behavior and assignment of clearly defined roles within the groups) were carefully developed, following the suggestions of Kagan (1986) and Cohen (1986). In sum, the instructional strategies modeled through the course included: (a) highly interactive learning activities, (b) heterogeneous groups, (c) cognitively demanding tasks, and (d) cooperative learning procedures. In order to directly demonstrate these principles, a bilingual curriculum that satisfies these conditions was selected for use during the initial phase of the course: the Finding Out/Descubrimiento (FO/D) curriculum (DeAvila, Duncan & Navarrete, 1987). During this initial phase, which lasted approximately 15 contact hours, students engaged in small-group sessions based on the FO/D curriculum. Following each initial session, the 414 TESOL QUARTERLY

instructor guided students through an introspective evaluation process that focused on specific aspects of group functioning that appeared to require further attention. During these evaluations, conducted with full class participation, students became accustomed to operating under the fundamental rules established for all subsequent learning activities in the course, including predominant use of Spanish in the class, a focus on learning activities drawn from the content areas, and strict adherence to cooperative learning procedures. During this initial phase of the course, students also completed readings on the theoretical underpinnings of the course, and received minilectures in sheltered Spanish on these principles. In order to prevent inappropriate generalizations regarding immersion education from being drawn, care was taken to stress through required readings and discussion the critical role of LEP childrens native language in achieving academic goals. In particular, students were asked to consider why immersion techniques, which might be appropriate for pursuit of second language goals in certain contexts (such as in this course with adults), could be damaging when pursuing academic goals for language minority children, particularly if the likely outcome is eventual replacement of the childs native language with the dominant language (Hernandez-Chavez, 1984), During the subsequent phase of the course (which lasted approximately 27 contact hours), students were required to plan and design learning activities from the content areas that they would then facilitate in the class, with their classmates role-playing elementary school students. The guidelines for these small-group activities stressed that they were to be highly interactive, cognitively demanding, and require cooperation for full completion of the task. They were also to be planned and implemented totally in Spanish, with the groups composed of participants possessing varying levels of proficiency. Finally, participants were required to keep dialogue journals in order to encourage development of Spanish writing skills. Students were required to make regular entries in their journals, and were encouraged to use the journal as an interactive device to communicate with the instructor. Specific instructions for the journal included reflecting on emotions students were feeling in connection with their language learning experience. In addition to serving as a powerful stimulus for increasing writing fluency, the journals provided a rich source of insight into the kinds of challenges that students face when required to deal with cognitively demanding tasks in their weaker language. PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 415

EVALUATING THE COURSE In order to determine whether the course was accomplishing its goals, during the 1987 and 1988 summer sessions data were collected, designed to answer the following questions: 1. Did participants increase their proficiency in Spanish? 2. Did participants increase their confidence in using Spanish for academic/learning purposes? 3. Did participants gain insight into the difficulties experienced by LEP students when having to deal with the curriculum in their weaker language? 4. Were insights gained with respect to the integration of language and content in bilingual classrooms? 5. Were insights gained with respect to the second language acquisition process, including strategies commonly used by second language learners? The data sources for the study included (a) pre- and postcourse data on language Proficiency, (b) student evaluations of the course, and (c) entries from dialogue journals. Language Proficiency Data Because of the relatively short duration of the course (42 contact hours per summer session), it was not really expected that measurable gains in Spanish language proficiency would be found. Nevertheless, pre- and posttests using four different measures of language proficiency were administered in order to explore possible gains in Spanish proficiency. The first measure was a dictation: in 1987, a 92-word selection taken from a teachers manual for a bilingual reading curriculum; and in 1988, a 108-word selection from a Mexican short story (high school level). In scoring, spelling errors were ignored. The second measure was a relatively simple cloze passage based on a childrens story. This test contained 24 items obtained by deletion of every sixth word, and was scored following the exact-word method. The third measure was a more difficult cloze passage based on a teachers guide for science curriculum. It contained 25 items obtained by deletion of every eighth word, and was corrected using appropriate word scoring. The first three measures were administered by the course instructor. The fourth measure (which was administered during summer 1987 by a research assistant, but not administered during summer 1988 because of lack of funding) was the Spanish version of the Language Assessment Scales (LAS) II. The LAS is an oral proficiency 416 TESOL QUARTERLY

measure that consists of five sections, each requiring a different task: discrimination and production of sounds, vocabulary identification, oral comprehension, and story retelling. The test generates scores that are converted to five proficiency levels, ranging from Level 1 (nonspeaker) to Level 5 (fluent speaker). Students took the pretest for each of the four measures on the first day of class, prior to any instruction; the posttest was taken on the final day of class. The results obtained by the first three measures of language proficiency are reported in Table 1. The overall pattern is one of
TABLE 1 Dictation and Cloze Test Results (Percentage Correct)

Note. Statistical significance tests reported are paired t-tests. The 2 years have been pooled. For each evaluation measure, summary measures and significance tests are based on all students whose gain scores were available. The significance of these multiple t-tests 2 is further supported by the results of a single multivariate procedure, Hotellings T test (p < .01), performed for all students who had gain scores available for all three evaluation measures.

PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES

417

gains, which were substantial in many cases, with only a few losses. For all measures, the average gain scores were statistically significant. Table 2 records gain scores on the LAS II test, which was administered only during the 1987 summer session. Six of the 9 students achieved gains of 10 points or more on this measure; no students posted losses. The average gain score was significant. It is interesting to note that 6 out of the 7 students who began the course with a LAS II score below Level 4 tested at Level 4 or above at the conclusion of the course 3 weeks later. It might be worth speculating on the 2 students who began the course at Level 4 or above (Students 1 and 2), and who posted no gains. There are at least two possible explanations for this: Either Students 1 and 2 failed to gain much proficiency because they were already at a fairly high level, or the test was too easy for them and thus did not adequately measure the gains that they may have made in Spanish. The latter explanation is strengthened by the fact that these two students achieved substantial gains on the two cloze tests (Table 1). It may be that the LAS II is not sensitive to gains by students who are already at a fairly high level of Spanish proficiency. Student Evaluations At the conclusion of the course, the students were asked to complete an evaluation form that focused, in part, on the extent to
TABLE 2 Gain Scores on LAS II Test

418

TESOL QUARTERLY

which they felt the course had helped them in three specific areas: (a) improving their Spanish proficiency, (b) understanding theories related to bilingual instruction, and (c) teaching a second language (see Appendix). The form required respondents to select an answer on a 5-point Likert scale with 3 points described: A Great Deal (scored as 5), A Fair Amount (scored as 3), and Not at All (scored as 1). Efforts were made to encourage frank responses: Students completed the forms anonymously, and forms were collected by a student. The forms were not returned to the instructor until after grades for the course had been turned in. Summary results from the student evaluations are reported in Table 3. High means (defined as above 4) were obtained for some of the items under the broad category of improving Spanish proficiency, but not for others. Specifically, students felt they had made significant improvement in their overall fluency and in academic language (including vocabulary and the ability to communicate concepts from the content areas), but not in grammatical knowledge or in writing mechanics. These results coincide closely with course goals, which focused heavily on oral communication related to teaching the bilingual curriculum, with only secondary stress on writing skills and grammatical knowledge.
TABLE 3 Student Perceptions of Course Effectiveness Means from Student Self-Report Data (1987, 1988 N = 17)

PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES

419

Under the category that relates to understanding theories of bilingual instruction, all four items had means above 4. Each of these items related to a specific area of the syllabus for which readings had been assigned and lectures had been presented: cooperative learning, integration of language and content instruction, second language acquisition theory, and the importance of native language development in bilingual settings. Similarly, high means were obtained for the three items related to second language instruction. These items corresponded to major course goals such as using small groups for second language teaching and creating an awareness of the kinds of problems encountered by LEP students in the classroom. The high means obtained for all the items included under these two categories, therefore, seem to suggest that the course may have led to important theoretical and practical insights related to effective bilingual and ESL instruction. Dialogue Journals Additional insights on effects of the immersion experience on course participants were obtained from the dialogue journals. The student entries are so rich as to merit separate treatment elsewhere, but a few insights briefly summarized here make possible a fuller understanding of the data presented. First, the intensity of emotions surrounding the immersion experience was evident in all the journals. One of the lower proficiency students wrote (uncorrected version):

This student, who was struggling a bit with her Spanish, was clearly concerned that her true personality could not be adequately revealed when she was denied access to English. This first-hand experience and the insight it engendered is extremely significant for teachers of LEP students. A second area explored in the journals related to the continuing need to foster cooperation among group members. One student wrote (uncorrected):

420

TESOL QUARTERLY

[Everyone works well and each person helps the othersbut there exists a certain tendency by some to avoid a particular groupI dont know. For me, I prefer to be with the native speakers.] In a later entry, this student noted that unless there was open encouragement on the part of the instructor and a conscious effort among participants, there was a tendency for communication to take place along ethnic lines. This tendency, which was noted on a number of occasions, underscores the necessity for a clear and deliberate strategy on the part of the instructor to maintain heterogeneity within groups, and to gently nurture full elaboration of each individuals role within the group. A third point raised was the effectiveness of dialogue journals for affective as well as pedagogical reasons. A student wrote (uncorrected):

Finally, the journals provided anecdotal evidence of students perceptions that their fluency in Spanish had improved notably. One Anglo student, whose former husband was a native speaker of Spanish, wrote (uncorrected):

DISCUSSION The experimental Spanish immersion course described here sought to accomplish a number of goals simultaneously: 1. To meet training needs for bilingual teachers through the PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 421

development of academic language proficiency in Spanish, particularly through a broadening of vocabulary range and an increase in fluency in dealing with the content areas. 2. To meet training needs for ESL teachers by helping them gain direct insight into the challenges facing an LEP student in dealing with subject matter in a weaker language, as well as helping ESL teachers working with Spanish-speaking populations to obtain a minimal functional proficiency in the home language of their students. 3. To serve as an experiential model for an alternative approach to ESL instruction that simultaneously meets linguistic and cognitive needs of language minority students. The pre- and posttest data suggest that course participants increased their proficiency in Spanish, but we do not know that these measurable gains were obtained as a result of the intensive immersion experience. Moreover, not all students posted impressive gains. Some participants seemed to improve substantially in their Spanish proficiency while others did not. In part, these results may reflect measurement problems that exist in attempting to demonstrate small-scale gains in language proficiency through integrative instruments such as the cloze (Madsen, 1983). What is clearly evident from the participants self-evaluations, however, is that the overwhelming majority of participants (15 out of 17), felt that their fluency in Spanish, as well as their vocabulary range, had substantially improved as a result of their experience. These self-perceived gains translate into greater confidence in the use of Spanish for instructional purposes by the bilingual teachers; this alone may be as significant as any real gains. The second objective, which relates to a need for ESL teachers to gain both empathy and understanding of the learning process of LEP students when functioning in their weaker language, appears to have been convincingly attained, based on both self-evaluations as well as on subjective comments included in the dialogue journals. In addition to this explicit goal, the dialogue journals reveal that additional insights related to interethnic communication in bilingual settings were apparently obtained by a number of the participants. Based on data from the course evaluations as well as from the dialogue journals, it appears that the third objective (providing an experiential pedagogical model) was firmly obtained by the participants. Without exception, teachers participating in this intensive immersion experience, within which they performed problemsolving tasks following a cooperative learning approach in linguistically heterogeneous groups, felt that they were now able to better 422 TESOL QUARTERLY

achieve the integration of language and content-area instruction in their classrooms. Although the number of students participating in this course is too small for broad generalizations to be drawn, the possibilities posed by this type of experientially grounded methods course appear to be promising indeed. One factor that needs to be explored is the extent to which the positive effects reported here are dependent upon a sustained, semi-intensive format such as existed in this summer school course, meeting daily for 2 hours or more. The potential for applications of this type of immersion experience in the in-service development of bilingual/ESL personnel merits further exploration. New means must be found through which classroom teachers may experience, in a direct and dramatic manner, the tremendous benefits of an instructional approach that thoroughly integrates language and content-area instruction. ACKNOWLEDGMENTS
Data from the first year of this study were presented at the National Association for Bilingual Education 17th Annual International Bilingual/Bicultural Education Conference in Houston, April 1988. An updated version of the paper was presented at the 2&d Annual TESOL Conference in San Antonio, March 1989. Funding from the Dean of the College of Social and Behavioral Sciences at the University of Texas at San Antonio is gratefully acknowledged. Special thanks go to Patricia Rosales and to Mara Espericueta for their assistance in completing the study, as well as to anonymous TESOL Quarterly reviewers for their helpful suggestions. Summary measures and significance tests in Tables 1 and 2 were computed by a TESOL Quarterly reviewer.

THE AUTHOR
Robert D. Milk, Professor and Director of the Division of Bicultural-Bilingual Studies at the University of Texas at San Antonio, has taught ESL and Spanish at the secondary level and worked as a teacher educator in Peru at the elementary school level. He currently directs two Title VII training projects for bilingual education teachers.

REFERENCES
Celce-Murcia, M. (1983). Problem solving: A bridge builder between theory and practice. In J. E. Alatis, H. H. Stern, & P. Strevens (Eds.), Georgetown University Round Table on Languages and Linguistics 1983: Applied linguistics and the preparation of second language teachers (pp. 97-105). Washington, DC: Georgetown University Press.

PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 423

Chamot, A. U., & OMalley, J. M. (1986). A cognitive academic language learning approach: An ESL content-based curriculum. Rosslyn, VA: National Clearinghouse for Bilingual Education, Chaudron, C. (1989). Second language classrooms: Research on teaching and learning. New York: Cambridge University Press. Clark, E. R., & Milk, R. D. (1984). Training bilingual teachers: A look at the Title VII graduate in the field. NABE Journal, 8 (l), 41-53. Cohen, E. (1986). Designing groupwork: Strategies for the heterogeneous classroom. New York: Teachers College Press. Crawford, J. (1989). Bilingual education: History, politics, theory, and practice. Trenton, NJ: Crane. Cummins, J. (1980). The entry and exit fallacy in bilingual education. NABE Journal, 4 (3), 25-60. DeAvila, E., Duncan, S., & Navarrete, C. (1987). Finding Out/Descubrimiento teachers resource guide. Northvale, NJ: Santillana. Enright, D. S. (1986). Use everything you have to teach English: Providing useful input to young language learners. In P. Rigg & D. S. Enright (Eds.), Children and ESL: Integrating perspectives (pp. 113-162). Washington, DC: TESOL. Garcia, E. (1987). Effective schooling for language minority students. Focus, 1, 1-12. Wheaton, MD: National Clearinghouse for Bilingual Education. Hernandez-Chavez, E. (1984). The inadequacy of English immersion education as an educational approach for language minority students in the United States. In Studies on Immersion Education (pp. 144-183). Sacramento: California State Department of Education. Johnson, D. (1983). Natural language learning by design: A classroom experiment in social interaction and second language acquisition. TESOL Quarterly, 17 (1), 55-68. Kagan, S. (1986). Cooperative learning and sociocultural factors in schooling. In Beyond language: Social and cultural factors in schooling language minority students (pp. 231-298). Sacramento: California State Department of Education. Krashen, S. (1985). The input hypothesis: Issues and implications. N e w York: Longman. Larsen-Freeman, D. (1983). Training teachers or educating a teacher. In J. E. Alatis, H. H. Stern, & P. Strevens (Eds.), Georgetown University Round Table on Languages and Linguistics 1983: Applied linguistics and the preparation of second language teachers (pp. 264-274). Washington, DC: Georgetown University Press. Long, M. (1981). Input, interaction, and second language acquisition. In H. Winitz (Ed.), Native language and foreign language acquisition. Annals of the New York Academy of Sciences, 379, 250-278. Long, M., & Porter, P. ( 1985). Group work, interlanguage talk, and second language acquisition. TESOL Quarterly, 19 (2), 207-228. Madsen, H. S. (1983). Techniques in testing. New York: Oxford University Press. 424 TESOL QUARTERLY

McGroarty, M. (1989). The benefits of cooperative learning arrangements in second language instruction. NABE Journal, 13 (2), 127-144. Merino, B., & Faltis, C. (1986). Spanish for special purposes: Communication strategies for teachers in bilingual education. Foreign Language Annals, 19 (1), 43-46. Milk, R. D. (1985). The changing role of ESL in bilingual education. TESOL Quarterly, 19 (4), 657-672. Milk, R. D. (1990). Integrating language and content: Implications for language distribution in bilingual classrooms. In R. Jacobsen & C. Faltis (Eds.), Language distribution issues in bilingual schooling (pp. 32-44). Clevedon, Avon, England: Multilingual Matters. OMalley, J. M., Chamot, A. U., Stewner-Manzanares, G., Russo, R. P., & Kpper, L. (1985). Learning strategies used by beginning and intermediate ESL students. Language Learning, 35, 21-46. Ovando, C. J., & Collier, V. P. (1985). Bilingual and ESL classrooms: Teaching in multicultural contexts. New York: McGraw-Hill. Politzer, R. (1970). Some reflections on good and bad language teaching behaviors. Language Learning, 20, 31-43. Politzer, R. (1978). Some reflections on the role of linguistics in the preparation of bilingual/cross-cultural teachers. Bilingual Education Paper Series, 1 (12). (ERIC Document Reproduction Service No. ED 161 274) Ramirez, A. G. (1985). Bilingualism through schooling: Cross-cultural education for minority and majority students. Albany, NY: State University of New York Press. Rigg, P., & Enright, D. S. (Eds.). (1986). Children and ESL: Integrating perspectives. Washington, DC: TESOL. Saville-Troike, M. (1984). What really matters in second language learning for academic achievement? TESOL Quarterly, 18 (2), 199-220. Texas State Board of Education. (1987). Standards for teacher education institutions. Texas administrative code and statutory citations (Chapter 137, Subchapter M, pp. 35-40). Austin: Texas Education Agency. Tikunoff, W. (1983). Significant bilingual instructional features study. San Francisco: Far West Laboratory for Educational Research and Development. van Lier, L. (1988). The classroom and the language learner. New York: Longman. Willets, K. (Ed.). (1986). Integrating language and content instruction. Los Angeles: University of California, Center for Language Education and Research. Wong Fillmore, L. (1982). Instructional language as linguistic input: Second-language learning in classrooms. In L. C. Wilkinson (Ed.), Communicating in the classroom (pp. 283-296). New York: Academic Press.

PREPARING ESL AND BILINGUAL TEACHERS FOR CHANGING ROLES 425

APPENDIX Course Evaluation I. Rate your level of proficiency in Spanish (circle one) II. To what extent do you feel that the course helped you in each of the following areas (circle one): A. Language proficiency (Spanish) 1. Overall fluency 2. Ability to communicate ideas and concepts related to content areas 3. Vocabulary 4. Awareness of mechanics for writing 5. Understanding of Spanish grammar 6. Awareness of local & regional Spanish variety B. Understanding theories 7. Second language acquisition in the classroom 8. Cooperative learning 9. Integrating language and content 10. Dual language development in bilingual settings C. Teaching a second language 11. Effective ideas/techniques 12. Use of small groups 13. Awareness of learners problems in dealing with content in a weaker language Low Intermediate High A = A Great Deal C = A Fair Amount E = Not at All

426

TESOL QUARTERLY

TESOL QUARTERLY, Vol. 24, No. 3, Autumn 1990

The TOEFL Test of Written English: Causes for Concern

ANN RAIMES
Hunter College, City University of New York

Owned and administered by the Educational Testing Service (ETS), the Test of English as a Foreign Language (TOEFL), taken by approximately 600,000 students a year, influences access to or exclusion from colleges and universities in North America. This article provides background information about the TOEFL and ETS; describes the development of ETS tests of composition for native speakers of English and of the most recent addition to the TOEFL testing program, the Test of Written English (TWE); and explores seven areas of concern with respect to the TWE: the comparability of topic types; the lack of topic choice; the lack of distinction between graduate and undergraduate students; the scoring system; the question of what the test measures; the question of whether both the TOEFL and the TWE are needed; and the backwash effect of the TWE, including the proliferation of coaching and test-specific instructional materials. The article urges careful scrutiny of new developments in ETS testing as they affect our students, and ends with seven recommendations for action.

On six Fridays and six Saturdays a year, thousands of people in places as far-ranging as Dallas, Texas; Gwynedd Valley, Pennsylvania; Whitehorse, Yukon Territory Victoria, British Columbia; Quito, Ecuador; Sydney, Australia; Alexandria, Egypt; Athens, Greece; Lome, Togo; Tel Aviv, Israel; and Kyoto, Japan assemble at test centers to take the TOEFLthe Test of English as a Foreign Language. At more than 500 test centers in the U.S. and Canada alone and 700 centers in 170 other countries, approximately 600,000 students a year take the TOEFL, and as many as 80,000 may test worldwide on one date (Educational Testing Service, 1989, p. 1). THE TOEFL AND THE EDUCATIONAL TESTING SERVICE The TOEFL testing program is owned and run by an organization that exercises considerable power over education and 427

professional life in the United States: the Educational Testing Service (ETS). ETS has an enormous impact on U.S. students lives with its Scholastic Aptitude Test (SAT), Graduate Record Exam (GRE), Law School Admission Test (LSAT), and numerous other standardized tests that influence access to higher education and to professions. TOEFLS influence also extends beyond academia, since many government agencies, school programs, and licensing/ certification agencies use TOEFL scores to evaluate English proficiency (Educational Testing Service, 1990a, p. 3). In some states of the U. S., you cant get to be a real estate salesperson, a firefighter, a plumbing engineer, a golf pro, or even a barber without taking an ETS test (Owen, 1985). The TOEFL expands not only the influence of ETS but also its revenues. Even in 1977, the TOEFL was ETSs seventh highest source of revenue, contributing a profit on gross revenues of more than 10% (Nairn & Associates, 1980). And in 1979, ETS, classified as a tax-exempt, nonprofit institution, grossed $94 million in annual revenues ( Nairn et al., 1980). Since that time the TOEFL has grown steadily, showing an average increase of about 41,000 students a year. Although the scores are used by more than 2,300 colleges in North America, the fees for the TOEFL are paid by the individual students. Currently fees range from $31 to $41. On the basis of simple arithmetic, one may infer that the 566,000 students who registered for the TOEFL in 1988-89 (Educational Testing Service, 1990a) contributed a great deal to ETS revenues. An organization with such influential, widespread, and lucrative testing programs is bound to generate criticism. A Ralph Nader organization has produced a highly critical report, The Reign of ETS: The Corporation that Makes up Minds (Nairn et al., 1980). This generated a series of critical examinations of ETS and particularly of the SAT. A no-holds-barred attack on ETSS domination of the testing field appeared in 1985 (Owen), closely followed by several academic research studies. Among these, Crouse and Trusheim (1988) have presented the results of a 6-year research study that shows in particular the adverse effect the SAT has on black and low-income applicants. Rosser (1989) has reported that the SAT is biased against women, underpredicting their grades. ETS counters many of the criticisms by referring to its own research findings. Two Harvard professors (Slack & Porter, 1980), however, in an attempt to dispute the validity of the SAT, have argued that ETSs rebuttals of criticism cite research studies in a way that is so highly selective as to be biased. They also provide evidence that ETS misrepresents data on validity coefficients for the SAT with high school records, concluding that the data that can be tracked down 428 TESOL QUARTERLY

and analyzed are often at variance with the numerous summary statements to be found in ETSs technical documents and promotional literature (p. 169). To deflect criticism, ETS has tried to involve professional experts to plan programs and generate policy. It has been the practice of ETS to form and confer with advisory groups, such as the TOEFL Policy Council. However, the members of the council and of its committees are appointed by ETS-governed boards or elected by the appointed members. If members are unhappy with ETS, they have little recourse. The tests belong to ETS. So do the data. The control over what data are released, what research is carried out and reported, and ultimately what is tested and how, remains the province of the ETS staff. ETS does appoint a TOEFL Research Committee composed of prominent experts in our field. But according to the description of current TOEFL research procedures presented at ETSs Second TOEFL Invitational Conference (October 1984), research studies are proposed and conducted by ETS staff members, not initiated by the committee (Holtzclaw, 1986). ETS unilaterally and unequivocally controls the form and content of the tests and the data they generate. Since ETS owns, designs, administers, and evaluates the TOEFL, it is in a position to exercise a great deal of influence over ESL/EFL students academic careers. That influence is spreading. Even as more and more students take the TOEFL each year, more tests, are being added. In the early 1980s, a separate Test of Spoken English (TSE) was added (with an additional fee for students ranging from $75 to $100). Then, in 1986, the Test of Written English (TWE) was added, administered at four of the twelve TOEFL administrations. As of now, students are not chargedan additional fee. The purpose of the TWE, as an ETS booklet rather euphemistically puts it, is not so much to test as to provide students with an opportunity to demonstrate their ability to write in English (Educational Testing Service, 1990b, p. 3). This demonstrating of ability, of course, leads to what higher education institutions in North America have been doing with TOEFL scores for many years: using them to admit or exclude students. TOEFL tests are also used to structure curricula of preparation courses, instruction, and materials. Such a widespread testing program has a considerable impact on teachers and on our constituency of ESL/EFL students. We need to examine itand continually reexamine itwith the utmost care. Bearing in mind that testing is power and that the teaching and testing of English are, in a large sense, political acts (White, 1986, p. 77), we need to look closely at this new writing test to examine its purpose, its structure, and its effects. THE TOEFL TEST OF WRITTEN ENGLISH 429

ETS TESTS OF COMPOSITION FOR NATIVE SPEAKERS OF ENGLISH First, however, it is instructive to examine ETS practice with respect to combinations of essay and standardized tests of composition for native speakers of English. The College Board (which later joined forces with ETS for its Admissions Testing Program) began testing composition with essay tests in 1916 (Godshalk, Swineford, & Coffmann, 1966). It was in 1947 that the College Board first experimented with testing writing objectively (without an essay). An objective section was included in the English Composition Test (ECT), one of the subject area Achievement Tests. When the Board compared the essay with the objective section and with course grades and teachers ratings, the findings were that a 60-minute essay would have markedly less predictive value [of teachers ratings of ability to write expository prose, and course grades in English] than a full length test composed entirely of objective material (cited in Davis & Davis, 1953, p. 178). Objective approaches then replaced the essay in the ECT. However, teachers protested and in 1953 a 2-hour essay exam, the General Composition test (GCT), was instituted for 3 years on an experimental basis. During this period, the College Board and ETS studied the results of three of their tests: the verbal portion of the multiple-choice SAT, the multiple-choice ECT, and the new GCT (reported in Hoffmann, 1962, and in Godshalk, Swineford, & Coffmann, 1966). The study compared the results of the three tests with a standard by which teachers of English composition at selected schools [rated] their students on the basis of many essays each student had written in school (Hoffmann, 1962, p. 116). The all-essay testthe GCTcorrelated least well. First and second were the verbal SAT and the ECT respectively. Multiple choice had won the day as an indicator of real-world writing ability. ETS eliminated the GCT. Yet, even though the study had shown that the SAT was a better indicator of essay writing ability than the ECT, the latter continued to be offered, and students continued to take and pay for both the SAT and the optional ECT as part of their college application process. Since teachers and administrators continued to insist on the value of asking students to write, ETS experimented with writing samples, editing tasks, and more essay tests until in 1977 they released an alternative version of the English Composition Test, with one of the multiple-choice sections replaced by a 20-minute essay holistically rated by two readers. Native speakers of English who now apply to so-called selective colleges and universities can 430 TESOL QUARTERLY

elect either the English Composition Test without essay or the ECT with essay. Scores on these tests are reported as a single numerical score on the familiar 200 to 800 scale. An ETS publication (College Entrance Examination Board, 1989) reports that the correlation coefficient between the multiple-choice and essay components of the English Composition Test is .47 (p. 21), lower even than the .66 correlation coefficient reported between verbal and mathematical scores of the SAT. Owen, who takes ETS to task in his book None of the Above: Behind the Myth of Scholastic Aptitude (1985), provocatively asks a series of questions about these two alternative composition tests and the research findings. The first question to ask, he suggests, is whether the ECT with essay is a "better measure of writing skills than the ECT without ." An answer to that question would lead to more questions: If one test is better, why give the other? If the tests are the same, why give both? In an exercise that is supposed to rank people scientifically on a scale from 200 to 800, how can you offer a choice of tests? (1985, p. 26). He contends that throwing out the multiple-choice version would imply what ETS has always denied: that an essay test, especially a teeny one, is better than a multiple-choice test. Alternatively, getting rid of the essay would appear to be confessing that holistic grading is ineffective or in Owens more colorful terms, a bunch of hooey (1985, p. 26). Meanwhile, students have to decide which test to take. THE TOEFL TEST OF WRITTEN ENGLISH There are clear parallels between the two versions of the ECT and the TOEFL/TWE. Long before the introduction of the TWE, the TOEFL used to have five multiple-choice sections: Listening Comprehension, English Structure, Vocabulary, Reading Comprehension, and Writing Ability. An ETS study in 1967 by Pitcher and Ra (cited in Hale, Stansfield, & Duran, 1984) looked at correlations between writing sample scores on four separate themes and the five sections of the TOEFL. The highest correlations were .74 for both the English Structure and the Writing Ability sections. A 1979 ETS study of the TOEFL by Pike (cited in Hale, Stansfield, & Duran, 1984) examined correlations between sections of the test with an essay test, a cloze test, and a rewriting test. This time stronger correlations (averaging .85) were found between the essay and the Writing Ability section of the test. Pikes recommendation after this research was that the TOEFL Writing Ability section should not be replaced with a writing sample but that the Writing Ability section should be combined with the English Structure section. Once again the research recommended the multiple-choice format for assessing writing ability. THE TOEFL TEST OF WRITTEN ENGLISH 431

According to ETS, it was pressure from faculty to institute an essay test that led to the development of the TWE. ETS officials writing in the TESOL Newsletter (Stansfield & Webster, 1986) described the preparation for the new test, mentioning a survey of academic writing in 190 departments conducted for the TOEFL program by Bridgeman and Carlson (p. 17). In fact, Bridgeman and Carlson (1983) did not collect actual academic writing samples from those 190 departments. They began by interviewing 30 faculty members from six universities in engineering, business, and English language institutes. Then after in-depth discussions with advisory committee members, research and program staff members at ETS and faculty and administrators at local institutions (p. 13), they developed a questionnaire, which included 10 possible topic types. Faculty members were to indicate the degree of acceptability (Appendix A, p. 7) for each topic and to indicate the topic types they considered the most appropriate. In other words, the choice of topic types came before the questionnaire was sent out, not after responses about actual writing assignments were received. The questionnaires were then sent to faculty in the following 7 disciplines: undergraduate English and 6 graduate departments electrical engineering, civil engineering, computer science, chemistry, psychology, and MBA programs. The same questionnaires were sent to graduate and undergraduate faculty. The respondents were asked to rate not just the topic types but specific examples of topic types. These, too, were developed by ETS staff. One hundred ninety questionnaires were returned. Again, it is important to note that these 190 departments were merely responding to categories established by ETS researchers, not contributing to the development of the categories. Despite these drawbacks, the researchers stated broad conclusions with confidence: Although some important common elements among the different departments were reported, the survey data distinctly indicate that different disciplines do not uniformly agree on the writing task demands and on a single preferred mode of discourse for evaluating entering undergraduate and graduate students (Bridgeman & Carlson, 1983, p. 56-57). Notwithstanding findings that questioned the viability of a generalized writing test, ETS went ahead with the TWE. Four features were adopted after the preliminary research: 1. Two topic types were selected. In one the student compares/ contrasts two opposing points of view and defends a position in favor of one; in the other, the student describes and interprets a chart or graph (Stansfield & Webster, 1986, p. 17). 432 TESOL QUARTERLY

2. Only one topic type appears on each test for a 30-minute response. 3. Scoring of the TWE is on a holistic 6-point scale by two readers. 4. Scores are reported to individual colleges. These features, and more recent adjustments to the test format, present at least seven major causes for concern. CAUSES FOR CONCERN First, we should be concerned about the comparability of the topic types selected, scrutinizing comparability both between topic types used for native versus nonnative speakers and between topic types used on different administrations of the TWE. Native speakers of English applying to undergraduate programs in colleges and universities in North America only write an essay on an ETS test if they select the English Composition Test with essay. For this, they write on topics such as People seldom stand up for what they truly believe; instead they merely go along with the popular view. Do you agree or disagree with this statement? (Essay topic for December 4, 1982, cited in Shostak, 1987, p. 11). This relatively straightforward position paper with one conceptual demand (to take a position on an issue) certainly seems less demanding than the two- or three-part task of the TWE: compare, contrast, and take a position; or describe and interpret a chart or graph. After Bridgeman and Carlsons extensive research (1983), two topic types were selected for the TWE. The official test guide asserts that further research by Carlson, Bridgeman, Camp, and Waanders (1985) reported that correlations among writing topics were as high across topic types as within topic types, suggesting that overall competency in composition could be adequately assessed using a variety of composition types (Educational Testing Service, 1989, p. 2). The researchers do indeed support the first statement about correlations, but their text does not offer the suggestion that ETS implies: Correlations were as high across topic types as within topic types. This result suggests that (1) the different topics did not elicit qualitatively different writing performance, and/or (2) the readers maintained a comparable scale for evaluating the writing samples, despite fluctuations from topic to topic, These positive results, however, should not be interpreted as evidence that papers written in response to any topic or type of topic would yield equivalent reliability [italics added]. The topics were selected on the basis of previous research indicating THE TOEFL TEST OF WRITTEN ENGLISH 433

that specific kinds of topics would serve as more appropriate stimuli to reflect the academic writing task demands experienced by students in higher education in the United States. Carefully controlled conditions of design and pretesting, and of scoring methods that emphasized functional academic English proficiency, would need to be replicated to attain similar results. (Carlson, Bridgeman, Camp, & Waanders, 1985, p. 76-77) The researchers caution is necessary. According to a senior ETS official, ETS and its Core Reader Group for the TWE have decided that the chart/graph topic raises so many concerns that it should not be used until more research is carried out (Henning, 1990). And, according to the Director of the TWE, it has been used only once in the last eight test administrations, when it was included only for research purposes as part of a comparability study (C. Taylor, personal communication, December 1990). According to Taylor, so much language is contained in the charts and graphs themselves that there is a concern that students simply reproduce that given language. However, although the chart/graph topic has been shelved, it is still listed as a TWE topic type in the official guide (Educational Testing Service, 1989). Interestinglyand confusingly for both students and teachersit appears there not with the originally announced compare, contrast, and defend topic, but with two different topic types: express and support an opinion and choose and defend a point of view (Educational Testing Service, 1989, p. 4). And indeed, one of the sample essay questions in the guide illustrates how far the TWE has moved not only from the chart/graph topic but also from the original compare, contrast, and defend topic that the research recommended: Inventions such as eyeglasses and the sewing machine have had an important effect on our lives. Choose another invention that you think is important. Give specific reasons for your choice. (p. 57) There was a good deal of literature to explain the decisions about topics in the new TWE, but very little to explain the quite radical changes that are occurring in the test. The speed of substitution of new topics for old does not appear to be the result of the carefully conducted research appropriate for a large-scale testing program that affects students academic careers. One might question the claim made by the Assistant Director of the TWE Division at ETS that since the introduction of the TWE test, carefully controlled item development, pretesting, pretest analysis, test administration, and postadministration scoring procedures have helped to maintain the comparability of TWE topics within and across administrations 434 TESOL QUARTERLY

(Fallon, 1989, p. 7). With swift changes in topic types, comparability must be difficult to maintain. Meanwhile, students take the test and have scores reported. Second, we should be concerned that students are presented with only one topic type, though different types of topics are offered at different test administrations. Greenberg (1986) points out that this administrative decision seems to fly in the face of the TOEFL researchers hypothesis that writing competence is a situational construct and, as such, should be elicited by tasks with different academic demands (p. 537). In fact, the ETS research conducted by Bridgeman and Carlson used a multidimensional scaling analysis to show that the two essay types initially selected for TWE seem to call for very different cognitive and linguistic skills (Greenberg, 1986, p. 537). Faculty perceived this without the benefit of research findings; Bridgeman and Carlson (1983) report that the chart/graph topic was seen as inappropriate by a majority of the English faculty (p. 56). In addition, the 1985 TOEFL research study by Carlson et al. noted that one writing sample is not necessarily a sufficient sample of writing performance (p. 81) and recommended assessment by means of more than one writing sample. Despite that, only one topic appears at any TWE test administration and is expected to have predictive validity for graduate and undergraduate students in any discipline. Third, we should be concerned about the curious blending of undergraduate and graduate levels that the TWE promotes. Of the 7 disciplines surveyed, only 1 (English) was at the undergraduate level. The faculty there recommended the compare and contrast topic, while the graduate disciplines (engineering, business, and the sciences) recommended the chart or graph-neither of which seem to be favored in current test administrations. While ETS distinguishes between the undergraduate SAT and the graduate GRE for native speakers, nonnative-speaking students applying for undergraduate and graduate schools all receive the same TOEFL and TWE. Yet they will face very different demands. Fourth, we should be concerned about several aspects of the scoring system. Colleges and universities to which nonnativespeaking students apply receive not the TWE writing sample itself but only a score on a 6-point scale. The score can thus only be useful for admission (or rejection); the writing sample is not available for an institutions specific diagnostic purposes. For students rattled enough by the ETS numbers game, the numerical score of the TWE can only add confusion and stress. Already they are baffled by the scoring system of the TOEFL, in which they receive a scaled score from 200 to 677, with a last digit THE TOEFL TEST OF WRITTEN ENGLISH 435

of 0, 3, or 7. The 1990-91 Bulletin of Information for TOEFL and Test of Spoken English informs students that scaled scores are reported; then, unrealistically, refers these students to journal literature on item response theory, and tells them to add the scaled scores for each section and multiply by ten thirds (Educational Testing Service, 1990a)! If they dont trust the computer scoring and the complicated scaling of the TOEFL, they can request restoring by handfor an additional $20. The TWE, holistically scored by two readers on a 6-point scale, allows for no restoring. Students know that if TOEFL and TWE scores begin to be formally linked for admissions purposes, students can be denied acceptance at a college or university on the strength of one point on a holistic scale. With such a numbers game, its no wonder that students become desperate and test forms are leaked. A recent New York Times article reported that 620 students who took the TOEFL in five centers around the southern Indian city of Bangalore on October 28 had their test results annulled because copies of the test had been widely sold beforehand (Crossette, 1990, p. A10). And when Qinghua University, wanting to limit the number of students entering U.S. schools, did not announce the place of registration for the TOEFL test, students lined up in random places until the final announcement caused a riot and one student jumped to his death. He was carrying a notebook with the words TOEFL, TOEFL, TOEFL, CONFIDENCE, CONFIDENCE, CONFIDENCE etched on the cover (Elliott, 1990, p. 23). Fifth, we should be concerned about establishing clearly what the TWE is or is not testing. ETS personnel frequently assert that TOEFL is not an aptitude testand in fact it does not correlate highly with students grade point averages (Light, Xu, & Mossop, 1987). ETS testing directors assert that the purpose of the TOEFL is to help institutions to determine whether a nonnative Englishspeaking applicant for admission has attained sufficient proficiency to study in an English-medium instructional environment (Stansfield & Webster, 1986, p. 17). But if the test is measuring attained proficiency, why does the writing test present tasks that are representative not of the writing the students have done in the pastthe types of tasks the IEA (International Association for the Evaluation of Educational Achievement) Study of Composition has identified (Purves, 1988) but of the type that they will be doing in the future? Why should international students be expected to have mastered college-level writing as a requirement for admission? Dont the faculty have a responsibility to teach field-specific forms of academic writing, rather than expect mastery from their entering students, particularly students from other countries? 436 TESOL QUARTERLY

Sixth, we should be concerned about whether we need both the TOEFL and the TWE. TOEFL researchers reporting on a correlation study have found that the correlation of scores on the writing sample with the TOEFL total scores and with GRE verbal scores is nearly identical, indicating that the writing sample scores serve as an indicator of English language skills (Carlson et al., 1985, p. 79). They justify the use of both TOEFL and TWE by saying that correlations between the holistic essay score and the TOEFL score indicate that the two measures evaluate English proficiency to a considerable degree, but that the overlap between the two instruments is not perfect. The writing sample contributes additional [italics added] information regarding English proficiency (Carlson et al., 1985, p. 79). For the recipient, that additional information is buried in a holistic, 6-point score. ETS researchers provide justification for the TWE, saying, TOEFL provides evidence of mastery of English language. skills, but not of higher order writing skills such as organization and quality of ideas (Carlson et al., 1985, p. 79). But since ETS rejected scoring on the three separate features of content organization, and grammar despite faculty members recommendations that this would be preferable to an overall score (Bridgeman & Carlson, 1983), we have no way of knowing what kind of organization or what quality of ideas a score reflects. Even preparation books for the TWE raise questions about the need for both the TOEFL and the TWE: The TOEFL Programs research has suggested that if you score around 550 on the TOEFL, you are most likely to score about a 4 on the TWE. If you score around 500 on the TOEFL, you are most likely to score about a 3 on the TWE (Hamp-Lyons, 1989, p. 2). Recalling the questions that Owen asked about the two versions of the ECT, one needs to ask similar questions about TOEFL and TWE: If one testis better, why give the other? If the tests are the same, why give both? Weve seen from ETSS handling of the ECT that the tendency is not to replace but to add on. So TOEFL has been joined by the TSE and the TWE. Finally, we should be concerned about the special instruction and the resulting inequalities that the tests generate, as well as the effect on instruction in general. Students from high-income families tend to receive high SAT scores (Admissions Testing Program of the College Entrance Examination Board, 1979), and coaching has been shown to be concentrated in the higher income groups (Federal Trade Commission, 1979), inevitably linking it to class and race. While ETS staff members have consistently denied that coaching can influence the results of standardized tests like the SAT or GRE THE TOEFL TEST OF WRITTEN ENGLISH 437

(Pike & Evans, 1972; Jackson, 1980), other data show otherwise (Pallone, 1961; Marron, 1965; Slack & Porter, 1980). Although not advised by ETS, coaching flourishes for the TOEFL as well as for the SAT. According to a Research Associate at Stanley Kaplan Educational Centers, for example, TOEFL preparation courses give students 100 hours of audiotape practice of TOEFL-type questions (L. Earl, personal communication, April, 1989). In these courses and in any curriculum leading up to a standardized test, the test itself creates a negative backwash effect. What the multiple-choice format of the TOEFL actually leads to in instruction is far removed from the communicative classroom advocated by most TESOL professionals (Savignon, 1986). TOEFL self-study test materials are as reductive as special courses. ETS provides its own TOEFL study materials at a cost of $39. In addition, many commercial publishers, seeing the spread of the test, have jumped on the bandwagon. It is not surprising that with the new Test of Written English, new TWE preparation books are also appearing, as indicated by a book review in this issue of the TESOL Quarterly. Teachers who have seen writing as a tool of learning, as a means of expressing ideas and creating language for a variety of purposes and audiences, will not welcome textbooks relegating it to repetitive practice of two ETS-generated topics. Also alarming is the fact that the topic types of the TWE seem to be changing so rapidly that as preparation books appear, they might be out of date. Students and teachers need to know how the test format is changing before they even consider buying and using specific test-preparation textbooks. Most important, we need to heed Roys words (1987) that we not allow consciousness of the TWE to foster a narrowly instrumental mode of teaching and a dependent mode of learning (p. 5). These seven causes for concern alert us to the fact that the new Test of Written English, as well as the literature surrounding it and developed from it, should be scrutinized with great care. To preserve symmetry, seven recommendations for action by ESL/ EFL teachers are presented. RECOMMENDATIONS 1. Individually, by telephone, letter, publication, and in public forums, we should continue to ask questions, more questions, and more probing questions about the TWE.1 We need to ask
1 One

forum for discussion of the TOEFL is the Fair Test Examiner, the newsletter of Fair Test (The National Center for Fair and Open Testing, 342 Broadway, Cambridge, MA 02139-1802), which works to reduce the overuse and misuse of standardized tests.

438

TESOL QUARTERLY

about test question development and content-related validation evidence. We also need to ask about construct-related validation evidence and about reliability. At present, ETS research uses only another of its tests, the TOEFL, to establish construct validity and reliability (Educational Testing Service, 1989). We need to know ETSs plans for the test: So far they have announced no plans to make the TWE a separate test with a separate fee, like the TSE, but will this policy continue once colleges have prescribed score cutoffs for admission? Consider how many students would be excluded from the possibility of higher education with the institution of a fee increase. 2. Those of us who teach ESL in postsecondary schools in North America need to be informed about the admissions policies of our own schools. How do they use TOEFL and TWE scores? ESL teachers cant let subject area faculty, college administrators, and ETS personnel make all the crucial decisions about which international students are admitted to our institutions and what skills they need. We are the experts. We need to reassert that position. 3. We should press for more information about the training of raters for the TWE. What factors influence assessment? Vopat (1982) provides a chilling description of formulaic essay writing and assembly-line grading (p. 45) for ETSs Advanced Placement essays. What kind of writing are raters trained to reward? Are they encouraged to recognize a wider variety of rhetorical modes than the ethnocentric deductive linear argument, which Land and Whitley (1989) view as situated within a particular sociopolitical context (p. 289)? 4. We should see to it that ESL/EFL students are not treated as one monolithic group. Tests should reflect the variety of students educational levels and purposes, distinguishing between graduate and undergraduate students. 5. We should avoid recommending and using reductive methods of instruction and materials for test preparation. Our students should be spending time learning English, not learning ETScoping skills. 6. We should examine the TOEFL and TWE in relation to other proficiency tests such as the oral ACTFL (American Council on the Teaching of Foreign Languages) proficiency tests. We should examine the TWE also in relation to research such as the IEA study of the types of writing students in secondary schools actually do (Purves, 1988). Perhaps this data base is the one that THE TOEFL TEST OF WRITTEN ENGLISH 439

should determine the kind of writing that undergraduate applicants are asked to produce. 7. Finally, we need to set up mechanisms to keep close watch on ETS tests that affect our students. Professional educational organizations (e.g., TESOL) might set up their own independent, active, and vigilant testing committees to review test development, procedures, scoring, and fees, to carry out research, and to disseminate results and recommendations. We should know the effects of tests on our student population in terms of economics, quality of instruction in English, and requirements for further study. When students inclusion or exclusion from higher education is at stake, the stakes are indeed high.

ACKNOWLEDGMENTS
An earlier version of this paper was presented at the 24th Annual TESOL Convention in San Francisco, March 1990. The author is grateful to colleagues Karen Greenberg and Kate Parry, and three anonymous TESOL Quarterly reviewers for their helpful comments. The views expressed here remain those of the author. Sandra Silberstein, TESOL Quarterly editor, posed searching, thoughtful questions and offered wise counsel.

THE AUTHOR
Ann Raimes is the author of many articles on writing research, theory, and teaching, and on ESL methodology. Her books include Techniques in Teaching Writing (Oxford University Press, 1983), Exploring Through Writing (St. Martins Press, 1987), and How English Works: A Grammar Handbook With Readings (St. Martins Press, 1990). She is Chair of the TESOL Publications Committee.

REFERENCES
Admissions Testing Program of the College Entrance Examination Board. (1979). National college bound seniors, 1979. New York: College Entrance Examination Board. Bridgeman, B., & Carlson, S. B. (1983). Survey of academic writing tasks required of graduate and undergraduate foreign students ( T O E F L Research Report No. 15). Princeton, NJ: Educational Testing Service. Carlson, S. B., Bridgeman, B., Camp, R., & Waanders, J. (1985). Relationship of admission test scores to writing performance of native and nonnative speakers of English (TOEFL Research Report No. 19). Princeton, NJ: Educational Testing Service.

440

TESOL QUARTERLY

College Entrance Examination Board. (1989). ATP guide for high schools and colleges, 1989-90. New York: Author. Crossette, B. (1990, January 2). Vast cheating forces many in India to retake scholastic test. The New York Times, p. Al0. Crouse, J., & Trusheim, D. (1988). The case against the SAT. Chicago: University of Chicago Press. Davis, C. D., & Davis, F. B. (1953). CEEB Achievement Test in English Composition. In O. K. Buros (Ed.), Fourth mental measurement yearbook (pp. 177-179). Highland Park, NJ: Gryphon Press. Educational Testing Service. (1989). TOEFL Test of Written English Guide. Princeton, NJ: Author. Educational Testing Service. (1990a). 1990-91 bulletin of information for TOEFL and TSE. Princeton, NJ: Author. Educational Testing Service. (1990b). TOEFL and TSE: 1990-91 test center reference list. Princeton, NJ: Author. Elliott, D. (1990, April 9). The lost generation. The New Republic, p. 23. Fallon, M. A. (1989). The TOEFL Test of Written English: An updated overview. ESL in Higher Education Newsletter, 9 (l), 7. Federal Trade Commission. (1979). Effects of coaching on standardized admission examinations: Revised statistical analyses of data gathered by Boston regional office of the Federal Trade Commission. Washington, DC: Federal Trade Commission, Bureau of Consumer Protection. Godshalk, F. I., Swineford, F., & Coffmann, W. E. (1966). The measurement of writing ability. New York: College Entrance Examination Board. Greenberg, K. L. (1986). The development and validation of the TOEFL writing test: A discussion of TOEFL Research Reports 15 and 19. TESOL Quarterly, 20 (3), 531-544. Hale, G. A., Stansfield, C. W., & Duran, R. P. (1984). Summaries of studies involving the Test of English as a Foreign Language, 1963-1982 (TOEFL Research Report No. 16). Princeton, NJ: Educational Testing Service. Hamp-Lyons, L. (1989). Newbury House TOEFL Preparation Kit: Preparing for the Test of Written English. New York: Newbury House. Henning, G. (1990, March). Current research on the TOEFL Test of Written English. Paper presented at the 24th Annual TESOL Convention, San Francisco, CA. Hoffmann, B. (1962). The tyranny of testing. New York: Collier Books. Holtzclaw, H. (1986). Current TOEFL research. In C. W. Stansfield (Ed.), Toward communicative competence testing: Proceedings of the second TOEFL invitational conference (TOEFL Research Report No. 21, pp. 3-9). Princeton, NJ: Educational Testing Service. Jackson, R. (1980). The SAT: A response to Slack and Porters Critical appraisal. Harvard Educational Review, 50, 382-391. Land, R. E., & Whitley, C. (1989). Evaluating second language essays in regular composition classes: Toward a pluralistic U.S. rhetoric. In D. Johnson &D. H. Roen (Eds.), Richness in writing: Empowering ESL students (pp. 284-293). New York: Longman. THE TOEFL TEST OF WRITTEN ENGLISH 441

Light, R. L., Xu, M., & Mossop, J. (1987). English proficiency and academic performance of international students. TESOL Quarterly, 21 (2), 251-262. Marron, J. E. (1965). Preparatory school test preparation: Special test preparation, its effect on College Board scores, and the relationship of affected scores to subsequent college programs (Study No. l-Al.02-63011). West Point, NY: U.S. Military Academy. Nairn, A. & Associates. (1980). The reign of ETS: The corporation that makes up minds. Washington, DC: Ralph Nader. Owen, D. (1985). None of the above: Behind the myth of scholastic aptitude. Boston: Houghton Mifflin. Pallone, N. J. (1961). Effects of short- and long-term developmental reading courses upon SAT verbal scores. Personnel and Guidance Journal, 39, 654-657. Pike, L. W. (1979). An evaluation of alternative item formats for testing English as a foreign language (TOEFL Research Report No. 2). Princeton, NJ: Educational Testing Service. Pike, L. W., & Evans, F. R. (1972). Effects of special instruction for three kinds of mathematics aptitude items (College Entrance Examination Board Report No. 1). New York: College Entrance Examination Board. Pitcher, B., & Ra, J. B. (1967). The relation between scores on the Test of English as a Foreign Language and ratings of actual theme writing (Statistical Report No. 67-9). Princeton, NJ: Educational Testing Service. Purves, A. C. (Ed.). (1988). Written Communication Annual: Vol. 2. Writing across languages and cultures. Newbury Park, CA: Sage. Rosser, P. (1989). The SAT gender gap. Washington, DC: Center for Women Policy Studies. Roy, A. (1987, March). Teachers and administrators concerns about the TOEFL Test of Written English. Paper presented at the Annual Conference on College Composition and Communication, Atlanta, GA. Savignon, S. J. (1986). The meaning of communicative competence in relation to the TOEFL program. In C. W. Stansfield (Ed.), Toward communicative competence testing: Proceedings of the second TOEFL invitational conference (TOEFL Research Report No. 21, pp. 17-30). Princeton, NJ: Educational Testing Service. Shostak, J. (1987). How to prepare for the College Board Achievement Tests in English. New York: Barrons Educational Series. Slack, W. D., & Porter, D. (1980). The SAT: a critical appraisal. Harvard Educational Review, 50, 154-175. Stansfield, C. W., & Webster, R. (1986). The new TOEFL writing test. TESOL Newsletter, 20 (5), 17-18. Vopat, J. (1982, November). Guilty secrets of an ETS grader. Washington Monthly, pp. 44-47. White, E. M. (1986). Pitfalls in the testing of writing. In K. Greenberg, H. Wiener, & R. Donovan (Eds.), Writing assessment: Issues and strategies (pp. 53-78). New York: Longman. 442 TESOL QUARTERLY

TESOL QUARTERLY, Vol. 24, No. 3, Autumn 1990

Student Input and Negotiation of Meaning in ESL Writing Conferences

LYNN M. GOLDSTEIN
Monterey Institute of International Studies

SUSAN M. CONRAD
Central Washington University

Research and practice in composition pedagogy suggest that student-teacher conferences play an important role in helping students become more effective writers. Many students, teachers, and researchers believe that conferences are valuable because they allow students to control the interaction, actively participate, and clarify their teachers responses. This paper reports the results of a study that examined the degree to which these characteristics were present in conferences between one teacher and each of three students enrolled in an advanced ESL composition course. In addition, the study looked at the students texts to determine how students dealt with the revisions discussed in the conferences and the role negotiation of meaning played in the success of such revisions. There were large differences in the degree to which students participated in the conferences and negotiated meaning. In addition, students who negotiated meaning made revisions in the following draft that improved the text. In contrast, when students did not negotiate meaning, even when they actively participated in the conference, they tended either not to make revisions or to make mechanical, sentence-level changes that often resulted in texts that were not qualitatively better than previous drafts.

Student-teacher writing conferences are widely recommended in composition pedagogy and many claims have been made about their role in helping students become more effective writers. These claims, however, remain unverified for second language writers because none of the research has examined the discourse that takes place in conferences or the relationship between this discourse and subsequent revision for these writers. In fact, most claims for both 443

native-speaker and ESL writers are based on the participants impressions of, or attitudes towards, conferences. In a study of native speakers, Carnicelli (1980) reviewed students evaluative comments towards their conferences. On the basis of these, he concluded that conferences are a more effective means of feedback than are written comments because conferences allow students to express their opinions and needs, and to clarify teachers comments when they are not understood: If a teachers response is unclear the student can simply ask for an explanation (p. 108). Zamel (1985) and Sokmen (1988) reach similar conclusions for conferences with nonnative speakers. Zamel discovered that ESL students often found written comments difficult to understand. Thus, she suggests that teachers need to hold conferences with students because dynamic interchange and negotiation is most likely to take place when writers and readers work together face-toface (p. 97). Sokmen concurs, stating that responding in conferences is more effective than in writing because you, the teacher, can interact dynamically with the students to understand the intent (p. 5). The above claims, however, are based, not on an examination of discourse that actually occurs in conferences, but on students and teachers evaluations of conferences. The few studies that have examined actual discourse have focused on native-speaker conferences. Freedman and Katz (1987) analyzed transcripts of several conferences and found that the discourse within these conferences had predictable parts: openings, student-initiated comments and questions, teacher-initiated comments and questions, reading of the paper, and closings. Examining one conference in detail, they discovered that the teacher and student followed interfactional rules that placed the conference somewhere between (p. 77) conversational turn-taking rules described by Sacks, Schegloff, and Jefferson (1974) and the rules of classroom turn-taking as described by Mehan (1979). While the teacher initiated many questions to guide the student, the student supplied the direction and content of the conference. Freedman and Katz hypothesized that a students input and control of the discourse accounts for the effectiveness of conferences in improving student writing. However, they did not actually look at the relationship between these factors and subsequent revisions or papers to test this. Walker and Elias (1987) compared the discourse in conferences rated highly by tutors and students to those rated poorly. Highly rated conferences were characterized by a focus on the student, with a discussion of criteria for successful writing and with an evaluation of the students work. Low-rated conferences were 444 TESOL QUARTERLY

dominated by the tutor and contained repeated requests for explanations, either by the tutor or student or both. Since success in this study is defined by tutor and student evaluation, there is no discussion of whether writing or revisions that occurred after the more successful conferences were more effective than those that occurred after less successful conferences. Researchers have also studied the variation among students in the discourse they produce within conferences. Freedman and Sperling (1985) examined the conferences of 4 native-speaker students: 2 high-achieving students and 2 low-achieving students. The highachieving students elicited more praise from the teachers while lowachieving students tended to nominate topics that alienated (p. 128) the teachers. Freedman and Sperling conclude that the interactions in conferences vary and that these differences in conversational interaction signal the possibility of differential instruction (p. 128). These researchers do not, however, examine the relationship between such differential instruction and student success. The relationship between the discourse created in conferences and subsequent revision or overall writing improvement has been studied by Jacobs and Karliner (1977). They compared the conferences of two native-speaker students to determine if the differences in the roles played by teacher and student corresponded to differences in the revisions made in subsequent drafts. They found that the student who engaged in exploratory talk and who initiated more discussion in the conference made revisions that contained deeper analysis of the subject. In contrast, the student who deferred to the teacher, with the teacher acting as an expert who gives suggestions even before hearing the students ideas, made more surface-level changes and never solved the deeper problems in content. Jacobs and Karliner conclude that the type of verbal interaction within the conference does influence the type of subsequent revision made. We must be cautious in extending the conclusions of these studies to ESL student-teacher conferences. First, there is very little research that examines actual conference discourse and/or conference discourse in relation to subsequent revision. Second, we cannot extrapolate from studies where the subjects were native speakers of English because we cannot assume that nonnative speakers will behave in conferences in the same ways that native speakers behave. THE STUDY In our study we sought to answer the following questions: STUDENT INPUT IN ESL WRITING CONFERENCES 44.5

1. To what extent do ESL writing conferences ensure student input? 2. To what extent is meaning negotiated in ESL writing conferences? (See Figure 1 for a definition of negotiation of meaning.) 3. What is the relationship between the discourse in the conference and successful revision in the subsequent draft?

The Educational Context Subjects were selected from 21 students in an advanced ESL composition class at a large urban university. The teacher was an experienced ESL composition instructor who had been using conferences as an integral part of her courses for the previous 4 years. The students wrote multiple drafts of expository papers, had a scheduled 20-minute conference every other week to discuss the draft they were working on, and received written feedback on another draft in the week between conferences. The teacher did not read the drafts that were discussed in conference until the actual conference and students were asked to be ready to identify areas they wanted to discuss when they came to conference.

METHODOLOGY Subjects Three students were selected from three different cultural backgrounds. The students had roughly equivalent proficiency, as determined by a holistic evaluation of all the papers each had written during the semester. They were in the last course of an ESL sequence that leads to Freshman Composition. Each demonstrated a working knowledge of academic rhetoric, and evidenced only relatively minor and infrequent sentence-level problems. Two women and one man, all in their 20s, participated in the study. All three were full-time matriculated students in their junior year majoring in a science. All had been in the United States for 6 years and were fluent speakers of English who evidenced no difficulty in understanding or participating in spoken discourse. Two of the subjects, Tranh (from Vietnam) and Zohre (from Iran), had attended high school in their native countries; Marigrace (from the Philippines) had attended public high school in the United States. 446 TESOL QUARTERLY

Data Collection With the students permission, the teacher taped all the conferences and collected copies of each draft of every paper. Oral data consisted of tapes of ten 20-minute conferences, 3 each for Zohre and Marigrace, and 4 for Tranh. Written data consisted of 2 drafts each of 10 papers (3 papers each for Zohre and Marigrace, and 4 for Tranh). One draft of each paper was written before the conference and was discussed in the conference, and the other draft of the paper was written after the conference. Conference Data Analysis The 10 tapes of the conferences were transcribed orthographically. Our first attempts to apply established discourse analysis systems to the data did not account for elements that appeared important in the conferences. As has been the case for other researchers (Walker & Elias, 1987; Freedman & Sperling, 1985; van Lier, 1988), it became obvious that our data should suggest the categories, rather than be made to fit imposed categories. Consequently, we looked for recurring patterns and variations across students that suggested to us how the discourse was structured and what the roles of each participant were in the discourse. As new patterns and variations emerged, we went back and coded them in conferences that we had already analyzed. Through this iterative process, we identified seven features (see Figure 1) for coding. After we had finished analyzing all the conferences once, we went through them two more times (independently and then together) to ensure that our analysis was consistent across all conferences. After the features were identified and coded on the transcripts, we obtained frequency counts per conference for types of discourse structures, topic nominations, invited nominations, turns per episode, questions, and negotiations. We then calculated mean frequencies per category for each students conferences. Analysis of Revision and Negotiation One of the goals of this research was to look at the relationship between what was discussed in conference and what was revised in the subsequent draft. We recognize that many other revisions may have occurred that were not discussed in conference and that many other rhetorical problems may have remained in the drafts. However, we decided to limit ourselves in this study to an examination of only those revisions that were discussed in the conference. Our overall goal was to determine what elements (if any) in the STUDENT INPUT IN ESL WRITING CONFERENCES 447

FIGURE 1 Discourse Features Episodes: These are subunits of conferences, with a conference made up of a series of episodes. Each episode has a unique combination of topic and purpose such that a change in either or both signifies a new episode. Episodes could be interrupted by others, continuing at the end of the interruption. Discourse Structure: Each episode was characterized by a particular discourse structure. Six types of structures emerged from the data. 1. Teacher talks and student backchannels1 2. Teacher questions and student answers. 3. Teacher talks and student talks. 4. Student talks and teacher backchannels. 5. Student questions and teacher answers. 6. A combination of the above. Topic Nomination: The participant who introduced either a new topic and/or new purpose, effectively changing to a new episode, was said to have nominated the topic of the new episode. Invited Nomination: An invited nomination occurs when the participant nominates the topic in response to a question such as, What would you like to discuss? Turns: A change of speaker signified a new turn, with the exception of backchannels. There are many theoretical positions on whether or not backchannels are turns (see van Lier, 1988, for example). However, these positions vary with the data being analyzed. Thus, we do not count backchannels as turns because, while they showed the listener was attending, in our data they do not expand, comment on, agree or disagree with, or ask for clarification of what the speaker was saying. Questions: We counted the number of questions asked both by student and teacher. This category contains only those questions not used for negotiation (see below). Negotiation: Two types of negotiation were identified Negotiation of meaning, identified in many second language acquisition studies (see, for example, Long, 1983) refers to confirmation checks, comprehension checks, and clarification requests. Negotiation of revision took place not when meaning needed to be clarified, but when revision strategies needed to be clarified. These consisted of (a) the student confirming the teachers suggestion of a need for revision or the use of a revision strategy (for example, saying, So you are suggesting that I should change the order of these); (b) either the teacher checking to see if the student had understood a discussion of revision options or a student checking (for example, the teacher saying, So what strategies can you use to revise this?); (c) the student checking, while the need for revision was being discussed, to see if it would be appropriate to revise in a certain way (for example, the student saying What do you think if I added this example here?); (d) the student stating that he or she did not understand either why a revision would be necessary or how to revise.
1 Backchannefs

are verbal devices such as um-hum, yeah, and um that indicate that a listener is attending to a speaker.

conference discourse concerning revision appeared to influence

whether and how the students revised those areas. After we had analyzed the conference transcripts, we looked at the student papers. We compared the conference draft to that written subsequently, examining those places in the papers that had been identified in the conferences as needing revision. Through this 448 TESOL QUARTERLY

process, a pattern began to emerge: Revisions seemed to occur when they had been negotiated in the conference. Working with this hypothesis, we went back to the transcripts and identified all the discussions of revision and categorized them on the basis of whether negotiation had taken place. We again compared the draft being conference with the one written after the conference, this time to discover which revisions had been made and how successful these had been. In the determination of successful revision, each of the researchers analyzed the written data individually, then compared and discussed categorizations, reaching consensus on what was successful or not, and why. Next, we compared negotiated discussions of revision to nonnegotiated ones to see the degree to which each resulted in successful, unsuccessful, or no revision. We defined successful revisions as those we judged had solved or improved upon a rhetorical problem discussed in the conference while being consistent with the writers purpose, main point, and audience. This also allowed us to credit as successful those revisions that solved the rhetorical problem under discussion even if, when a strategy had been discussed in conference, the student chose to use a different one. To illustrate, Tranh wrote a first draft about discrimination in which he had confused types of discrimination with causes of discrimination and in which he had arrived at a superficial discussion. In the conference, the teacher elicited the fact that Tranhs purpose was to examine the causes of discrimination so that people could arrive at solutions to it. The teacher and Tranh then went on to examine whether or not his purpose had been achieved, discovering that it had not been and that Tranh was confused about the difference between cause and type of discrimination. As the conference unfolded, they jointly generated possible causes of discrimination and discussed how to focus on and develop only those parts of his paper related to his purpose. After the conference, Tranh rewrote his introduction making it fit his purpose; he kept in causes he had discussed in the previous draft while he eliminated any discussion from the previous draft on types of discrimination; he expanded his discussion of causes by adding ones he had not mentioned in his previous draft; he provided concrete illustrations for some of the causes he was writing about; he completely rewrote his conclusion to be consistent with his purpose. These revisions were judged successful since they solved rhetorical problems of the previous draft. Importantly, he was able to decide on his own which parts of the text fit his purpose and should remain and which didnt and should be removed. Also, he STUDENT INPUT IN ESL WRITING CONFERENCES 449

was able to generate causes that were not in his previous draft and that had not been discussed in conference. It is important to note that our definition of success is local: We were examining only the relationship between revisions discussed in the conference and the revisions that appeared in the subsequent draft. We recognize that future research needs to address rhetorical issues not discussed in the conference, as well as the long-term effects of conferencing on writing quality and revision. RESULTS Conference Data The mean scores for each discourse feature and discourse structure are displayed in Tables 1 and 2. These scores demonstrate that there was much variation across the students in the amount of interfactional work they did in their conferences. Frequencies for individual conferences are not reported because there was little or no variation across each students conferences.
TABLE 1 Comparison of Student input Discourse Features

The three students differed greatly in the amount of input they contributed. First, the degree to which each set the agenda can be seen in the percent of nominations (see Table 1). While Zohre and Tranh contributed roughly half of the topic nominations, Marigrace contributed only one fifth. Second, these three students differed in how much interfactional work they did building the discourse (see Table 2). Tranh consistently did more work than Zohre, who in turn consistently did more than Marigrace. Although Zohre and Tranh made about the same percent of topic nominations (Table 1), twice 450 TESOL QUARTERLY

as many of Zohres nominations were invited nominations (41% for Zohre vs. 20% for Tranh). Marigraces nominations, in addition to being relatively infrequent (19.50%), were often invited (42.88% of her nominations were invited). Marigrace was also more often invited to contribute input to the conference in other ways. For example, the teacher frequently used questions (14 times per conference) with Marigrace; in contrast, she asked far fewer questions of Zohre and Tranh (6 and 6.75 per conference respectively) who more often voluntarily contributed to the conferences. The students also differed in the degree to which they clarified meaning (Table 1). Marigrace was responsible for only 33.20% of the meaning negotiations (mean per conference = 1.66); Zohre was responsible for 55.75% (mean per conference = 6.30) and Tranh for 60.78% (mean per conference = 7.75). This is another measure demonstrating that Tranh did the most conversational work in the conferences, Zohre the next, and Marigrace considerably less than either of the other two. The amount of work that the students did is also reflected in the degree to which they used each type of discourse structure (see Table 2). Those episodes where the student did less work than the TABLE 2 Comparison of Student Input: Discourse Structure (%)

teacher (teacher questions/student answers, and teacher talks/ student backchannels) occurred most frequently with Marigrace (60.60%), the next most frequently for Zohre (50.00%), and STUDENT INPUT IN ESL WRITING CONFERENCES 451

considerably less frequently for Tranh (14.28%). In contrast, episodes where the student did more of the work (student questions/teacher answers, and student talks/teacher backchannels) never occurred in Marigraces conferences, occurred 8.33% of the time in Zohres conferences, and 14.28% of the time in Tranhs. Episodes in which student and teacher shared the work occurred least frequently for Marigrace (21.21%), more frequently for Zohre (36.11%) and most frequently for Tranh (53.57%). The Relationship Between Revision and Negotiation Table 3 presents the results of the analysis of the relationship between the revision of the written drafts and negotiation of revisions. These results support our hypothesis that there is a positive relationship between negotiation and successful revision.
TABLE 3 Negotiations of Revisions (%)

When teacher and student negotiated revisions, the ensuing revisions were almost always successful (see Table 3). In the following excerpt from one of Zohres conferences, for example, the teacher and Zohre discuss the need to include more concrete details in a paper written as a letter to convince a friend to come for a visit. In lines 9-10 Zohre negotiates by asking if a certain revision strategy would be appropriate, and in lines 27-29 she checks her understanding of the number of examples needed: T: Um:: (teacher reading from Zohres text) In addition there are many cheap ethnic restaurants in which can satisfy the taste of an adventurist person. Heres another place where I think that 452 TESOL QUARTERLY

someone could get a real sense of the place um by describing a little bit the kind of food you might get in one particular restaurant a Thai restaurant or something like that Um the sensations the taste what the food looks like (.) like ok in different restaurants like I say what kind of food they have Umhum, um: you know if I were if I were gonna write to someone and I was gonna say um you can find many cheap ethnic restaurants you can sa-satisfy the taste of an adventurist person that means a taste Ive never tasted before urn if Im an adventurous person Im gonna try something new. So Id try to think of something exotic you know for example you could try Thai food and you can taste the hot and sweet flavors in combination with each other with coconut milk urn um make their mouths water ya you know in a sense. You dont have to the purpose of the paper isnt to describe the restaurant so you dont have to go into any great detail but if you could just have one line the dominant flavors of a particular cuisine I think would make it very vivid o.k. just one example ya would be enough ya In the subsequent draft of the paper Zohre adds details that give the reader a more vivid picture of one of the restaurants: there are many cheap restaurants in which can satisfy the taste of an adventurist person. For example, there is a Moroccan restaurant which serves you with a spicy lentil based soup, platters of Arabic bread and different entrees, most of which are chicken or lamb stewed with various combinations of fruits and vegetables. In this restaurant you eat with your fingers like North African tradition. These details were of Zohres own making, not a copy of those given as an illustration by the teacher. She extracted the principle and applied it to her own writing. Table 3 shows that all three students, not only Zohre, had a higher percentage of successful revisions when negotiation had taken place. Every time Marigrace and Zohre negotiated, their subsequent revisions were successful. When Tranh negotiated, 91.66% (11/12) of his revisions were successful. None of Tranhs were unsuccessful, but there is one instance where he did not revise STUDENT INPUT lN ESL WRITING CONFERENCES 453

despite negotiation. This lack of revision, however, may be due to a discussion between the teacher and Tranh in which this revision was determined to be of relatively minor importance. In contrast, when the students did not negotiate (i.e., when the teacher made revision suggestions and the student backchanneled), the subsequent revisions were often either unsuccessful or not attempted at all (Table 3). For example, in the same conference referred to above, Zohre did not negotiate when the teacher suggested using more specific details in another part of the paper. Zohre only backchanneled while the teacher spoke T: . . . where I fel I felt you needed the detail and (teacher reading Zohres text) it has really nice and big campus this Z: uhuh T: word nice means nothing Z: oh T: O.K. what does nice mean Z: O.K. like I know T: um you might want to describe the campus briefly Z: o.k. T: here its set on a hill lots of green and the architecture you know Z: ya The only change Zohre made in her next draft is from it has really nice and big campus to It has a nice and big campus. She did not address the need for specific details. We can also contrast this excerpt to the previous one While Zohre did not do a lot of conversational work in either excerpt, in the previous one she did negotiate the revision suggestion, and she acted on that suggestion successfully. She did not act on the nonnegotiated suggestion even though both suggestions addressed the need for more detail. The relationship between lack of negotiation and unsuccessful/ unattempted revision holds for all three students (Table 3). When revisions had not been negotiated, Zohre either revised unsuccessfully (40%) or did not attempt revision at all (60%). While Tranh had some successful revisions (20%) when he didnt negotiate, the majority were unsuccessful (60%) or not attempted (20%). Although Marigrace had the highest number of successful revisions in nonnegotiated instances (33.33%), she still produced a greater number of unsuccessful revisions (66.67%). In three out of the four cases of nonnegotiated successful revision (both of Marigraces and one of Tranhs), the discussion included specific instructions for very mechanical revisions, such as the switching of the order of two 454 TESOL QUARTERLY

sentences. And, in these cases, the instructions were restated. The simplicity of these revisions and the restatements may explain why, even without negotiation, these revisions were successful. DISCUSSION Our results do not support some of the claims that have been made for conferences. Much of the literature suggests that the very act of conferencing (see, for example, Carnicelli, 1980; Zamel, 1985) leads students to contribute input setting the agenda, making their needs known, expressing their ideas and opinions, and asking questions and clarifying meaning. However, we have not found this to be the case for all the students in this study. Like Jacobs and Karliner (1977) and Freedman and Sperling (1985), we have found variation across students in the way they interact with the teacher in a conference. Marigraces conferences were characterized by the teacher generating most of the input and doing most of the conversational work: The teacher nominated the topics, the teacher did most of the talking, and the teacher used questions to engage Marigrace in the interaction. Marigrace primarily backchanneled. Tranhs conferences contrasted sharply with those of Marigrace. His were characterized by student and teacher equally contributing topic nominations, questions and talk, and backchannels; they shared in the building of the discourse. Zohres conferences fell between these extremes. Thus, while a student may contribute input to the conference, may set the agenda, and may negotiate meaning, these are not guaranteedeven in conferences with the same teacher. Each student who participates in a conference brings to that conference a unique personality that may affect the ways in which that student behaves in the conference. For example, the teachers impression, before the study began, was that Tranh was the most assertive, Zohre the next, and Marigrace the least. If this is the case this might be one explanation of why Tranh contributed the most input, Zohre somewhat less, and Marigrace the least. The teachers role in producing variation in the conference discourse needs to be considered. One possibility is that the teacher may have adjusted to the students individual discourse style, thus reinforcing it, whether or not this resulted in the student actively participating in the conference. For instance, the greater amount of conversational work done by the teacher in Marigraces conferences, asking many questions for example, may be an adjustment to Marigraces lack of voluntary input and may have encouraged her to continue to rely on the teacher to do most of the interfactional work in the conference. STUDENT INPUT IN ESL WRITING CONFERENCES 455

However, it is also possible that the teacher gave differential treatment to students for reasons other than the teacher adjusting to the students own discourse styles, as Freedman and Sperling (1985) suggested in their study. In our study, once students behaviors in conference and in class had been observed, the teacher may have subconsciously behaved in ways consistent with her expectations of the students. The teacher may have accepted less participation from Marigrace in the conference because she evaluated her as a less capable student on the basis of her initial conferences and revisions. On the other hand, the teacher may have been more encouraging of discussion with students such as Tranh and Zohre who more actively participated in conferences and who revised their papers more successfully. In addition, as members of diverse cultures, ESL students come with rules of speaking that may conflict with those of U.S. classrooms and with those teachers might like to see operate in conferences. These rules of speaking may also play a role in the students perceptions of theirs and their teachers roles in a conference. As Phillips (1972) has demonstrated, for example, students often bring to the classroom rules of speaking from their own cultures that work differently from those of the new culture. In our study, it is possible that the variation we have seen across the three students may result, at least in part, from these students using culturally diverse rules for how much teachers and students control the discourse when interacting with each other. Students may have also acquired rules of speaking from typical U.S. classrooms that may also conflict with those of the conference. For example, in many U.S. classrooms it is the teacher who typically initiates and questions, the student who responds, and the teacher who evaluates (see, for example, Mehan, 1979). Again, this may result in some students contributing more input than others. In our study, it is possible that Marigrace had been influenced by her high school education in the United States and was consequently following that discourse structure. In the end, however, regardless of why variation across students existed, the results show that conferences do not necessarily do what the literature claims they dothey do not necessarily result in student input. In sum, instructional events such as conferences are dynamic, lending themselves to the myriad influences and interpretations of their participants. Conferences also do not necessarily result in revision, and when revision occurs after a conference, it is not always successful. Our data suggest that negotiation of meaning does play a role in subsequent revision and we need to ask why negotiation would lead 456 TESOL QUARTERLY

to more successful revisions. First, just as negotiation clarifies meanings in ordinary conversations, negotiation in the conference may clarify the need for revision and the strategies to undertake the revision. Students, therefore, may understand more clearly what to revise, how to revise, and why they need to do so. In addition, negotiation may lead to better retention of what has been discussed. Negotiation requires the student to be more actively involved in the discussion either by asking questions or answering them, which may lead to better retention (see, for example, Stevick 1976). Finally, it is also possible that students negotiate points where they most clearly see the need for revision; they may already be predisposed to revising in the area being negotiated and may be more interested in discussing how to do so. For example, in Zohres case, there are several instances where she shows very little interest in the revision the teacher has nominated for discussion, and in fact she does not make these revisions in subsequent drafts. Although we do not know all the characteristics of discourse that might lead to successful revision, this study suggests that negotiation plays an important role. The student who was conversationally active (Tranh) and the student who was more dependent on teacher input and direction (Marigrace) both demonstrated more successful revisions when negotiation occurred. However, we have seen that despite the claims made, conferences do not ensure that negotiation will take place any more than they necessarily result in a great deal of student input and control. IMPLICATIONS AND SUGGESTIONS FOR FUTURE RESEARCH We cannot expect that students will come to writing conferences understanding the purposes of such conferences, the rules of speaking, and the respective roles of the participants. Since the quality of their conferences and revisions can be affected by participant expectations, we must teach students the purposes conferences can serve, and stress that the discourse and the teacherstudent relationship can vary greatly between a conference and classroom. In a sense, we need to give students permission to break the rules they may have learned previously and we need to teach them new rules for a new speech event. This can be accomplished in several ways. Teachers can have students discuss the rules of speaking the students feel govern classroom behavior, making these rules explicit. The teacher can then discuss conferencing with students in terms of the goals of conferences, the roles of participants, and the rules of speaking. Conferences and classrooms can be compared and contrasted so that STUDENT INPUT IN ESL WRITING CONFERENCES 457

students understand the differences and gain permission to behave differently in conferences. Furthermore, our results suggest that students might benefit from explicit instruction concerning the importance of their conversational input and of the negotiation of meaning; in addition, students need to be taught concrete ways to achieve these goals. We have experimented with bringing the transcripts of the students in this study into our ESL writing classes and having our students analyze them. From this, the students have seen the differences among the conferences, and they have learned specific techniques for contributing input and negotiating meaning and revisions. As teachers, we need to examine our own behaviors as well. One means of doing so is to tape our conferences (with permission from our students) and then examine them with particular questions in mind. For example, we can ask if we control the discourse, thereby discouraging our students from participating in the conference. By coding how and the degree to which the teacher and student nominate topics, and the relative amounts of teacher and student talk, we can begin to answer this question. We can also compare the treatment we give to different students, seeing if all are given equal opportunities to contribute input and negotiate meaning. In addition, we can examine the degree to which we negotiate meaning when we want to clarify for ourselves or for the students. There are many questions to be asked, and taping and analyzing the discourse in our conferences is one means of answering these questions. In composition research we must move beyond an assessment of the effectiveness of conferences based primarily on student and teacher evaluations. While it is important to know the participants attitudes towards conferences, and the criteria by which students and teachers judge the effectiveness of conferences, we need to understand how discourse is jointly built by the participants, and what characteristics of the discourse influence success, defined as either improvements in subsequent revisions or in terms of more positive student attitudes. We also need studies that compare the success of revisions made after conferences with those made after written comments so that we can examine the relative effectiveness of these different forms of feedback. Finally, ESL composition teachers are indebted to those who teach native speakers and who have conducted research with native-speaker writers. They have taught us much about composing, and over time we have discovered that their findings are often applicable to ESL students. However, while the results of this study are similar to those of Jacobs and Karliners (1977) study of native 458 TESOL QUARTERLY

speakers, we should keep in mind that ESL students bring with them diverse cultures and languages. This fact argues for more research conducted with an ESL population. There may be, for example, many student characteristics, such as culture, that potentially affect how students conference or how their teachers respond to them. For that matter, teachers may differ greatly from each other in how they interact with their students in conferences (see, for example, Katz, 1988). These factors, among many others, need to be systematically studied since writing conferences are not stable entities but rather, dynamic events affected by context and participants.

ACKNOWLEDGMENTS
This paper is a revised version of presentations made at the 22nd Annual TESOL Convention in Chicago, March 1988; the Second Language Acquisition Forum in Honolulu, March 1988; and the 1989 Conference on College Composition and Communication in Seattle, March 1989. We would like to thank the students who participated in this study and Anne Katz, Joanne Cavallero, Kathi Bailey, Tim Hacker, and two anonymous TESOL Quarterly reviewers for their valuable suggestions on the paper.

THE AUTHORS
Lynn M. Goldstein is Associate Professor of TESOL/applied linguistics and the coordinator of campus-wide writing courses at the Monterey Institute of International Studies. She has published articles on second language acquisition and on dialogue journals, and is the 1987 recipient of the TESOL/Newbury House distinguished research award. Susan M. Conrad teaches ESL at Central Washington University. She has also taught ESL and composition in California, New York City, and Korea, and as a Peace Corps volunteer in Lesotho. She has published on dialogue journals, and has made presentations on discourse analysis and composition.

REFERENCES
Carnicelli, T. A. (1980). The writing conference A one-to-one conversation. In T. R. Donovan & B. W. McClelland (Eds.), Eight approaches to teaching composition (pp. 101-131). Urbana, IL: NCTE.

STUDENT INPUT IN ESL WRITING CONFERENCES

459

Freedman, S., & Katz, A. (1987). Pedagogical interaction during the composing process: The writing conference. In A. Matsuhasi (Ed.), Writing in real time: Modeling production processes (pp. 58-80). New York: Academic Press. Freedman, S., & Sperling, M. (1985). Written language acquisition: The role of response and the writing conference. In S. Freedman (Ed.), Acquisition of written language: Response and revision (pp. 106-130). Norwood, NJ: Ablex. Jacobs, S., & Karliner, A. (1977). Helping writers to think: The effect of speech roles in individual conferences on the quality of thought in student writing. College English, 38, 489-505. Katz, A. (1988). Responding to student writers: The writing conferences of second language learners. Unpublished doctoral dissertation, Stanford University, Palo Alto. Long, M. H. (1983). Native speaker/non-native speaker conversation and the negotiation of comprehensible input. Applied Linguistics, 4, 126-141. Mehan, H. (1979). Learning lessons: Social organization in the classroom. Cambridge, MA: Harvard University Press. Philips, S. U. (1972). Participant structure and communicative competence: Warm Springs children in community and classroom. In C. Cazden, V. Johns, & D. Hymes (Eds.), Functions of language in the classroom (pp. 370-394). New York: Teachers College Press. Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematic for the organization of turn-taking for conversation. Language, 50, 694-735. Sinclair, J. McH., & Coulthard, R. M. (1975). Towards an analysis of discourse. London: Longman. Sokmen, A. A. (1988). Taking advantage of conference-centered writing. TESOL Newsletter, 22 (l), 1,5. Stevick, E. W. (1976). Memory, meaning, and method. Rowley, MA: Newbury House. van Lier, L. (1988). The classroom and the language learner. London: Longman. Walker, C. P., & Elias, D. (1987). Writing conference talk: Factors associated with high- and low-rated writing conferences. Research in the Teaching of English, 21, 266-285. Zamel,7V. (1985). Responding to student writing. TESOL Quarterly, 19 (1), 79-97.

460

TESOL QUARTERLY

TESOL QUARTERLY, Vol. 24, No. 3, Autumn 1990

Teaching the English Articles as a Binary System

PETER MASTER
California State University, Fresno

The English article system can be taught as a binary division between classification (a and ) and identification (the). All the other elements of article usage can be understood within this framework, allowing a one form/one function correspondence for a and the. Furthermore, the notions of classification and identification can be introduced as distinct concepts before the various rules for article usage are taught. This simplified schema is presented as a pedagogical tool for selecting the appropriate article, a universally acknowledged difficulty for nonnative speakers of English.

The English article system is one of the most difficult aspects of English grammar for nonnative speakers and one of the latest to be fully acquired. It appears deceptively easy to most native speakers, who usually have difficulty articulating the rules for article usage much beyond It sounds right. And since the articles are either unstressed (a(n) and the) or invisible (the zero article []), it is difficult for students to gain sufficient input from native speakers to acquire the system. Furthermore, the articles, like the other lateacquired elements rarely cause misunderstanding when misused in spoken language. It is usually only when ESL/EFL students have to write that they become aware that they lack the basic concepts necessary to guide them in choosing the correct article. There are comparatively few attempts in the literature to provide a coherent grammar for teaching the articles as a system. Whitman (1974) bases his pedagogical sequence on the assumption that English article structure is a sequence of quantification and determination rather than a choice between specified and unspecified (p. 253). He delineates six steps for teaching the system:
1. Quantity (Singular and plural count nouns)

John has a book versus John has [] four books.

461

2. Generic plural All apples are red versus [] Apples are red. 3. Noncount nouns. (Noncount vs. count and a lot of vs. much and many) John drank a lot of water versus John bought a lot of books. Do we have much water? versus Do we have many books? 4. Determiners (which- NP questions and first/subsequent mention) Which books are red? The red books are on the table. I read a book. The book was called Dracula. 5. Quantity and determiner One of the books on that table is blue. 6. Generic articles Elephants never forget. An elephant never forgets. The elephant never forgets. Whitman maintains that generic usages of a/n and the are not that commonly found and are probably best delayed considerably in teaching the article system (p. 261). McEldowney (1977) takes a form/content approach to the teaching of the articles. She says that four types of meaning are communicated by the presence or absence of a, the, or -s in various combinations in noun phrases: (a) general or particular, (b) any or special, (c) countable or uncountable, and (d) singular or plural. She then cites three universal types of error which she claims occur irrespective of Ll: (a) omission of a/the/-s, (b) wrong insertion of a/the/-s, and (c) confusion of a/the/-s. With these taxonomies in view, McEldowney proposes the following stages of learning (p. 110): 1. Classification a + N (any one) versus the + N (the special one). Choose a bag. versus Take the red bag. N + s (plural classification) These are bags. 2. Plurality some + N + s (any ones) Choose some bags from the collection. the + N + s (the special ones) Take the red bags. 462 TESOL QUARTERLY

3. Mass or substance N (the substance in general) Mud is found at the bottom of rivers. some + N (any substance) Some mud is grey; some mud is black. the + N (the special substance) Point to the black mud. 4. Numbered specific; generic numeral N + s (any numbered ones) Choose six pens from the collection. a + N/the + N (ones in general)/the + N + s An elephant never forgets. The elephant never forgets. The elephants never forget. McEldowneys sequence links the English articles to three concepts: any (a) to mark choice, special (the) to mark specification, and general (-s and later a and the) to mark generalization. Her decision to resort to the -s plural in order to avoid referring to the zero article, however, results in the neglect of the relatively frequent + noncount noun category (e.g., This battery needs water). Pica (1983) argues not for a new pedagogical sequence but for the inclusion of discourse-related rules in the presentation of the English article system. She based her research on a perusal of the kinds of article rules typically presented in ESL/EFL grammar texts and compares them with the article use of native speakers in requesting and giving directions and ordering food at restaurants. She concludes that article use may have more to do with communication and communicative competence than with grammar and linguistic competence (p. 231) and makes the following recommendations for instruction: 1. Since articles are often not necessary in immediate environments, activities like ordering food should be practiced first as a nonfrustrating lesson for beginning students (p. 232). 2. First mention a and subsequent mention the are easy to teach from a pedagogical point of view but are not used as frequently as preforms (i.e., possessive pronouns) in natural speech. 3. Since assessing the knowledge of the hearer is often no simple matter, students should be encouraged to always use the with a qualifying description rather than just a bare noun (e. g., the nearest post office vs. the post office; the university bookstore vs. the bookstore). TEACHING THE ENGLISH ARTICLES 463

4. Dialogues should be used to provide students with relevant examples of article use and the effect of using an incorrect article should be discussed with the class to increase awareness of native usage. 5. Students should be engaged in experiences outside the classroom to foster natural acquisition. Picas points are generally well taken, especially if spoken communicative competence is the goal and especially for students at lower levels of proficiency. With more advanced students, especially if the goal is written competence (where article errors really stand out), Picas suggestions would need to be supplemented with more detailed aspects of the article system. Master (1983) presented a detailed schema for teaching the English article system, which was subsequently refined in other publications (Master 1986a, 1986b, 1988a, 1988b). This system presented a hierarchical sequence of six questions which must be asked about each noun in a piece of discourse: 1. Is the noun countable (singular or plural) or uncountable (singular)? 2. Is the noun indefinite or definite? 3. Is the noun postmodified or not? 4. Is the noun specific or generic? 5. Is the noun common or proper? 6. Is the noun nonidiomatic or idiomatic (e.g., a set phrase, a title or label reduction)? (Master, 1986b, p. 204) A controlled study (Master, 1986c) in which this 6-point schema was used in the experimental group versus no systematic treatment of the article system in the control group found significant improvement on a fairly reliable article test (r = .79) between pre- and postadministrations. The treatment, consisting of 6 hours of instruction spread over a period of 9 weeks, provided the intermediate university-level experimental group with a systematic approach to the article system in which the six steps were presented one at a time, each building on the former in a hierarchical fashion. The control group received no explicit, systematic approach to the article system, although articles were corrected on written compositions. JUSTIFICATION OF THE BINARY SCHEMA Although the schemas described above cover the majority of situations in which articles occur, they are somewhat unwieldy for 464 TESOL QUARTERLY

students to use. Lisovsky (personal communication, 1987) created an exercise in which he asked his students to identify whether a noun was [ count], [ definite], [ postmodified], or [ generic] before they selected the correct article. He found little correlation between the students ability to classify the noun and their choice of the correct article. Thus, despite considerable time spent on teaching these distinctions, students appeared not to be able to use this knowledge in choosing the article in an exercise. It would thus seem that the significant improvement on the article test described in Master (1986c) arose from the focusing of students attention on the need for articles in English rather than from any explicit method for choosing the article correctly. What is needed is a description of the articles that would conform to Bolingers (1977) notion that the natural condition of a language is to preserve one form for one meaning, and one meaning for one form (p. x), i.e., a generalized function of a and a generalized function of the. Both students and teachers would welcome a straightforward rule of thumb that accounts for article usage in the greatest number of cases. A one form/one function correspondence can be approximated when is used to classify a noun and the to identify it (as a is derived from the word one and therefore only applies to singular countable nouns, it is considered a variant form of rather than a separate category of articles). This binary division results from a rejuggling of various descriptions of the English articles in order to simplify their pedagogical presentation. The term classification, or variations of it, has been used before in describing the article system. Kruisinga (1932) used the term classifying (p. 242) to describe the secondary or generic function of the definite article whose function was to make a class noun into a synonym for the whole group (p. 245). He thus used the term for a binary division, but only to separate specific from generic usage. McEldowney (1977) used the word classify (p. 110) in describing plural countable nouns that do not require an article (i. e., take the zero article). As for identification, Grannis (1972) spoke of a conspiracy of uniqueness (p. 275) in which he sought to give a single meaning to the various uses of the; this exactly parallels the aim of the present study. He did not, however, apply the same notion to a and . In formal linguistic terms, determining the correct article in English requires the simultaneous consideration of four features: definiteness [ definite], specificity [specific], countability [count], and number [singular]. Number really only applies to [count] nouns and should therefore only be considered a feature of that subset. The four possibilities that result from combining the TEACHING THE ENGLISH ARTICLES 465

features [definite] and [specific] are shown in the following examples: la. [definite] [+specific] A tick entered my ear. b. [definite][specific] A tick carries disease. c. [+definite] [+specific] The computer is down today. d. [+definite][specific] The computer is changing our lives. The new approach that the binary schema proposes is the collapsing of the features [definite] and [specific] into a single feature [identified]. The feature [identified], or identification, thus includes the features [+definite] [+specific] whereas [identified], or classification, includes the features [definite] [specific]. The binary [ + identified] schema is essentially an argument for the primacy of Examples lb and lC over la and 1d. In other words, it suggests emphasizing the feature [ + definite ] and subsuming the feature [ + specific ] for pedagogical purposes. The feature [ + identified] thus comes close to the traditional feature [ + definite ]. A justification for this attempt to simplify the article system is set forth in Sterns (1983) description of pedagogical grammars: A pedagogical grammar is an interpretation and selection for language teaching purposes of the description of a language, based not only on linguistics but also on psychological and educational criteria (p. 186). The justification for the new terminology is that the terms identified and classified embrace a larger concept than definiteness, and, while they reduce the descriptive adequacy required in formal linguistics, they allow a better description and explanation of the article system for educational and psychological purposes. The Effect of Ignoring Specificity in Indefinite Nouns Ignoring the feature [specific] in indefinite nouns is equivalent to saying that [definite] embraces [specific] or that all uses of a (and ) are essentially generic. The specific/generic distinction indicates when a noun phrase is a real or actual noun as opposed to when it is the idea or concept of a noun. In the sentence A tick entered my ear, a tick is specific because an actual tick entered my ear. On the other hand, in the sentence A tick can carry disease, a tick is generic because it does not refer to an actual tick but rather to anything that can fit into the class of things called ticks. This is obviously a very subtle distinction. Burton-Roberts (1976), for example, argues that a scientist in John is a scientist is specific whereas a scientist in A physicist is a scientist is generic. Generic 466 TESOL QUARTERLY

NPs are difficult to identify because they are based entirely on context, i.e., they can only be determined from discourse considerations and/or the nature of the sentence. This led some researchers (e.g., Chafe 1969) to conclude that it was really the verb that determined genericness and not the noun phrase. Celce-Murcia and Larsen-Freeman (1983) speak of the potential ambiguity of the indefinite article in a sentence such as I needed a book, which allows both the specific interpretation but I didnt have it and the generic interpretation but I didnt have one. The binary schema thus seeks to diminish the importance of the difference between Examples la and lb. The argument is that whether or not we mean a specific, actual tick in Example la or a generic one in Example lb, we still classify that tick when we use the article a. These sentences could be paraphrased in the following way: 2a. Something that can be classified as a tick entered my ear. b. Something that can be classified as a tick carries disease. The major consequence of the feature [specific] (or [+generic]) is that subsequent mention constraints do not apply to the generic noun phrase, generic noun phrases do not allow the unstressed determiner some, and generic a cannot occur with nonrestrictive relative clauses. However, spending class time on a distinction that requires the same article (a in this case, although the same applies to the zero article) seems unnecessary in all but the most advanced levels of ESL/EFL instruction. This is especially true when the [ + specific] status of a noun phrase is ambiguous, as shown in the following passage from Newsweek (Lost Signals, 1989): Husbands and wives [spec] tend to bring their own scripts [spec] to a relationship [spec] and assume, mistakenly, that their spouse [+spec?] can read their emotional signals [+spec?] loud and clear. The result [+spec] of this unwitting breakdown [+spec] in communication [spec] is often a vicious cycle [spec] of attack and retaliation [spec]. Now, through adaptation [spec] of a technique [spec] called cognitive therapy [spec], counselors [spec?] are helping to patch up troubled marriages [spec?] by teaching couples [spec?] to become better senders and receivers [spec?] of emotional messages [spec?]. (p. 3) The problem in this passage is that the author shifts from a clearly generic introduction into an example, and it is hard to say (and not particularly important, in my view) whether the example refers to actual [+specific] or to representative [specific] spouses or situations, TEACHING THE ENGLISH ARTICLES 467

The Effect of Ignoring Specificity in Definite Nouns Ignoring the feature [specific] in definite nouns is equivalent to saying that [+definite] embraces [+specific] or that all uses of the are essentially specific. As already noted, the specific versus generic distinction shows when a noun phrase is the idea or concept of a noun versus when it is a real or actual noun. The same distinction is also apparent in the sentences The computer is changing our lives versus The computer is down today. Although this is a more substantial difference and one that must be recognized, generic the is a comparatively infrequent usage. In the proposed binary schema, generic the would count as [+identified] and not as classified, thus posing a potential problem for the schema. There are some who believe, however, that generic the is not so very different from specific the. Burton-Roberts (1976) states: Generic NPs mention individuals (ones that happen to be classes) just as my brother and the sun mention individuals. Such NPs then are fundamentally distinct from NPs determined by generic a, which do not mention individuals. . . . definite NPs appearing . . . in sentences which . . . mention individuals can be acceptably interpreted as generic (for the simple reason that they are themselves individuals, and have the same distribution as other NPs mentioning individuals). (p. 435) Burton-Roberts implies that the identified quality of even a generic NP like the computer is retained. That is, we do not immediately summon up the classifying notion one of a group for this NP until we have decoded the sentence in which it occurs. And even when we understand that the NP requires a generic interpretation, we seem to interpret the class through the individual. For this reason, generic the is described as the identification of a class in the binary schema and is considered to be [+identified]. This is further justified by the fact that generic the has certain qualities (e.g., it is often preferred when the noun is an agent of change) which distinguish it from generic a and (see Master, 1987, who found generic the to occur comparatively infrequently, with just under 350 instances in 50,000 words of text from Scientific American, a genre in which one would expect considerable generic usage). PEDAGOGICAL APPLICATIONS It was suggested earlier that one of the pedagogical advantages of the binary schema is that classification and identification can be presented to students as concepts before the linkage to articles is 468 TESOL QUARTERLY

explained. After this dichotomy has been explained, the countability of the noun must also be considered. Classification can be introduced by having students sort a pile of objects into categories: These are books/These are pencils/This is paper. Or a teacher can go around the classroom asking, Whats this? (Its a blackboard/Its a light switch/Its chalk.) For more advanced students, one could set classifying situations, e.g., How would you classify or describe the school/this student/the film youve been watching? (Its a language school/Shes an Italian/Its a comedy.) Or, since definitions always require classification, one could ask how a student would define a thermometer/a calculator/ a paragraph. All responses to the questions presented above require or a in the answer, the distinction entirely dependent on the number and countability of the noun. Students will sometimes spontaneously recognize this on their own. If not, it must be pointed out to them. Identification can be introduced by having students identify specific members within each category: This is the blue book/ These are the red books. The questions Which one is this?/Which ones are these? force an identifying response. Although the classifying question Whats this? could also elicit an identifying response in the real world (e.g., Its the key you accused me of stealing/its the remains of the steak youve been broiling for the last half hour), the use of the question words what for classification and which for identification is recommended in order to keep the pedagogical distinction clear. Ultimately, the student should be able to understand the difference between the two. For example, the teacher could say, pointing to a student: 3a. How would you classify this person? (Shes a student.) b. How would you identify this person? (Shes the student with the red hat./its Joan.) Names (proper nouns) identify, as do possessive determiners (e.g., my, his, their), possessive -s (e. g., Johns, the girls), demonstratives (e.g., this, that), and certain other determiners (e.g., either/neither, each, and every). The articles and a classify, as do the determiners some/any, no and one. Countability and Number The purpose of the binary schema is to simplify article choice by reducing the number of features required to correctly determine the article from four to three. We turn now to the features of TEACHING THE ENGLISH ARTICLES 469

countability and number, which the schema cannot dispense with. Countability must be considered only when the noun is classified ([identified) because identified nouns require the whether the noun is countable or not. The primary article occurring with uncountable and countable nouns is . In a tally of all the articles used with common nouns in an issue of Newsweek (1989) (N = 5004), 46% of the nouns took , 35% took the, and 19% took a. If a and are combined, the function of articles, at least in this genre, is to classify (65%), nearly twice as often as it is to identify (35%). The tally also underscores the importance of the zero article, which has often been neglected in article studies because it is difficult to count (e.g., Brown, 1973; Lamotte, Pearson-Joseph, and Zupko, 1982). One of the reasons occurs more often is that it applies to both noncount nouns and to plural count nouns, not to mention the numerous cases in which occurs with a singular count noun (e.g., at school, on edge, the smell of onion, hunting fox). Thus, nouns are typically classified with , and only with a discrete, singular count noun is a required. In other words, a fully separate feature indicating [ singular] is really not justified given the comparative infrequency of a. This reduces the number of features governing the articles from three to two: [ identified] and [count ]. The best way to assign the articles, as other writers have pointed out (e.g., Huckin & Olsen, 1983), is through a flow chart in which the classification/identification dichotomy is invoked first, followed by the count/noncount dichotomy and finally by singular/ plural considerations, as shown in Figure 1.
FIGURE 1 Chart for Determining the English Articles

470

TESOL QUARTERLY

Teaching the Details of Article Usage Within the Binary Framework After the concepts of classification and identification have been presented and practiced, a chart (see Figure 2) can be drawn on the blackboardthe left side for classification, the right for identificationto show how the details of article usage can be interpreted in a binary manner. The count/noncount distinction is then discussed as a subset of classification as this distinction is not necessary when the is present.
FIGURE 2 Summarized Aspects of Classification and ldentification

First and subsequent mention. The next step is to teach the notion of first and subsequent mention. First mention, which requires or a, is simply a form of classification in the binary schema. The first time we are introduced to a new noun, it is simply a member of a class. The classifying article can be paraphrased with the words n o particular or one we havent seen before. So we might say: "A man [no particular man] is walking down a road [no particular road] with some wood [no particular wood] ." The subsequent mention of that noun of course requires the: The man is old, the road is long, and the wood is heavy. But the reason the is required is that the nouns are now identified. To use the questions described earlier, we could ask of the first mention picture, Whats that? (Its a man/Its a road/its wood.) However, in the subsequent mention picture we could ask, Which man/road/wood is that? (Its the old man who is carrying some wood/its the road that the old man is walking TEACHING THE ENGLISH ARTICLES 471

along/Its the wood that the old man is carrying.) Notice that if we had asked the which question of the first mention picture, a logical response would have been How should I know? or, imagining an identifying response, I suppose its the one the speaker/writer is talking about. On our chart (Figure 2), we would put first mention on the classification side and subsequent mention on the identification side, and then provide students lots of practice in which they can apply just this principle. Ranking adjectives and shared knowledge. Ranking adjectives and shared knowledge automatically identify nouns, so they always require the. They both belong under the identification column of the chart. A typical way to paraphrase the in English is to say the only one. Thus, when we say the most beautiful/the next/the only city, for example, representing the three kinds of ranking adjectives (superlative, sequential, unique), we know that there is only one city that can be meant. Similarly, with shared knowledge, when we say the moon/the school/the window, representing the two most common kinds of shared knowledge (universal and regional/local), we know that only one is being referred to. It must therefore be identified. Postmodification. In teaching the effect of postmodification on article use, an amorphous shape can be drawn on the blackboard to represent a noncount noun and the word water written (using and thereby classifying it) inside that shape. The word salt/spring/lake is then written in front of that word and students are asked what the article should now be. Invariably, they will say the. This response can be used as a jumping off point to introduce the notion of postmodification because it is only when an uncountable noun is postmodified that the is required. This is introduced by drawing a broken line to isolate a small part of the amorphous shape, to limit the quantity. In this way, a sentence like Water is necessary for life can be contrasted with The water in this glass is dirty. In terms of classification/identification, when we speak of water or salt/ spring/lake water, we are speaking of no particular water but rather a type of water, which naturally comes under classification. When we postmodify water, essentially limiting it, then we have identified that water. So limiting postmodification is placed under the identification column because one of the ways to identify something is to postmodify it. After students have had a chance to practice moving from a classified to an identified noncount or plural count noun by means of postmodification, another example is introduced: A thermometer is an instrument that measures temperature. This is an example 472 TESOL QUARTERLY

where postmodification does not require identification, which is always the case in definitions or, for that matter, in any postmodifying phrase whose function is to classify or define rather than to identify. The distinction is shown in the following example: 4a. Houdini was the man who could open any lock. [identification] b. Houdini was a man who could open any lock. [classification] In the first sentence, we single out Houdini (i.e., identify him) as being the one who was perhaps the best at this particular skill. In the second, we place Houdini in a group of like others (i.e., we classify him). Defining postmodification is therefore placed under the classification heading opposite limiting postmodification. Descriptive versus partitive. A slightly more complicated version of this technique applies to postmodification with of-phrases. If the ofphrase serves to describe the headnoun (e.g., the diameter of a circle, the length of a room), then it limits that noun, which serves to identify it because there is usually only one. Furthermore, such phrases can be inverted into possessive structures (a circles diameter, a rooms length) and we have already seen that possessive determiners always serve to identify nouns. If, on the other hand, the headnoun of the of-phrase represents a portion, part (hence the term partition), or measure of the object of the preposition of (e.g., a cup of coffee, a length of eight feet), then it presents one of many possible divisions of that object (we could have a pound/bag/ teaspoon of coffee or a height/diumeter/thickness of eight feet), which serves to classify it. Partitive phrases cannot be inverted into possessive structures (*coffees cup, *eight feets length). Thus, descriptive of-phrases is placed under the identification column of the chart and partitive of-phrases under the classification column. Intentional vagueness. One special use of descriptive of- phrases with rather than the occurs frequently in scientific prose. A phrase like the replication of cells, a typical descriptive of-phrase, which usually takes the, is sometimes rendered with the zero article, replication of cells. Christophersen (1939) noted that when continuate-words [i.e., noncount nouns] and plurals are used in zero-form [i.e., with the zero article]. . . . only the common properties of the object denoted are thought of, not special features, and as for quantity, the limits are imagined as vague and indefinite (p. 66). This usage is commonly referred to as intentional vagueness. Replication of cells represents a less focused notion and hence more a type of classification than the replication of cells, TEACHING THE ENGLISH ARTICLES 473

which represents a focused, identified activity. Thus, intentional vagueness is also placed in the classification column. Other classifying conditions. One of the more important classifying conditions is the description of general characteristics, often with the verb have: 5a. A zebra has [] stripes. b. Jupiter has a red spot. c. San Francisco has a population of 800,000. Another is the use of classifying noun phrases after existential there and it: 6a. There is a book on the table. b. There are [] holes in your sweater c. There is [] paint in your hair. d. Its a boy. e. Its [] sugar. This category subsumes the whole notion of generic noun phrases, that is, abstract representatives of a class rather than actual representatives, because an abstract representative is a classification. Proper nouns. One thing that the classification/identification dichotomy cannot simplify or explain is the use of the and with proper nouns. It seems entirely arbitrary, for example, that rivers require the (the Amazon, the Mississippi) whereas parks require (Yosemite Park, Yellowstone National Park). For this reason Huebner (1983) and others count articles with proper nouns as an entirely separate class. Those proper nouns that take , however, can be both classified and identified, as in the following examples: 7a. Theres a Mr. Smith to see you, sir. b. This was not the London I knew. One observation concerning proper nouns, although it does not fit the generalization that the identifies and classifies may, however, be useful to ESL/EFL teachers. The names of political divisions (China, California, Chicago) typically take whereas the titles of political divisions (the Peoples Republic of China, the state of California, the city of Chicago) take the. Similarly, the names of individual people, mountains, and islands (John Smith, Mt. Everest, Wake Island) take whereas the names of families, mountain ranges, or groups of islands (the Smiths, the Himalayas, the Hawaiian Islands) take the. 474 TESOL QUARTERLY

Idiomatic phrases. Another aspect that the classification/identification dichotomy cannot really explain is article use in idiomatic phrases. The use of with formal names of diseases (typhoid, cancer, meningitis) as opposed to the with less formal names (the flu, the bends, the plague), the fact that we can use with half (half a loaf) but must use a with other fractions (an eighth of a loaf), the use of with few and little to indicate a negative context (few people [=not many] remember him) in contrast to the use of a with the same words to signify a positive or neutral context (a few people [=a small group] came to the hospital) all these, not to mention the numerous set phrases (e.g., hand in hand, go by the board, at sea, all in a dither), are the arbitrary phrases that characterize idiomatic usage. Idiomatic usage remains, for the student, in the realm of things which must be learned and memorized and for which there is rarely a productive rule. There are some cases, however, where the difference between the adverbial and the nominal usage of an idiomatic phrase reflects the classification/ identification dichotomy (e.g., with in case of, in fact, and at last) although the phrase only retains its true idiomatic sense as an adverbial: Adverbial Usage (classified) In case of fire, break the glass. Nominal Usage (identified) In the case of fire, the insurance wont cover it.

In fact the earth is a minor planet. He lied in the fact that he didnt tell the whole truth. The plane arrived at last. At the last, he admitted his greatest crime.

Student Errors with Classification and Identification The most important justification of a pedagogical methodology is that it truly meets a student need. Do students actually make errors with classification and identification? The following are examples of confusion over the dichotomy from student writers whose first languages do not contain an article system: 1. Used classification (a); required identification (the) The hull is a lower part of the ship. Net sales represents a total amount of activity of a merchandising firm. The computer is a control unit of the robot. TEACHING THE ENGLISH ARTICLES 475

2. Used classification(); required identification (the) An income statement . . . shows relationship between two important parts of the firms actiuity. The manager can decide on a profitable plan for next period based on the income statement. 3. Used identification (the, possessive determiner); required classification (a) The worst of forecasters occasionally produce the very good forecast. How does one open oneself to Zen and get the clear mind? It [the end affecter] functions like our human hand. 4. Used identification (the); required classification () 1 think this exercise needs the discussion because each student might have a different answer. The line has the variations such as its length, bending, and thickness. Selecting the familiar topics for the students is important for improving their motivatwn. Such errors will no doubt be familiar to ESL/EFL composition teachers. They show that students do indeed commonly make errors in deciding whether to classify or to identify a noun phrase. CONCLUSION Many examples of article usage have been discussed that can be understood in terms of a binary classification/identification dichotomy. The greatest advantage of the dichotomy is that it provides a framework in which a/ has one clear role and the another. Another advantage is that there is no need to present the generic/specific distinction. And a third is that the notion of intentional vagueness takes on a more principled application. The weakness of the dichotomy is that proper nouns and idiomatic phrases still need to be covered separately. However, even with these, some principles of classification/identification apply. Little has been said in this discussion regarding articles and discourse. Rutherford (1987) summarizes the need to recognize article usage as a discourse phenomenon with its own binary constraints: given and theme require the whereas new and rheme require a/ (p. 77). The contrast is shown in the following examples from Rutherford: 8a. On stage appeared a man and a child. The child [given/theme] sang a song. b. Last on the programme were a song and a piano piece. The song was sung by a child [new/rheme]. (p. 167) 476 TESOL QUARTERLY

In the classification/identification dichotomy, if a noun phrase is given (i.e., thematic), it is identified; if it is new (i.e., rhematic), it is classified. Discourse considerations clearly play a decisive role in article selection in first and subsequent mention environments, including some regional/local aspects of shared knowledge and the subsequent mention aspects of postmodification. However, in selecting the article with ranking adjectives (e.g., the tallest mountain), with world shared knowledge (e.g., the sun), with descriptive versus partitive of-phrases (e.g., the diameter of a circle, a pound of onions), with intentional vagueness (e.g., replication of) cells), and with proper nouns and idiomatic phrases, there is no need to go beyond sentential boundaries unless first/subsequent mention is a factor (e. g., I loved London, but this was not t h e London I knew). Thus, discourse is an important factor but not the only one required for article choice. The classification/identification dichotomy, on the other hand, can be applied whether or not discourse is the controlling factor. In conclusion, it has been my purpose not to undermine the careful theoretical linguistic descriptions that have refined our understanding of the article system but rather to provide an understandable pedagogical tool by means of which nonnative speakers of English might better hope to grasp this elusive aspect of English grammar.

THE AUTHOR
Peter Master is an Assistant Professor in the Department of Linguistics at California State University, Fresno. He is the author of Science, Medicine, and Technology: English Grammar and Technical Writing (Prentice Hall, 1986) and is interested in the acquisition and teaching of the English article system.

REFERENCES
Bolinger, D. (1977). Meaning and form. London: Longman. Brown, R. (1973). A first language. Cambridge, MA: Harvard University Press, Burton-Roberts, N. (1976). On the generic indefinite article. Language, 52 (2), 427-448. Celce-Murcia, M., & Larsen-Freeman, D. (1983). The grammar book. Cambridge, MA: Newbury House.

TEACHING THE ENGLISH ARTICLES

477

Chafe, W. (1969). English noun inflection and related matters from a generative semantic point of view (POLA Report No. 2-6). Berkeley: University of California. Christopherson, P. (1939). The articles: A study of their theory and use in English. Copenhagen: Munksgaard. Grannis, O. (1972). The definite article conspiracy in English. Language Learning, 22 (2), 275-289. Huckin, T., & Olsen, L. (1983). English for science and technology. N e w York: McGraw-Hill. Huebner, T. (1983). A longitudinal analysis of the acquisition of English. Ann Arbor, MI: Karoma. Kruisinga, E. (1932). A handbook of present day English (Vol. 2). Groningen, The Netherlands: Noordhoff. Lamotte, J., Pearson-Joseph, D., & Zupko, K. (1982). A cross-linguistic study of the relationship between negation stages and the acquisition of noun phrase morphology. Unpublished manuscript, University of Pennsylvania, Philadelphia. Lost signals of marriage. (1989, January 9). Newsweek, p. 3. Master, P. (1983, March). Teaching the art of the article. Paper presented at the 17th Annual TESOL Convention, Toronto, Canada. Master, P. (1986a). Science, medicine, and technology: English grammar and technical writing. Englewood Cliffs, NJ: Prentice Hall. Master, P. (1986b). Teaching the English article system to foreign technical writing students. The Technical Writing Teacher, 13 (3), 203-210. Master, P. (1986c). Measuring the effect of systematic instruction in the English article system. Unpublished manuscript, University of California, Los Angeles. Master, P. (1987). Generic the in Scientific American, ESP Journal, 6 (3), 165-186. Master, P. (1988a). Teaching the English article system (Part 1), English Teaching Forum, 26 (2), 2-7. Master, P. (1988b). Teaching the English article system (Part 2), English Teaching Forum, 26 (3), 18-25. McEldowney, P. L. (1977). A teaching grammar of the English article system. International Review of Applied Linguistics, 15 (2), 95-112. Newsweek. (1989, January 9). Pica, T. (1983). The article in American English: What the textbooks dont tell us. In N. Wolfson & E. Judd (Eds.). Sociolinguistics and language acquisition (pp. 222-233). Rowley, MA: Newbury House. Rutherford, W. E. (1987). Second language grammar: Learning and teaching. London: Longman. Stern, H. H. (1983). Fundamental concepts of language teaching. Oxford: Oxford University Press. Whitman, R. L. (1974). Teaching the article in English. TESOL Quarterly, 8 (3), 253-262.

478

TESOL QUARTERLY

TESOL QUARTERLY, Vol. 24, No. 3, Autumn 1990

Attitudes of Native and Nonnative Speakers Toward Selected Regional Accents of U.S. English
RANDALL L. ALFORD and JUDITH B. STROTHER
Florida Institute of Technology

Although some research has been done on the attitudes of native speakers of English toward various regional varieties of U.S. English, few studies have been done on nonnative speakers reactions toward regional accents. This empirical investigation sought to determine the attitudes of both L1 and L2 listeners toward specific regional accents of U.S. English and to compare and/or contrast those attitudes. The subjects were 97 university students from Florida Institute of Technology, half of whom were L2 listeners (advanced ESL students) and half of whom were L1 listeners. Through the use of a modification of the matched guise technique, the students listened to tapes of the same passage read by a male and female native speaker from each of the following accent groups: (a) southern (South Carolina), (b) northern (New York), and (c) Midwestern (Illinois). Respondents then recorded their attitudes about each of the readers using a Likert scale. The results indicated that the judgments of L2 subjects differed from those of L1 subjects and that L2 subjects were able to perceive differences in regional accents of U.S. English.

Dialects are varieties of a language, usually mutually comprehensible by a particular group of people. Although this seems to be a fairly standard definition, there are debates among authors about precise definitions of and differentiations among such terms as dialect, variety, and accent. Chaika (1982) has included the following as characteristics of dialects: the way words are pronounced, syntax, and word choice between speakers, in addition to differences in timbre, tempo, and paralinguistic features. Wolfson (1989) states: From the point of view of sociolinguistic description, a dialect is best regarded as a regional variant. That is, the dialect or variety of a language used by particular speakers is determined in large part by where 479

they come from. Pronunciation is one of the most obvious differences separating regional dialects, but syntactic and semantic patterns also differ, as do some sociolinguistic rules. (p. 3) When the focus narrows to strictly pronunciation (phonological or phonetic distinctions), the term accent is used (Fromkin & Rodman, 1983; Pealosa, 1981). Pronunciation differences are probably the major factor in U.S. English regional varieties, with vowel differences being the most crucial distinguishing feature. Rather than being recognized as having various pronunciation rules, regional accents are often characterized by popular labels such as drawl, twang, nasal, and flat (Christian & Wolfram, 1979). Although dialectologists have carefully analyzed regional dialects for such features as lexical or phonological variations (e.g., Carver, 1987), and some research has been done to record attitudes of native speakers of English toward selected accents of U.S. English, there is a paucity of research dealing with perceptions of L2 speakers toward various regional accents of U.S. English. However, if the media is any indication, popular stereotypes abound, available to native and nonnative speakers. It is not uncommon to find references to various regional dialectal groups in the popular press, especially for humorous, condescending, or derogatory purposes. In an article on southern stereotypes on television, Blount (1988) complained that the stronger a characters Southern accent, the dumber and/or less honest the character. . . . The license to assume that Southerners are morons still holds on TV today (p. 28). Derogatory images are certainly not confined to southerners. For example, New Yorkese is considered by some to be both crude and loud (Hunt, 1986). U.S culture is saturated with caricatures of various ethnic and regional peoples (e.g., Lil Abner, Snuffy Smith, the Dukes of Hazzard, Roseanne, Archie Bunker, and the Honeymooners). These generalized impressions become stereotypes of the group they are purported to represent and in many cases such stereotypes become part of ones cultural background ones frame of reference. According to Gallois and Callan (1981), studies of how people form impressions have consistently shown the readiness of individuals to use language as a cue to classify others into groups. When people know little about an individual, they tend to attribute to that person various traits that they associate with the group(s) to which they assume the person belongs (Tajfel & Turner, 1979). In such situations, virtually any cue to group membership may serve as the basis for ascribing a stereotype. Therefore, a key part of stereotype formation is the value judgment a person makes about different languages or dialects (Sledd, 1969). This is clearly 480 TESOL QUARTERLY

demonstrated in Goldens study (1964, cited in Sledd, 1969) showing that southern speech elicited negative reactions among employers in Detroit. Stereotypes may sometimes be formed by individuals as a result of direct experience with members of the stereotyped groups. For the most part, however, such impressions are learned by word of mouth or from books and films (Karlins, Coffman, & Walters, 1969). The mass media are probably the most common source for international students, who, even before coming to the United States, may have formed their entire impressions of the stereotypical American from movies, television, and the press. These impressions may not include perceptions of regional variation in speech patterns; no research exists on the question of whether or not exposure to movies and television results in nonnative listeners recognizing differences in varieties of English. Such research might ask the following types of questions: (a) Are nonnative listeners able to perceive the phonological variations in speech by speakers of different varieties of U.S. English? (b) If they do detect differences, do they attach value judgments to those differences? and (c) What factors enter into these value judgments? Some studies have dealt with how L1 speakers perceive groups who speak different varieties of English. For example, Labov (1969), who studied black English vernacular in New York, confirms that many features of pronunciation, grammar, and lexicon are closely associated with black speakersso closely as to identify the great majority of black people in the northern cities by their speech alone. He goes on to point out that while many white northerners, particularly those living in close proximity to black communities, share some of these speech characteristics and some black northerners have none, or almost none, of these features, we are dealing with a stereotype that provides correct identification in the great majority of cases, and therefore with a firm base in social reality. Such stereotypes are the social basis of language perception; this is merely one of many cases where listeners generalize from the variable data to categorical perception in absolute terms. (p. 242) Labov points out that a speaker who uses a stigmatized form 20%-30% of the time will be assumed to be using this form all the time. Labov played tapes with sizable extracts from the speech of 14 individuals. He asked his subjects to identify the family backgrounds of each and found that no one even came close to a correct identification of black and white speakers. Labov concluded that this result does not contradict the statement that there exists a socially based black speech pattern; it supports everything that I have said on ATTITUDES TOWARD U.S. REGIONAL ACCENTS 481

this point. The voices heard on the test are the exceptional cases: blacks raised without any black friends in solidly white areas; whites raised in areas dominated by black cultural values; white southerners raised in predominantly black areas. . . . The speech of these individuals does not identify them as black or white because they do not use the speech patterns that are characteristically black or white for northern listeners. The identifications made by these listeners, often in violation of actual ethnic membership categories, show that they respond to black speech patterns as a social reality. (p. 243) In a study (Alford & Strother, in press) of southerners opinions of selected regional accents, southerners reactions were found to be highly sensitive to differences between northern and southern speakers and between northern and Midwestern speakers; however, they did not register significant differences between southern and Midwestern speakers. This might suggest that southerners react more positively to a Midwestern accent because they perceive it as being more standard, more acceptable, and more similar to their own. Since southerners rate northerners as significantly different from both themselves and from midwesterners, it could be assumed that this judgment is based on strong stereotypes. Gallois and Callan (1981) examined the reactions of native-born Australian subjects to Australian, British, and some nonnative accents using sex of subject, nationality of speaker, and sex of speaker as independent variables. They found a significant main effect for sex of speaker (p < .01), indicating that males were perceived more negatively than females, and a significant interaction (p < .01) between nationality and sex of speaker. For these subjects, for example, Italian male voices were rated more negatively than any other group; Italian females as positive as any other group and more positively than speakers with native Australian accents, and male British speakers were rated quite favorably. This is in direct contrast to the findings of the current study in which males were almost always rated more positively than females from the same dialectal regions. Anisfeld, Bogo, and Lambert (1962) found that when the same speakers used Jewish-accented English, they were rated much less positively on personality characteristics and were labeled immigrants on the basis of accent alone. Tucker and Lambert (1969) used three groups of college students as subjects (one northern white, one southern white, and one southern black). These subjects used adjective checklists to evaluate recorded readings by six U.S. English dialect groups: network (the speech of television newscasters on the major networks), educated white southern, 482 TESOL QUARTERLY

educated black southern, Mississippi peer, Howard University, and New York alumni. They found that both northern white and southern black judges rated the network speakers most favorably and the educated black southerners next. The network speaker, followed by the educated white southerner, received the most favorable ratings by the southern white subjects. Both groups of white subjects rated the Mississippi peer least favorably while the black subjects ranked the educated white southerner the least favorably. In a study on bidialectal differences between French-speaking and English-speaking Canadians speaking English, Lambert (1967) made the following generalization: A technique [matched guise] has been developed that rather effectively calls out the stereotyped impressions that members of one ethnic-linguistic group hold of another contrasting group. The type and strength of impression depends on characteristics of the speakerstheir sex, age, the dialect they use, and very likely the social class background as this is revealed in speech style. The impression also seems to depend on characteristics of the audience of judgestheir age, sex, socio-economic background, their bilinguality and their own speech style. (p. 100) While these studies have focused on L1 subjects reactions to accented speech, relatively little has been done to discover how L2 learners react to various U.S. English speech varieties. Eisenstein and Verdi (1985) used three varieties of Englishstandard, New Yorkese (New York nonstandard English), and black Englishto investigate the ability of L2 working class subjects to discriminate among varieties. They found that learners are able to discriminate among the different varieties in the early stages of language acquisition, but that the attitudes and stereotypes for these speakers did not develop until learner proficiency in English increased (Wolfson, 1989). Few studies have measured L2 responses to regional accents in standard U.S. speech. A study by Strother and Alford (1988) examined the relationship between the quality of an L2 speakers pronunciation and his/her ability to perceive differences in L1 speakers accents. No significant correlation was found between scores on a pronunciation test and the ability to detect differences in accent, a finding that may be explained in part by the fact that the L2 subjects did rate the various accents as significantly different. The purpose of the study reported in this article was to compare the reactions of L1 and L2 subjects to selected standard U.S. English accents. ATTITUDES TOWARD U.S. REGIONAL ACCENTS 483

METHOD Subjects The subjects, all of whom were students at Florida Institute of Technology, included 31 L1 and 66 L2 speakers of English. All students were tested in their linguistics, foreign language, or ESL classes. Of the L1 group, 16 were from the North, 9 from the South, and 6 from the Midwest. The average age of the group was 23, and 61% was male. The L2 subjects came from 24 different countries and represented 15 languages; Arabic (36%) and Chinese (27%) were the largest groups. The average age of the predominantly male (84%) L2 group was 23.2 years; L2 subjects had spent an average of 6.37 months in the United States at the time of testing. The L1 subjects were undergraduates, while the L2 subjects included both undergraduate and graduate students. Materials and Procedure The matched guise technique, which was developed at McGill University by Lambert, Hodgeson, Gardner, and Fillenbaum (1960), is a subjective reaction test used to reveal how people feel about characteristics of others based solely on tape-recorded speech of individuals who are bilingual or bidialectal (Anisfeld, Bogo, & Lambert, 1962; Webster & Kramer, 1968). Subjects indicate twice (once to each guise) how they react to each trait for each speaker by marking a scale divided into any odd number of segments. Numbers are assigned to each fine distinction on a continuum and averages of each trait are calculated. However, a modification of the matched guise techniquewhich has also been used by Anisfeld, Bogo, and Lambert (1962); Markel, Eisler, and Reese (1967); Tucker and Lambert (1969); Carranza and Ryan (1975); Ryan and Carranza (1975); Williams, Hewett, Miller, Naremore, and Whitehead (1976)was employed in this study. This modification uses several speakers from each accent group, all of whom speak with their normal accents. This technique utilizes natural rather than counterfeit accents, which may only represent actual stereotypes of the speakers. This also prevents speakers from varying their voice quality and style in an attempt to distinguish among the various accents. All subjects listened to a taped text, which dealt with the culturally neutral topic of what to do in case of an earthquake (Morley, 1979), read by a male and a female native English speaker from each of the following regional accent groups: North (New York), South (South Carolina), and Midwest (Illinois). These 484 TESOL QUARTERLY

middle class, white speakers were college educated and had been screened to control for variations in style, voice quality, and age. Therefore, syntax, word choice, and voice quality variables were controlled for, leaving accentpronunciation and intonationas the variables under consideration in this study. The recording was produced in a professional studio to control volume, clarity, and overall sound quality. The same text was read twice by each speaker in random order, taking from 1.25 to 1.50 minutes to read. This was done to ensure that subjects were consistent when rating a particular speaker both times they heard that speaker. Selection in the order of speakers was randomized by having them draw numbers (1-12) from a hat. A bipolar rating scale was constructed by using adjectives which could be clearly understood by L2 as well as L1 raters following Tucker and Lamberts (1969) suggestion that to be most useful, the rating scales provided listeners for evaluating speakers should be developed specifically for the sample of subjects to be examined (p. 404). The following 24 positive and negative traits were paired: very intelligent/not very intelligent good family training/poor family training; well educated/poorly educated; ambitious/lazy; self-confident/not self-confident; professional/nonprofessional; trustworthy/untrustworthy; sincere/insincere; friendly/unfriendly; patient/impatient; gentle/harsh; and extrovert/introvert. After listening to each speaker, the subjects were asked to evaluate personality characteristics of that speaker, using speech style and voice characteristics as cues, by marking their responses on a 7point Likert scale. Each subjects reactions to each speaker were recorded in the form of a numerical index for each region. This index was obtained by summing the ratings for each trait for each speaker. In the ranking, a score of 1 was the most negative, and 7 was the most positive. It should be noted that a rating of 7 for each of the characteristics on the Likert scale may not represent the most positive evaluation for all subjects. Whereas most L1 subjects would value such traits as extroversion or ambition, this may not be the case for all L2 subjects. However, since differences are being examined, the data are still valuable. LIMITATIONS OF THE STUDY The results of this study should not be overgeneralized because (a) there was a relatively small sample size, especially for L1 subjects, and (b) both L1 and L2 subjects had geographically diverse backgrounds. ATTITUDES TOWARD U.S. REGIONAL ACCENTS 485

While a modification of a true matched guise avoids a feigned accent and while every effort was made to control for style, voice quality, age, and educational background, every voice has its own personality cue value (Webster & Kramer, 1968, p. 239) in addition to the words or accents being used. RESULTS AND DISCUSSION To determine subjects reactions to the six speakers, several statistical analyses were completed. To test for intrarater reliability, a two-tailed Pearson Product-Moment Correlation was calculated (averaged using the Fisher z transformation) to determine the degree of correlation in the way subjects rated each speaker the two times that the speaker read (r= .455). The hypothesis that differences exist between reactions of L1 and L2 subjects toward specific U.S. English accentsnorthern male, northern female, southern male, southern female, Midwestern male, and Midwestern femalewas tested using an analysis of variance with repeated measures. The hypothesis was supported (p< .0001). There was a significant difference (p< .0001) among the regional accents as perceived by all subjects. There was also a strong interaction between the native-speaker status (Ll vs. L2) of the subjects and their reactions to the accents (p < .0016). The data confirm the hypothesis that L2 subjects, as well as L1 subjects, are able to perceive differences among the regional accents and that the perceptions of L2 subjects differ from the perceptions of L1 subjects. One might speculate that U.S. English speakers would provide uniform responses on a Likert chart for regional dialect groups without even listening to a tape. This would be because of strong cultural biases based on both personal experience and impressions from stereotypes given in the media for each group of people represented by a regional accent. Katz and Braly (1933) conducted a pioneer study of verbal stereotypes, looking at the five primary traits used by 100 Princeton undergraduates to characterize 10 different racial and national groups. There was an impressively high level of agreement of these verbal descriptions, which yielded a distinctive set of population labels for each of the 10 groups. In reviewing this study, Karlins, Coffman, and Walters (1969) noted that since most students had no contact with members of the stereotyped groups, it was obvious that they had simply absorbed the prevalent images of their day and culture (p. 1). The results of the Katz and Braly study were confirmed at the University of California, Los Angeles, by Centers 486 TESOL QUARTERLY

(1951) and by Reed (1971) with white southern college students as subjects. International students, for the most part, do not have the same cultural framework as native students. It is important to note that the L2 subjects had only been in the United States for slightly more than 6 months, barely enough time to form surface-level value judgments about the area in which they were living (in this case, Florida), much less to form complex opinions about the individual characteristics of and the interrelationships among various parts of the country. Their reactions to the various speakers in our study are all the more interesting in this light. It can be assumed that, aside from television and movies, a large part of their value judgments toward the various dialects was different from the cultural biases native speakers have. However, some of the subjects own cultures no doubt are influencing factors. Perhaps the strongest evidence of this is seen in the differences in rating male/female speakers, independent of accent, for some characteristics. In addition to regional accent, voice qualities of maleness or femaleness may have an important effect on listener perceptions. OLearys (1977) literature review on the stereotyping of male and female personality characteristics shows that, as in Gallois and Callens (1981) study described earlier, male and female subjects did not differ in their impressions of male and female speakers. Men and women tend to share sex-role stereotypes, indicating that both sexes share expectations about the characteristics of a typical male and typical female. Key (1975) has summarized a number of studies which link these stereotypes to the voices and speech styles of males and females. On the other hand, as Gallois and Callan (1981) have pointed out, a speakers sex has not been considered in studies dealing with language and accent-based stereotypes. There is certainly a strong possibility that these two variablessex and accent-based stereotypesinteract in determining what perceptions people have of accented language. While it is beyond the scope of this study to analyze in detail the rationale behind each L1 and L2 raters decisions about male and female speakers, it is an area of study that should be pursued. Due to the variety of cultures within the L2 group (representing 24 counties), an analysis of the relevant cultural factors is not feasible for this study. Table 1, a descriptive presentation of L1 subjects ratings of characteristics of. regional accents, indicates some interesting patterns. In each category, the highest rating is italicized, with a rating of 7 being highest and therefore the most positive. Most ATTITUDES TOWARD U.S. REGIONAL ACCENTS 487

TABLE 1 L1 Subjects Mean Ratings of Regional Accents (n =31)

people in the U.S. consider network standard to be the most acceptable accent since it is considered both regionally and socially neutral (Marckwardt, 1980). Many native speakers of U.S. English consider the Midwestern accent to be closest to this network model, a finding supported by a recent survey conducted by Strother and Alford (1989). However, the results of the current study indicate some differences. The summary figures at the bottom of the table show that this group of 31 native speakers accorded the southern male the highest overall rating (5.3). The second highest ranking was assigned to the Midwestern female (5.1), with the Midwestern male and the southern female tying for third place at 5.0. It is especially surprising that, with 52% of the L1 subjects coming from the North, they ranked both the northern male and the female quite low (4.5 and 3.8). In individual characteristic ratings, the Midwestern male received more high rankings than the other speakers. The Midwestern female received the highest ranking for good family training and for patience. It is also worth noting that together the Midwestern male and female were highest on 8 of the 12 categories. The other 4 characteristics on which they were rated lowertrustworthiness, sincerity, friendliness, and gentlenessare of an interpersonal nature. In contrast, the southern male was rated highest in 488 TESOL QUARTERLY

trustworthiness and sincerity, and the southern female in friendliness and gentleness. Table 2, a descriptive presentation of L2 subjects ratings of characteristics of regional accents, shows that this group of 66 L2 subjects accorded both the midwestern male and the southern male an equally high rating (5.5). In all characteristics except one, either the southern male or the midwestem male received the highest rating or tied for it. The one exception was the characteristic of friendliness in which the southern female was rated highest.
TABLE 2 L2 Subjects Mean Ratings of Regional Accents (n= 66)

Characteristics

Speakers

Just as in the L1 data, both the southern and Midwestern females received higher overall ratings than either the northern male or the northern female. The lowest ratings are also noteworthy. Again the northern female received the lowest overall rating. Several individual ratings were unusually low. For example, the northern male got a 3.5 rating in patience. Both the northern and southern females (3.7 and 3.6) were rated very low in extroversion. Although determining the reason for this was not a part of this research, a study could be developed to uncover the rationale behind these judgments. The bar graphs clearly show differences in the summary ratings by L1 and L2 respondents. Figure 1 presents L1 subjects ratings of ATTITUDES TOWARD U.S. REGIONAL ACCENTS 489

regional accents and Figure 2 shows the ratings by L2 subjects. As an example of a single characteristic, the ratings for professional/ nonprofessional have been graphed in Figure 3 for L1 subjects and in Figure 4 for L2 subjects.
FIGURE 1 L1 Subjects Mean Ratings of Regional Accents

North Male Female

South

Midwest

FIGURE 2 L2 Subjects Mean Ratings of Regional Accents

490

TESOL QUARTERLY

FIGURE 3 L1 Subjects Mean Ratings of a Single Characteristic Professional/Nonprofessional

7

7-

When L1 subjects react to U.S. English accents, we assume that various cultural and dialectal stereotypes, including male/female differences, are part of their frame of reference. Future research ATTITUDES TOWARD U.S. REGIONAL ACCENTS 491

needs to ascertain the reasons for the development of L2 subjects strong opinions of U.S. accents. It can be assumed that their stereotypes of male and female characteristics in their countries are deeply ingrained in their cultures and may be transferred to these ratings. Especially within some of the individual characteristics, strong male/female differences become more pronounced. CONCLUSIONS AND IMPLICATIONS The statistical analyses show that L2 speakers of English are able to detect regional accent differences in U.S. English. The subjective ratings of characteristics of each of the regional groups also show that L2 speakers are able to rate their perceptions of a speakers favorableness (characteristic by characteristic) based primarily on pronunciation variations separate from native speakers regional cultural biases. The exception to this may be differences in rating male versus female speakers where the subjects own culturesnot their perception of a region of the United Statesmay dictate their ratings. There are several important implications that can be drawn from the conclusions. It would seem beneficial for students learning English to know as much as possible about the distinctions that exist within the language they are learning since information regarding language diversity is an excellent introduction to the social and cultural background of that group (Wolfson, 1989). Wolfram and Christian (1989) suggest that the key to attitudinal changes lies in developing respect for the diverse varieties of English (p. 22). Knowing how people react to language features on the basis of what is considered most crucial or stigmatizing to students gives teachers an insight that they can apply in the classroom situation (Shuy, 1969). The type of sociolinguistic research represented by this study provides a valuable tool in making all of us aware of the stereotyped attitudes we have toward other groups. This knowledge should, therefore, be used to counteract such beliefs for both the native and nonnative alike. RECOMMENDATIONS FOR FURTHER RESEARCH In a current project, this study is being replicated in the specific regions represented (North, South, Midwest). The same materials and empirical procedures are being used with both L1 and L2 subjects in New York, South Carolina, and Illinois, the regions represented by the accents of the speakers on the tape. The results of this comprehensive collection of data will be compared to the 492 TESOL QUARTERLY

results of the current study to determine (a) whether L1 subjects react more or less favorably to their own accent than do subjects outside the region, and (b) whether the region in which L2 subjects are living affects their ratings of the various regional accents. A number of valuable research studies would add to this attempt to measure L2 speakers reactions to variations in regional accents of U.S. English. In addition to the suggestions for further research discussed throughout this paper, this study should be replicated with other speakers from the same regions in order to confirm these findings. With confirmation, the results would be more accurately generalizable.

THE AUTHORS
Randall L. Alford, Assistant Professor of Applied Linguistics and Foreign Language, is the Chair of the Florida Institute of Technology Division of Languages and Linguistics and Director of its Language Institute. He is a former President of Gulf Area TESOL. Judith B. Strother, Assistant Professor of Applied Linguistics at Florida Institute of Technology, does research in English for special purposes/English for science and technology. She has authored Kaleidoscope (Holt, Rinehart and Winston, 1988) and Syntax in the ST Register: Effect on Writers Choices and Readers Comprehension (Wibro, 1990), and contributed to Research in Reading in English as a Second Language (TESOL, 1988).

REFERENCES
Anisfeld, M., Bogo, N., Lambert, W. E. (1962). Evaluational reaction to accented English speech. Journal of Abnormal and Social Psychology, 65, 223-231. Alford, R. L., & Strother, J. B. (in press). A southern opinion of regional dialects. Perspectives on the American South. New York: Gordon and Breach. Blount, R. (1988, July 2). My, how they kiss and talk. T.V. Guide, pp. 26-29. Carranza, M. A., & Ryan, E. B. (1975). Evaluative reactions of bilingual Anglo and Mexican American adolescents towards speakers of English and Spanish. International Journal of the Sociology of Language, 6, 8-104. Carver, C. M. (1987). American regional dialects: A word geography. Ann Arbor: University of Michigan Press.

ATTITUDES TOWARD U.S. REGIONAL ACCENTS

493

Centers, R. (1951). An effective classroom demonstration of stereotypes. Journal of Social Psychology, 34, 41-46. Chaika, E. (1982). Language: The social mirror. Rowley, MA: Newbury House. Christian, D., & Wolfram, W. (1979). Dialects and educational equity Exploring dialects. Arlington, VA: Center for Applied Linguistics. Eisenstein, M., & Verdi, G. (1985). The intelligibility of social dialects for working-class adult learners of English. Language Learning, 35 (2), 287-298. Fromkin, V., & Rodman, R. (1983). An introduction to language (3rd ed.). New York: Holt, Rinehart & Winston. Gallois, C., & CalIan, V. J. (1981). Personality impressions edited by accented English speech. Journal of Cross-Cultural Psychology, 12 (3), 347-359. Hunt, G. W. (1986). On many things. America, 155 (16),2. Karlins, M., Coffman, T. L., & Walters, G. (1969). On the fading of social stereotypes: Studies in three generations of college students. Journal of Personality and Social Psychology, 39, 1-16. Katz, D., & Braly, K. (1933). Racial stereotypes of one hundred college students. Journal of Abnormal and Social Psychology, 28, 280-290. Key, M. R. (1975). Male/female language. Mehuen, NJ: Scarecrow Press. Labov, W. (1969). The logic of nonstandard English. (Georgetown Monographs in Languages and Linguistics No. 22). Washington, DC: Georgetown University, Center for Applied Linguistics. Lambert, W. E. (1967). A social psychology of bilingualism. Journal of Social Issues, 23 (2), 91-109. Lambert, W. E., Hodgeson, R. C., Gardner, R. C., & Fillenbaum, S. (1960). Evaluational reactions to spoken languages. Journal of Abnormal and Social Psychology, 60 (1), 44-51. Marckwardt, A. H. (1980). American English (2nd ed.). NY: Oxford University Press. Markel, N. N., Eisler, R. M., & Reese, H. W. (1967). Judging personality from dialect. Journal of Verbal Learning and Verbal Behavior, 6, 33-35. Morley, J. (1979). Improving spoken English. Ann Arbor: University of Michigan Press. OLeary, V. (1977). Toward understanding women. Monterey, CA: Brooks/Cole. Pealosa, F. (1981). Introduction to the sociology of language. Rowley, MA: Newbury House. Reed, J. S. (1971). The enduring south: Subcultural persistence in mass society. Lexington, MA: D.C. Heath Lexington Books. Ryan, E. B., & Carranza, M. A. (1975). Evaluative reactions of adolescents toward speakers of standard English and Mexican accented English. Journal of Personality and Social Psychology, 31, 855-863. Shuy, R. (1969). The relevance of sociolinguistics for language teaching. TESOL Quarterly, 3 (l), 13-22. 494 TESOL QUARTERLY

Sledd, J. (1969). Bi-dialectalism: The linguistics of white supremacy. English Journal, 58 (9), 1307-1329. Strother, J. B., & Alford, R. L. (1988, March). The relationship between L2 speakers pronunciation and their ability to detect variations in dialects of American English. Paper presented at the 22nd Annual TESOL Convention, Chicago. Strother, J. B., & Alford, R. L. (1989). [Survey of attitudes of regional dialects]. Unpublished raw data, Florida Institute of Technology, Division of Languages and Linguistics, Melbourne. Tajfel, H., & Turner, J. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33-48). Monterey, CA: Brooks/Cole. Tucker, G. R., & Lambert, W. E. (1969). White and Negro listeners reactions to various American-English dialects. Social Forces, 47, 463-468. Webster, W. G., & Kramer, E. (1968). Attitudes and evaluational reactions to accented English speech. Journal of Social Psychology, 75, 231-240. Williams, F., Hewett, N., Miller, M., Naremore, R. C., & Whitehead, J. L. (1976). Explorations of the linguistic attitudes of teachers. Rowley, MA: Newbury House. Wolfson, N. (1989). Perspectives: Sociolinguistics and TESOL. New York: Newbury House/Harper& Row. Wolfram, W., & Christian, D. (1989). Dialects and education: Issues and answers. Englewood Cliffs, NJ: Prentice Hall Regents.

ATTITUDES TOWARD U.S. REGIONAL ACCENTS

495

REVIEWS
The TESOL Quarterly welcomes evaluative reviews of publications relevant to TESOL professionals. In addition to textbooks and reference materials, these include computer and video software, testing instruments, and other forms of nonprint materials. Edited by HEIDI RIGGENBACH
University of Washington

Recent Publications in Sociolinguistics

Sociolinguistics and Second Language Acquisition
Dennis Preston. Oxford: Basil Blackwell, 1989. Pp. xv + 326.

Perspectives: Sociolinguistics and TESOL Nessa Wolfson. Cambridge: Newbury House, 1989. Pp. xvi + 319. Second language acquisition (SLA) research is now proceeding in two general directions: First is the investigation of learner internal issues, that is, the representation of the internal grammars of the L1 and L2; second is the investigation of external variables. Dennis Prestons Sociolinguisties and Second Language Acquisition and Nessa Wolfsons Perspectives: Sociolinguistics and TESOL make significant contributions to the field in the second area. Prestons book, in particular, provides a wealth of background to both SLA and sociolinguistics and successfully shows how the two fields can contribute to each other. However, the two volumes differ considerably in what they take to be the important contributions of sociolinguistic methods and findings to SLA and TESOL. Prestons book begins with a brief review of sociolinguistic research that demonstrates the importance of a variety of interfactional factors and learner characteristics. The section on individual characteristics provides a detailed description of participants; the discussion of interfactional factors expands Hymess (1972), SPEAKING taxonomy. Preston then looks at research in which these variables have been investigated in both first and second language sociolinguistics. Most of the book comprises reviews of major variationist theories of SLA. The discussion begins with Krashens Monitor Model (1981), in which variation is attributed entirely to the use of the monitor. Preston points to Krashens conflation of the sociolinguistic concept of attention to form and the psycholinguistic concept of 497

monitoring as a major weakness in this approach to variation. In reviewing Tarones (1983) continuous competence model, Preston criticizes the construct attention to form as too primitive to capture the complexity of the factors that may differ across what Tarone calls tasks (p. 59). Last, Preston reviews Elliss variable competence model which, he maintains, adds some psycholinguistic depth to Tarones capability continuum. This is an especially important section since, for many, Elliss Understanding Second Language Acquisition (1985) is their first exposure to SLA research and perhaps even to sociolinguistic research. Preston does point to serious problems, however, in Elliss presentation of basic concepts in the work of both Bickerton and Labov. In contrast, much of Perspectives: Sociolinguistics and TESOL (Perspectives), focuses on native-speaker behavior as part of the field the author calls microsociolinguistics. The late Nessa Wolfson was a tireless champion of the systematic examination of the rules of speaking across speech communities. In particular, she argued for the use of authentic, spontaneous data as the best way of accurately documenting what constitutes communicative competence in a given speech community. This perspective dominates much of her book. Although the title suggests the same integration of the fields of sociolinguistics and TESOL that is achieved for sociolinguistics and SLA in Prestons book, the topics covered (such as rules of speaking, sociolinguistic methodology, cross-cultural speech act research, and male/female language) demonstrate the dominance of sociolinguistic concerns in this volume. In fact, very little is offered in Perspectives in the way of teaching applications. This, Wolfson maintains, is because we know so little about the speech conventions in our own community and therefore are not yet in a position to provide instruction on such matters to nonnative speakers. For Preston, the goal of sociolinguistics is the investigation of variation. He maintains, in contrast to Wolfsons position, that there is no need to wait for the continued investigation of native-speaker behavior and that the insights from sociolinguistic methodology can be used to inform SLA research in two basic ways. First, as in native-speaker research, it is crucial that analyses of second language learner data take into account a wide range of variables: ascribed and acquired, individual characteristics as well as interfactional factors. These are the kinds of factors that sociolinguistics and ethnographers, Hymes in particular, have suggested are important in describing and analyzing variation in language use. Preston goes beyond this, however, to his second claim that the quantitative/sociolinguistic methodologies that have been used to analyze variation in language use in a single speech community (for 498 TESOL QUARTERLY

example, by Labov) due to their sensitivity to the variability of grammar, are excellent means of capturing synchronic and diachronic aspects of interlanguage systems, no matter what the source of variation [italics added] (p. 198). However, the study of second language variation cannot completely parallel native-speaker sociolinguistic research because there is no speech community. Preston does not emphasize sufficiently the important distinction between first and second language research. While methods of analysis may remain useful in investigating SLA, the assumptions on which they rest are very different. However, Prestons thesis that sociolinguistic methods are appropriate and necessary for SLA data analysis allows him to provide an excellent introduction to Labovian quantitative methods used to study sociolinguistic variation. He examines first how an analysis of all of the factors in his taxonomy might help to explain codeswitching in data from Fishman (1972). He then discusses a study in which quantitative methods were used to determine the factors influencing the use of plural marking by second language learners (Young, 1989). It is particularly important to stress that this second study must be viewed simply as the application of a procedure for data analysis. Although both Young and Preston claim that they are working within the framework set forth originally by Hymes (1972), this is misleading. Hymes never intended that his SPEAKING framework be used for componential analysis; rather, that the importance of each of the components be explored and understood within the context of the speech community. Perspectives explores a number of other sociolinguistic traditions, focusing more on qualitative approaches than does Prestons volume. Wolfson particularly stresses the work of Hymes; her book is more in keeping with the spirit of Hymess SPEAKING taxonomy and his analysis of speech events than is Prestons. Wolfson also briefly explores philosophical and ethnomethodological traditions, as well as Brown and Levinsons (1978) pragmatic work on politeness and face. Much of her book, however, is devoted to Wolfsons concern for how sociolinguistic data are gathered and interpreted. Several chapters are devoted to Wolfsons work, as well as that of others, on rules of speaking and cross-cultural investigations into speech act production and interpretation. Such speech acts include compliments, apologies, greetings, refusals, and requests. It is Wolfsons contention that when a nonnative speaker makes pragmatic errors, such as in the misuse or misinterpretation of these speech acts, native speakers are less forgiving than when the errors are grammatical or lexical. Often such behavior is interpreted as rudeness, obtuseness, and so on. For this reason, REVIEWS 499

Wolfson argues, it is essential that such concerns become fundamental to the field of language teaching. In order to do this, however, native-speaker teachers and researchers need to develop a better understanding of their own behavior. The remainder of Perspectives addresses other topics in sociolinguistics, but their treatment is somewhat cursory. These include language and gender, variation across social classes, multilingualism, and bilingual education. The strength of this volume clearly lies in Wolfsons discussions of research methods and findings in microsociolinguistics and their potential contribution to the field of TESOL. Her death will leave a void in this area of the field that will indeed be difficult to fill. Although these two books clearly take very different approaches to how sociolinguistics, TESOL, and SLA are related, the two volumes are complementary and together would make an excellent introduction to important issues in these fields. While other books may be preferred for broader perspectives on such issues as the history of nonstandard dialects in the United States, immigrant languages, and approaches to multicultural education, these two should be required reading for anyone interested in the social context of SLA and language teaching, which arguably should include every researcher in our field. REFERENCES
Brown, P., & Levinson, S. (1978). Universals in language usage: Politeness phenomena. In E. Goody (Ed.), Questions and politeness: Strategies in social interaction (pp. 56-289). Cambridge: Cambridge University Press. Ellis, R. (1985). Understanding second language acquisition. Oxford: Oxford University Press. Fishman, J. (1972). The sociology of language. Rowley, MA: Newbury House. Hymes, D. (1972). Foundations in sociolinguistics. Philadelphia: University of Pennsylvania Press. Krashen, S. (1981). Second language acquisition and second language learning. Oxford: Pergamon Press. Tarone, E. (1983). On the variability of interlanguage systems. Applied Linguistics 4 (l), 42-63. Young, R. (1989). Ends and means: Methods for the study of interlanguage variation. In S. Gass, C. Madden, D. Preston, & L. Selinker (Eds.), Variation in second language acquisition: Psycholinguistic issues (pp. 63-90). Clevedon, Avon, England: Multilingual Matters. JESSICA WILLIAMS
University of Illinois at Chicago

500

TESOL QUARTERLY

Newbury House TOEFL Preparation Kit: Preparing for the TOEFL Daniel B. Kennedy, Dorry Mann Kenyon, and Steven J. Matthiesen. New York: Newbury House/Harper& Row, 1989. Pp. xi + 262. Newbury House TOEFL Preparation Kit: Preparing for the Test of Written English Liz Hamp-Lyons. New York: Newbury House/Harper & Row, 1989. Pp. vi + 134. Why must I take the TOEFL? Does the TOEFL cost the same in each country? Should I guess if Im not sure about an answer? If I need to guess, what strategy should I use? Are there any general test taking strategies 1 should use? Detailed answers to these and 38 other questions are found in Preparing for the TOEFL, one of two books in the Newbury House TOEFL Preparation Kit. The book contains everything youve always wanted to know about the TOEFL, but never had a chance to ask. Written clearly and simply, it contains careful explanations about every part of the TOEFL. An entrance requirement at 2500 colleges and universities in the U. S., Canada, and other parts of the world, the TOEFL is one of the most widely used language proficiency tests in the world. In order to make maximum use of the available test time, students must be prepared for the kinds of questions that will be asked and need to develop good test-taking strategies. For this reason, the number of TOEFL preparation books on the market has proliferated despite the publication by the Educational Testing Service of the practice tests contained in Understanding TOEFL, published in 1980, and Reading for TOEFL, published in 1987. The publics demand for more preparatory materials indicates the importance of passing the test. The Newbury House TOEFL Preparation Kit consists of two books, one with accompanying materials. Preparing for the TOEFL includes three practice tests with answer sheets; accompanying the book are two tapes for practicing the Listening Comprehension section of the TOEFL, and a pamphlet containing a typescript and answers for the three practice tests in the book. The second book is Preparing for the Test of Written English. (All components can also be purchased separately.) Preparing for the TOEFL begins with a basic description of the REVIEWS 501

TOEFL examination, including sample questions. The next part provides detailed answers to over forty questions commonly asked about the TOEFL program and its policies (p. iv). The next chapters describe the seven major types of questions within the sections on Listening Comprehension, Structure and Written Expression, and Vocabulary and Reading Comprehension. The authors aim to help students recognize the different types of questions to expect so that they will be better able to answer each question correctly. Although other study materials with which I am familiar contain a review of English grammar, none is as specifically tailored to the TOEFL items as is this guide. Each grammar point is defined in simple language and followed by seven sample questions illustrating how the feature is tested on the TOEFL. No extraneous material is introduced to confuse students. The authors provide sensible, practical suggestions concerning the Vocabulary portion of the TOEFL. They have observed that two criteria for the appearance of a word in the Vocabulary section are (a) whether it can be found in a variety of contexts in universitylevel reading, and (b) whether it can easily be given a synonym. Test developers have no choice but to select such words if the test is to be valid. This means the student need not study most technical words, idioms, or phrases; nor need one study types of birds, fish and other animals, foods, clouds, minerals, etc. Since the Vocabulary section of the TOEFL is essentially a test of synonyms, the authors of Preparing for the TOEFL tell students not to waste time reading the whole sentence but to look only at the underlined word and choose its synonym. To illustrate that the context doesnt help, the authors give sentences in which they blank out the key word being tested, but include the four response choices. Any of the four choices would be possible in the sentence. Preparing for the TOEFL deals straightforwardly with elusive issues concerning the content of reading comprehension passages. It is explained that unhappy (p. 1) subjects, such as divorce and war, will not be included on the TOEFL. One should expect materials with women in positive roles and ones describing the contributions of U.S. ethnic minorities, Written texts will be academic in nature, coming from encyclopedias and textbooks but not from scientific or literary journals. However, the authors explain that knowledge of the subject is not very important. Do not worry if you are not familiar with the topic being discussed. You will be tested on your comprehension of the English language, not your general knowledge (p. 2). The TOEFL is geared toward students who will be living and 502 TESOL QUARTERLY

studying in the U.S. or Canada, and thus it contains references to North American events and figures. In their discussion of inference questions in the Listening section, the authors suggest how the reader may prepare for this fact. To answer these questions, you should try to become familiar with the kinds of settings students in North America may find themselves in (p. 42). Two appendices are provided. One contains a list of the TOEFL scores required by 100 U.S. colleges and universities having the largest international student enrollments. Students can use this list to compare their TOEFL scores with the entrance requirements of the institutions they plan to enter. The second appendix provides instructions and score conversion tables for translating raw scores on the practice tests given in the book to scaled scores. After deriving the scaled scores, students can compare their individual scores with the scores of others who have taken the TOEFL by referring to a table of norms for TOEFL test takers. One of the strengths of Preparing for the TOEFL is the care and detail of the explanations. Each question is discussed in terms of linguistics on the microlevel (vocabulary and syntax of the sentence) and the macrolevel (sociolinguistic considerations of the logic of a situation, the tone of the speaker), as well as in terms of test development logic and test-taking strategies. Fine details about the test are also provided: Note again that the numbers in this set of options are listed in ascending order, as always on the TOEFL (p. 53). Most TOEFL preparation kits help students recognize the format of each subtest, how to follow directions, and how to record answers. They familiarize the students with the appearance of the TOEFL. They do not analyze question functions. This is the first preparatory guide to explain the linguistic and sociolinguistic rationale as well as the testing function of each item type. The authors do not mislead the student about the value of their TOEFL preparation materials. They warn students that merely reading study materials will not guarantee a high score. Students are told that if, after they work through the book, their practice test scores do not improve, their problem is a lack of English language proficiency, and they are advised to enroll in an English language course while continuing to review the kit. Should students need further practice, the authors advise them to order the ETS TOEFL Test Kits. The extent to which the writers of Preparing for the TOEFL took pains to produce questions as authentic as possible is impressive. As explained in the introduction, the book is based on an examination of thousands of TOEFL test questions, the TOEFL Test and Score REVIEWS 503

Manual, and the TOEFL Research Reports. Based on this analysis, the authors identified and described subcategories of the seven TOEFL question types, wrote similar questions, and developed three practice tests. What is unique is that these questions underwent pretesting and retesting, just like the TOEFL items themselves. The practice tests were administered to international students studying at universities in the U.S. and elsewhere, and items were statistically analyzed and reworded when necessary. This procedure yielded questions that not only look like TOEFL questions, but also function like TOEFL questions. The final version of the three practice tests was administered with a real TOEFL (p. iii) to students at several U.S. universities; scores on the practice tests and the TOEFL were correlated, yielding a score conversion table. Preparing for the TOEFL is the only TOEFL preparation program I have seen that has been developed in such a professional way. The second book, Preparing for the Test of Written English b y Liz Hamp-Lyons, is for those who wish to take this optional part of the TOEFL. The half-hour test is offered four times a year with the TOEFL, and, in almost all cases, is administered before the TOEFL. The TWE requires one essay, to be written in 30 minutes. There is only one question, no choice of topics. Who should take this test? In the prefix, the author states that the book assumes that [the reader] already [has] an English proficiency level of 450 or higher on the TOEFL (p. iii) and directs the reader to a page in the book with suggestions for determining whether one is ready to take the TWE. Hamp-Lyons defines academic writing as the kind of writing that students and teachers do at universities and colleges. She discusses successful academic writing, gives examples of real essays written by students on topics and questions like those on the official TWE, discusses the strengths and weaknesses of these essays, and justifies the scores given them by TOEFL-trained essay scorers. The book also includes practice writing tasks and a section that teaches students how to score their own practice essays. The book has been designed to enable students to work through it with or without the help of a teacher, in groups or individually. The discussion of the structure and organization of academic writing is very clear, although the instructions would be more suitable for a teacher than for the student. A student who is able to read this book is well on the way to being able to write a composition. (The language proficiency assumed for readers of this book is somewhat higher than for Preparing for the TOEFL.) 504 TESOL QUARTERLY

In her introductory remarks to the student, the author states, Writing is like driving a caryou can only learn by actually doing it, not by reading about how to do it! (p. iii). She implements this approach through the book. Her purpose is to teach specific skills that prepare the student for the TWE or any other academic writing test. In fact, the book contains comparisons with two other academic writing tests, the British Councils English Language Testing Service (ELTS) writing component M2 and the University of Michigans English Language Assessment Battery (MELAB). Although these comparisons are not meant to be exact, their inclusion helps the reader assess his or her writing level on various tests and supports the idea that tests of academic writing are similar and can be valid and reliable. This book also contains a 45-page chapter called Self-Scoring Practice. It explains that the TWE is a criterion-referenced test based on an absolute standard of what good academic writing is. It explains and elaborates on the 6-level scale (6 = competence, 1 = incompetence) developed by ETS for the TWE. For example, ETS describes an essay receiving a score of 6 as one that clearly demonstrates competence on both the rhetorical and syntactic levels, though it may have occasional errors (p. 76). Hamp-Lyons further describes each level in terms of organization, use of supporting detail, coherence, and appropriateness. She gives examples of students handwritten essays with a scorers comments typed in the margins, and essays in which the student has made revisions during the test itself. Thus students using the book are provided many examples with which to compare their own writing. The inclusion of students handwritten essays provides additional authenticity. Each handwritten essay is also found in printed form elsewhere in the book. Including 20 TWE-like prompts, the book provides ample opportunity for practice in writing. Hamp-Lyons describes the structure of typical academic tasks: The student is given the situation and the problem, and must suggest a solution and make an evaluation (SPSE) (p. 8). She then discusses academic organization and invention strategies, such as mind-mapping, a technique for showing pictorially how major ideas are related. The two books are excellent preparatory guides for the TOEFL and the TWE. The authors are sensitive about the dangers of studying for tests. They warn students about the need to develop English language skills and make suggestions on how to do this. Hamp-Lyons focuses on teaching academic writing skills. In addition to the value of these books as guides for the tests, REVIEWS 505

because the authors have carefully identified and implemented rules underlying the development of language testing, the kit would also be useful for those interested in standardized language testing. Teacher training programs may also wish to include this kit in courses on second and foreign language test development.
MARSHA BENSOUSSAN
Haifa University

506

TESOL QUARTERLY

BOOK NOTICES
The TESOL Quarterly welcomes short evaluative reviews of print and nonprint publications relevant to TESOL professionals. Book notices may not exceed 500 words and must contain some discussion of the significance of the work in the context of current theory and practice in TESOL.

Making it Happen: Interaction in the Second Language Classroom. Patricia A. Richard-Amato. New York: Longman, 1988. Pp. xviii + 426.

This is an excellent textbook for a basic graduate class in TESL methods. The text covers the theory and practice of TESL as well as the latest thinking in international language instruction and bilingual education. Although ambitious in scope, the book provides good depth of material. Part I provides a theoretical perspective, including a brief history of second language teaching methodologies and a careful examination of the theoretical work of Krashen and Vygotsky. The underlying philosophical framework of this section is that a second language classroom should be an interfactional one in which communication is emphasized (p. xiii). Part II, Exploring Methods and Activities, could be a separate book. It provides excellent, detailed, practical teaching activities designed to help the new instructor develop a genuinely interfactional classroom. The suggested activities provide for a variety of options, including music, drama, storytelling, and games. Part III, Some Practical Issues, covers such practical topics as classroom management, testing, computers, and textbooks. Each chapter in the first three sections of the book is followed by an annotated list of suggested readings and by questions for reflection and discussion. Part IV, Programs in Action, shows the numerous settings in which ESL instruction can take place. This is one of the most valuable sections of the book. Teachers-in-training often think of only one type or level of ESL instruction without realizing the tremendous variety of opportunities available. For example, although a teacher may plan to work in a collegelevel intensive institute, a more attractive position in an elementary school may arise. In presenting interesting examples of diverse professional options, this book performs a service to the profession. Part V, the last quarter of the book, comprises a series of related supplemental readings by scholars in the field including Noarn Chomsky, Stephen Krashen, Lev Vygotsky, John Oller, and Jim Cummins. These related readings are frequently referred to in the preceding text. Their presentation here invites productive supplemental assignments since the reader can conveniently examine the primary sources on which the text is based. 507

Making it Happen is a comprehensive methods text. It can probably best be utilized in a two-semester sequence although, with carefully selected assignments, it would also be excellent in a one-semester course. DOROTHY S. MESSERSCHMITT
University of San Francisco

The Foreign Teaching Assistants Manual. Patricia Byrd, Janet C. Constantinides, and Martha C. Pennington. New York: Collier Macmillan, 1989. Pp. xii + 193. The Foreign Teaching Assistants Manual is a resource book to be used in international teaching assistant (ITA) training programs. The target audience, those currently teaching or preparing to teach U.S. undergraduates, is assumed to have an advanced level of skills in English. The 25 chapters of this activity-oriented manual are arranged in five parts. Part One, Preparatory Activities, provides questionnaires that guide ITAs in collecting information on U.S. culture and educational styles. Part Two, Background to Teaching, introduces a variety of teaching responsibilities, such as organizing a course, presenting in class, using audiovisual aids, leading a discussion, and record keeping. Part Three, Hearing and Pronouncing American English, deals with pronunciation of individual sounds, stress patterns, fluency, intonation, and voice quality, Part Four, Practice for Teaching, includes explanations, guidelines, and evaluation forms for a variety of presentations to be videotaped in class. Part Five, Observation, provides several open-ended worksheets to guide ITAs observation of undergraduate classrooms. Printed on 8 by 11 punched and perforated pages, this book is designed for integrated rather than sequential use. Each chapter begins with an overview section with a persuasive rationale, chapter objectives, and introductory content. Chapters are subdivided into nearly autonomous sections that typically provide basic information, descriptions of student activities, and accompanying worksheets. Throughout the book, the authors communicate directly with the reader using the second person. The overall tone is respectful of other cultures yet firm in its recommendation that ITAs adjust to American English and U.S. educational styles. While it is acknowledged that U.S. undergraduates are not always culturally sensitive or patient with outsiders (p. ix), the focus is on the ITAs skills and cultural adjustment for success in the U.S. classroom. The segmented design and individualized approach are this manuals dominant characteristics. These allow great flexibility for use with students at different levels, for various curricula, and for courses ranging from a few days to a full semester. Although the introduction states that the whole 508 TESOL QUARTERLY

book could be covered in a one-semester course, selective use of materials and/or allowing two semesters would be a more feasible approach. While suggestions for selecting and coordinating the materials are given in a onepage appendix, teachers using this book will need to carefully analyze the materials to coordinate them for their particular setting. They will also need to plan and prepare thoroughly before using the materials in class, since teacher-provided resources, such as videotapes of undergraduate classes, are frequently required. This book provides sound recommendations, multiple activities, and many worksheets. For teachers who have or can develop their own syllabus, it supplies excellent resource materials for a variety of ITA training programs. WANDA FOX
Purdue University

Languages and ChildrenMaking the Match. Helena Anderson Curtain and Carol Ann Pesola. Reading, MA: Addison Wesley, 1988. Pp. xv + 352. This is an up-to-date and comprehensive book on the field of elementary school foreign language instruction. It is designed by practitioners primarily for practitioners (p. xi) and is intended as a methods text and as a practical guide for teaching language to children. The text is an excellent resource not only for new teachers or teachers-in-training but also for school administrators who are beginning implementation of a K-8 foreign language program. Languages and ChildrenMaking the Match focuses on communication as the highest priority in teaching a language class. In an informative introduction, the fundamental principles of the communicative approach on which this book is based are summarized in 12 Key Concepts for Elementary and Middle School Foreign Languages (pp. xiv-xv). These include: "Successful language learning occurs in a meaningful communicative context, " Successful language learning activities for children incorporate opportunities for movement and physical activity, Successful language learning activities are organized according to a communicative syllabus rather than according to a grammatical syllabus, and Successful language learning activities establish the language as a real means of communication. The first four chapters deal with elementary school foreign language programs, including their history, the selection of an appropriate program model, and program planning. Chapter 5 provides a comprehensive overview of recent theories of second language acquisition that serve as a foundation of effective programs. Chapter 6 provides an overview of immersion programswhat they are, what makes them work, what are their results. Chapter 7 describes useful techniques of content-based instruction used in immersion programs, such as thematic webbing and BOOK NOTICES 509

semantic mapping. The success of these programs demonstrates that communicative competence can be developed-as students are provided with situations that require language use in real communication. The most readable part of the volume is Chapters 8, 9, and 10, which demonstrate meaningful communication in the context of a holistic approach to learning (p. 117). These three chapters illustrate the wide range of the holistic approachesfor example, how to create an environment for communication, how to give children experiences with the culture, and how to plan a communicative classroom day-to-day. In other chapters, different dimensions of a holistic approach are discussed: evaluation, materials, classroom activities, and teacher preparation. Unfortunately, compared to the clear presentation of the language experience approach, the very brief presentation on the topic of evaluation is not entirely satisfactoryonly 14 pages are devoted to it. Some readers might wish this chapter were more detailed. However, an introduction to language instruction based on meaningful communication is successfully made by this book owing to the multidimensional practical examples of teaching strategies. There are excellent questions for discussion in each chapter. This book is highly recommended for foreign language teachers and also for teachers working with limited English proficient (LEP) students. NORIKO ISOGAWA
Purdue University

Conversation Gambits: Real English Conversation Practices. Eric Keller and Sylvia T. Warner. Hove, England: Language Teaching Publications, 1988. pp. 96. What Im trying to say is; Sorry, I dont follow you; I dont understand, can you explain? These are examples of conversation gambits. A gambit is a conversational strategy that promotes discussion. In their introduction to the student, the authors say: We use gambits to introduce a topic of conversation; to link what we have to say to what someone has just said; to agree or disagree; to respond to what we have heard (p. 4). While gambits have little content, they have much meaning: They show our attitude to the person we are speaking to and to what (s)he is saying (p. 4). The authors state that using gambits will make students English sound more natural, will make it easier to converse and be understood. If we never use gambits in our conversation, other people may think we are very direct, abrupt, and even rudethey will get a wrong picture of us as people (p. 4). Useful in a secondary school or an adult ESL program for intermediateadvanced students, Conversation Gambits is an excellent supplement to other conversation or written activities. This well-structured text does what other ESL materials often fail to do: It teaches students native-like 510 TESOL QUARTERLY

conversational phrases to enhance their English skills and at the same time offers insight into cultural expectations by explaining the common and proper usage of particular phrases or idiomatic expressions. For example, the book suggests that l wonder if is a way of giving an open opinion while inviting others to comment too (p. 23). There are 63 lessons divided into three sections: (1) Opening Gambits (e.g., Excuse me for interrupting, but, Could you tell me); (2) Linking Gambits (e.g., How about, In addition, Whats more); and (3) Responding Gambits (e.g., Thats a good idea, but, Youre absolutely right, Exactly! and Ill have to think about that). Each one-page lesson focuses on a single topic, e.g., Changing the Subject, "Generalizing," Getting to Know Someone. A list of gambits is clearly printed on the side of the page. Classroom activities for pairs or small groups are suggested. Interspersed through the text are pictures, charts, graphs, games, and stories about life situations. The last lesson, longer than the others, is called Mini-Conversations. It provides many suggestions for practice of conversation gambits, e.g., Tell your partner a problem (p. 85). These suggestions offer good conversation starters that can easily be adapted to incorporate content from students own lives. (They also could be used as starting points for writing assignments.) Timely topics for conversation are given at the end of this sectione.g., living in the city, politics, healthy eating, smoking, teenage drug abuse. This text fosters a natural conversational approach in the ESL classroom and may also prove useful in English composition classes for monolingual English students. While it is in no way a complete curriculum, Conversation Gambits provides the ESL teacher with a good resource for teaching some of the subtle language messages we often fail to convey in our classes. TERESA GRANELLI
Hofstra University

Crazy Idioms: A Conversational Idiom Book. Nina Weinstein. New York: Collier Macmillan, 1990. Pp. vii + 62. Designed to enrich the ESL curriculum, this book focuses on a vocabulary of 45 common cultural idioms that often are not treated in the classroom. The presentation and exercises are simple enough to be used with high-beginners, and the books cartoon illustrations of each idiom appeal to students of widely varied ages. Each of the nine units introduces five idioms from a single category, such as animal names, foods, colors, body references, or double meanings. Idioms are introduced with an intuitive, discovery learning approach, giving students generous exposure to the idioms before providing BOOK NOTICES 511

definitions. The exercises lead the students to apply the idioms repeatedly at increasingly complex levels. Each unit begins by asking students to guess the meaning of five idioms. This effective reading preparation exercise can be handled in class discussion. A simple, situation-based dialogue then presents the idioms in a context that allows students to infer their meaning. In the third activity, students match each of five context-based cartoon illustrations with the proper idiom. Next, the text briefly defines each idiom; and students, in pairs or as a class, compare each expressions idiomatic meaning with its meaning in the students own cultures. The final activity provides for directed conversation, requiring each student to find a member of the class who fits each idiom in some specific way. At the end of the book is a review of all 45 idioms, again requiring each student to find someone in the class to whom the idiom applies in a particular way. An answer key for the picture-matching exercises is also provided, along with an index of expressions with the page number on which each idiom was introduced. The introduction states that the book comes with an audio cassette. Published in paperback, with glued binding, the thin book is printed on heavy 7 by 9 paper and should wear well. Although spaces are provided in the book for student answers, the book could easily be reused if students write answers on other paper. The books use of reading, writing, and speaking skills to repeat the material on progressively complex levels is effective; however, a few other aspects of the books presentation should be considered. For instance, the idioms presented are common in everyday speech but are rarely used in formal English. At times it is difficult to match the pictures with the correct idiom. In addition, the short definitions at times miss some of the idiomatic meaning and its logical connection with the literal meaning, and the double-meaning idiom section does not discuss the suggested dangerous meanings. Even with these considerations, the material is presented in a lighthearted way, which in itself will enhance students learning. An enterprising teacher could apply the effective format to other idioms as well. Crazy Idioms can provide effective enrichment for an ESL curriculum. EDA ASHBY
Brigham Young University

Words at Work: Vocabulary Building Through Reading. Betty Sobel and Susan Bookman. New York: Collier Macmillan, 1989. Pp. xiv + 122. This text is written for young adults and adults at the high-beginner and low-intermediate level. It is designed to cultivate a comfortable, positive, fearless attitude toward reading (p. ix) by emphasizing lexical 512 TESOL QUARTERLY

development in using high-interest readings and practice exercises that allow students to work alone, with partners, and in groups (p. xi). The 10 chapters constitute self-contained lessons. Each chapter follows the same format. It begins with preview questions on the reading topic that follows. Next, there is a short reading followed by a true-false comprehension exercise. The topics for the reading are either on life in the United States (i.e., surrogate parenting, or television and children) or on personal development (i.e., shyness, or money and credit). Students are asked to fill in sentences with words selected from the reading and then proceed to definition exercises in which they first define words and then fit them into sentences that form a dialogue. Next there is practice with derived forms. The lesson continues with a new paragraph that summarizes the original reading using the new vocabulary. Each lesson ends with a Wrap-Up Activity promoting student interaction. Is this text successful? On the positive side, students are provided many opportunities to manipulate the highlighted lexical items. The exercises developed for such work are generally well executed, although in some fill-in activities students would be able to figure out correct responses on the basis of grammatical, not lexical, knowledge. The items selected for practice are appropriate for students at the targeted ESL/EFL level. Thus, one hopes students reading abilities will increase as a result of their expanded vocabulary. However, the text fails on other accounts. Effective reading requires more than good lexical knowledge; good readers can deduce meaning from context. Words at Work provides no practice in such skill development. Another problem is the claim that each article is written in natural English, similar to the style of magazine articles (p. ix). Since the reading passages are not credited, the claim for natural language is undocumented. The articles appear to have been adapted, which may perhaps disappoint those instructors who, on the basis of the introduction, expect authentic texts. In a similar vein, the claim that the text is an interactive reader (p. xi) apparently means that students engage in some activities with partners or groups, not that the students interact with the text along the lines suggested by current reading research. In short, those seeking authentic texts and natural language or an interactive approach to reading instruction will need to search elsewhere. Finally, the topics chosen and some of the activities related to them seem culturally biased. For example, shyness is portrayed as a liability that should be overcome (Lesson 1). For some readers, the topics may be either offensive or irrelevant. Students are expected to debate a case of surrogate parenting (Lesson 10) and to devise a will (Lesson 5). Are these topics appropriate for all cultures within and outside of the United States? Is there a subtle U. S., middle-class bias involved in these readings? In summary, while Words at Work probably can deliver on its promise of building vocabulary, as a general reading text that would build other BOOK NOTICES 513

reading skills and appeal to ESL/EFL students, it is deficient in several key areas. For those considering using this text I would advise caution. ELLIOT L. JUDD
University of Illinois at Chicago

Techniques and Resources in Teaching Grammar. Marianne Celce-Murcia and Sharon Hilles. New York: Oxford University Press, 1988. Pp. 189. In their introduction the authors present a carefully constructed case for the necessity of teaching grammar to certain second language learners. The first chapter counters the current antigrammar stance by citing the profile of students who show a pattern of high vocabulary but fossilized inaccurate grammar arrived at through street learning or through communication first programs and who are unsuccessful at increasing their linguistic ability (Higgs & Clifford, 1982, cited in Celce-Murcia & Hilles, p. 3). The authors proceed to outline a framework for teaching grammar based on three aspects of language: social roles and communicative functions (e.g., politeness), semantic notions expressed through grammatical structures (e.g., prepositions), and discourse factors (e.g., word order and topic continuity). To these aspects they link the teaching techniques of dramatizing, responding to realia, and manipulating texts. The teacher new to ESL will find a wealth of concrete ideas for teaching grammar. However, while the practicality of the text is beyond question, there is an overly structuralist tone in many of the chapters despite the interweaving of communicative classroom techniques. The primary justification given for this is that we are more often obliged to teach district- or school-prescribed syllabuses, which usually are structurally based (p. 23). This is a disappointing caveat for those who believe that grammar must play an integral part in the communicative classroom. It would have been preferable for the authors to have derived the grammatical syllabus from actual student errors, and to have described the teaching of grammar in the context of teaching the other ESL skills, particularly reading and writing. Interwoven with the teaching examples are numerous comments concerning classroom management, such as the benefits of cooperative learning and the strict use of English in the classroom. One problem is that no mention is made of the differences that might arise in an EFL situation where, for example, the practicality of using the students native language may at times be justified. A weakness of the book is the unstated assumption that the readers will be teaching ESL classes in an adult school setting. Despite these problems, the text serves well as an introduction to the teaching of grammar. It provides many examples of simple, practical techniques that make use of everyday materials, from classroom furniture to pop songs to magazine pictures. The authors emphasize the fact that 514 TESOL QUARTERLY

their ideas are only suggestions to encourage the teacher to find innovative ways to contextualized grammar instruction. Through example, the book demonstrates how a teachers creative approaches can help ESL students gain maximum benefit from instruction in grammar. PETER MASTER
California State University, Fresno

A Linguistic Study of American Punctuation. Charles F. Meyer. New York: Peter Lang, 1987. Pp. xv + 159. Meyers purpose with this book is to set forth a comprehensive treatment of contemporary U.S. (termed Modern American) punctuation based on linguistic principles. Specifically, his focus is on those punctuation marks whose uses have not been rigidly conventionalized: that is periods, question marks, exclamation marks, commas, dashes, semicolons, colons, and parentheses (p. xiii). Chapter 1 overviews the history of U.S. punctuation and its linguistic basis. Chapters 2 through 4 outline the syntactic, semantic, and prosodic bases of punctuation norms. Chapter 5 specifies the pragmatic bases for overriding these norms. Chapter 6 details the close correspondence between actual practice and the prescriptive rules given in style manuals. Finally, Chapter 7 summarizes the rules and general principles of punctuation. Several appendixes conclude the book. Meyers thesis is that punctuation should be viewed as a system consisting of a small set of general principles. For example, in considering coordinate constructions, Meyer discusses a punctuation hierarchy and one general principle. The hierarchy ranks the period, the semicolon, and the comma in descending order. Meyers principle states that the lengthier or more complex a coordinated construction . . ., the greater the need to punctuate it with a mark higher on the punctuation hierarchy (p. 119). In a sentence like My mother went to the bank (punctuation mark) and my father went to the store, each clause is short and simple. Hence a comma (lowest on the hierarchy) is appropriate. This view of punctuation as a system of general principles has straightforward implications for the teaching of U.S. English. For example, suppose that in a choppy, hard-to-read essay, a student uses a period to separate the following clauses: My mother went to the bank. And my father went to the store. Using Meyers system, the instructor can present the punctuation hierarchy and principle, pointing out that the clauses are short, simple, and structurally parallel, thereby arguing for a mark lower on the punctuation hierarchy. Once the students master the principle, they can apply the same criteria to a huge class of coordinated constructions. Without Meyers system of general principles, the instructor is restricted to idiosyncratic correction (e.g., use comma here) or reference to a specific rule in a style manual. BOOK NOTICES 515

Meyers approach also has implications for nonnative speakers learning U.S. punctuation. Since different languages utilize different linguistic structures in different ways, one question researchers might ask is how a persons native linguistic competence affects or interferes with the acquisition of the punctuation of U.S. English. With over 250 examples and 35 different tables, Meyers book should also serve as a valuable summary of the data and principles of punctuation. Moreover, Appendix 1 details how 13 different style manuals treat 53 different punctuation usages (e.g., the acceptability of the dash for emphasis as in The manan incredible foolshould be fired). In sum, Meyers view of punctuation as a system of principles and his extensive surveys of punctuation data should prove especially useful to TESOL specialists and others engaged in either teaching or analyzing written U.S. English. K. SCOTT FERGUSON
Harvard University

Longman Preparation Course for the TOEFL. Deborah Phillips. New York: Longman, 1989. Pp. vi + 282. The Longman Preparation Course for the TOEFL comes close to meeting its back cover claims of giving students the skills, strategies, practice, and confidence they need to increase their scores on this important exam. Although the book does not touch on the Test of Spoken English and deals with the Test of Written English very superficially, the materials and exam practice given for the three main sections of the test are exemplary. The book claims to be all things to all people: a core text for TOEFL preparation, a self-study tool for students preparing for the test, and a supplementary text in a more general ESL course. The latter use would seem inadvisable, as the text deals as much with test-taking strategies as it does with language and skills content. Additional materials are available to supplement the materials in this volume. A separate typescript and answer key are needed to-work fully with the text. The complete Longman program for the TOEFL includes another book, Longman Practice Tests for the TOEFL, with its own typescripts and answer keys included. Cassettes (two in each set) go with each book. The practical, frankly beat-the-test orientation shows up in such instructions as choose the answer that sounds different (p. 18) or listen to the second line of the conversation (p. 50), both in the section on Listening Comprehension; and do not spend a lot of time looking for contextual clues to the meanings of the words (p. 208) in the section on Vocabulary and Reading Comprehension. Strategies for each question type in the three main sections of the TOEFL are described and the teacher is reminded to do the practice exercises in class as it is important 516 TESOL QUARTERLY

to keep the students under time pressure (p. 5) while they do the exercises. Unlike most TOEFL preparation courses, Phillips work provides little real TOEFL practice; the text begins with a pretest and ends with a posttest, as does each section dealing with each of the three sections of the TOEFL. These tests seem to have greater motivational than instructional import; students are instructed in converting their scores to a 200-700 TOEFL-like scale and then chart their progress (p. 270). Many exercises in the book adopt the TOEFL format but deal only with a specific skill or language item in focus. The real strength of the text lies in predicting areas in which students are probably going to have difficulty (based on errors actually made on the TOEFL) and then systematically dealing with these areas, most successfully in the Structure/Written Expression section. Here such common errors as subject-verb agreement are dealt with through brief explanatory notes, exercises, and finally TOEFL-like practice. A problemsolving approach follows the current trend in learner accountability: Teachers are a resource and their job is to assist the students in finding the various ways that the sentences can be corrected (p. 5). Material based on specific problems permits teachers to omit certain sections of the text if they dont like the way the problem is addressed or if the language point does not present a problem for their students. The Longman Preparation Course for the TOEFL has a pleasing appearance, contains realistic testing material and answer sheets, and provides adequate remedial work and test-taking strategies for the three main sections of the exam. This is what it was written to do and the author, publisher, and those who adopt it should not expect it to do any moreor any less. TERESE THONUS
Cultura lnglesa, Rio Grande do Sul, Brazil

Doublespeak, William Lutz. New York: Harper & Row, 1989. Pp. xiii + 290. Although we are language professionals, ESL teachers may not fully appreciate the degree to which the manipulative language of advertisers, bureaucrats, and politicians has come to pervade the public discourse. In an effort to draw attention to the trend, William Lutz has compiled into book form examples of doublespeak, which he defines as language that can mislead, distort, deceive, inflate, circumvent, obfuscate (p. 2). At its least offensive, doublespeak is inflated language (p. 9). Consisting mostly of material previously published in the Quarterly Review of Doublespeak, a publication of the National Council of Teachers of English edited by Lutz, this catalogue of examples from across the spectrum of public discourse is a soberingand entertainingaccount of the phenomenon of doublespeak. BOOK NOTICES 517

The term doublespeak was inspired by the work of George Orwell. In his book 1984, Newspeak is the official state language; doublethink is exemplified by the slogan War is Peace. Lutz quotes Orwells 1946 essay, "Politics and the English Language" The great enemy of clear language is insincerity. When there is a gap between ones real and declared aims, one turns as it were instinctively to long words (p. 8). Lutz notes that the problem is with the intent of the speaker or writer, not with the language but with those who use it. Doublespeak is not simply another inoffensive manifestation of the change all living languages are subject to. The existence of doublespeak is by design, not by natural evolution. In too many cases, its manifesto is manipulation and its primary goal is to dissemble. Teachers might use Lutz book to raise awareness of doublespeak and some of its subtleties. They could develop lessons in critical thinking, especially for reading courses where objectives include inferencing, distinguishing fact from opinion, and identifying tone and bias. Doublespeak is a rich resource to use in the classroom. One familiar phrase that Lutz dissects yields not only to grammatical and lexical, but also to pragmatic analysis: Twice as much of the pain reliever doctors recommend most (p. 95). Twice as much as what? he wonders. And just what is the pain reliever doctors recommend most, anyway? According to Lutz, its plain aspirin; the hollowness and pretension of the phrase begin to emerge. Teachers can help students wade through other odious language, such as this response to a request for a raise: Because of the fluctuational disposition of your positions productive capacity as juxtaposed to government standards, it would be monetarily injudicious to advocate an increment (p. 215). Most of the examples in Doublespeak are drawn from the published equivocations of politicians and bureaucrats. For example, there is the helicopter that, as the National Transportation Safety Board once reported, failed to maintain clearance from the ground (p. 214). And who is trying to hide what behind an expression like, predawn vertical insertion (p. 7)? (Remember Granada?) Ironically, teachers are partly to blame for the proliferation of doublespeak. In a lengthy section on education doublespeak, Lutz finds that educational journals and reports are replete with the stuff. Even English teachers have fallen for it. Lutz cites a study in which English teachers are found to prefer convoluted passages like those above to passages that say the same thing in simpler prose. Lutz encourages teachers to be aware of doublespeak: They should be leading the fight against doublespeak by teaching their students how to spot it, how to defend themselves against it, and how to eliminate it in their own writing and speaking (p. 63). VINCENT G. BARNES
University of Washington

518

TESOL QUARTERLY

BRIEF REPORTS AND SUMMARIES

The TESOL Quarterly invites readers to submit short reports and updates on their work. These summaries may address any areas of interest to Quarterly readers. Authors addresses are printed with these reports to enable interested readers to contact the authors for more details.

Edited by GAIL WEINSTEIN-SHR

University of Massachusetts at Amherst/ Temple University

Listening Perception Accuracy of ESL Learners as a Variable Function of Speaker L1

GEORGE YULE, SUSAN WETZEL, and LAURA KENNEDY
Louisiana State University

As we incorporate into our ESL speaking classes a greater number of interactive tasks involving nonnative-speaker/nonnative-speaker (NNS/ NNS) dyads or groups, trying to create the optimum conditions for learners to benefit from negotiated input (Long, 1983; Doughty & Pica, 1986), we have tended to move away from exercises that focus on linguistic form, particularly in terms of the pronunciation and perception of English sound contrasts at the syllable or word level. This would seem to be justified if we could be sure that, as an inevitable part of receiving negotiated input, learners were in fact developing sufficient accuracy in the production and perception of those features of spoken English that play a crucial part in comprehension. According to Long and Porter (1985), accuracy does not suffer when learners take part in interactive pair work with other NNSs. However, this claim was primarily based on the use of grammatical structures and seems, in our experience, to be less tenable when we think of some learners whose accuracy in spoken production does seem to be subject to some variation. We do not know of any studies focusing specifically on level of pronunciation and perception accuracy in NNS/NNS pair work, but we have observed that learners seem to get by with fairly inaccurate pronunciations (in terms of the target) when their NNS partner is very familiar with their speech, particularly in the EFL context where interlocutors share the same L1 (Kenworthy, 1987). We wondered if ESL learners actually found it easier to identify English words when these were spoken by other learners than when they were spoken by English native speakers. In an attempt to answer this question, we designed a listening perception exercise to investigate whether learners became more or less accurate in their identification of English words as a function of the L1 of the speaker. 519

THE STUDY A group of 25 students of high-intermediate to low-advanced ability, enrolled in sections of a Spoken English class at Louisiana State University, voluntarily took part in our study. There were 13 Spanish first language (SL1) speakers, 6 Vietnamese (VL1) speakers and 6 Chinese (CL1) speakers. They were individually recorded reading two different sets of 40 sentences. Every sentence came in two versions, to provide a minimal pair contrast. For example, when one set contained Everyone was present, the other set contained Everyone was pleasant. We also had a U.S. English speaker (EL1) record both sets. From these recordings we created, for each individual learner, cassette tapes with sets of sentences such that each learner had to listen to an English NS, a same L1 NNS, a different L1 NNS and him/herself. Sample sentences for each listener were randomly selected from each input source, with the result that there was no control over which particular minimal pair contrasts were represented on each listeners cassette. When the learners listened to their cassette tapes, they had to decide which word had been spoken in each sentence. For example, the learner would hear a spoken version of Put these in the bag, and have to indicate on an answer sheet which member of a pair of words had been spoken: Put these in the (bag/back). (For a more detailed description of the technique, see Yule, Hoffman, & Damico, 1987). We thereby collected perception accuracy scores for our three L1 groups listening to four different English input sources: NS, same Ll, different Ll, and self. The accuracy results, expressed as mean percentages for each L1 group along with the results of the analysis of variance (ANOVA) by group are presented in Table 1.
TABLE 1 Mean Perception Accuracy (%) and ANOVAs by Listener Group

RESULTS AND DISCUSSION Considering all the learners in our study as a single ESL population, we find that there is a substantially higher level of accuracy when the 520 TESOL QUARTERLY

individual learners try to identify what they themselves have said, with between self and each of the others.1 This should not be surprising since individuals speaking a second language must, via the self-monitoring process, become their own most common personal input providers, and inevitably the most familiar (Gass & Varonis, 1984). The same pattern as for the overall group (mean perception accuracy for self differing significantly from each of the other speaker conditions, but no real differences between any of those other conditions) is repeated group, only the difference between listening to self and listening to same SL1 groups, the relatively minor differences observed between their perception accuracy scores when trying to identify what was said by an English NS and a NNS (either with same or different Ll) would tend to offer support to those who claim that the source of L2 input need not be a native speaker. However, the results for the Chinese L1 group present a quite different picture. For this group, there is no difference in accuracy between listening to self and to others with the same L1. Here the critical difference in mean difference in listening to self versus an English native speaker and listening to a same L1 speaker versus an English native speaker. That is, in marked contrast to the SL1 and VL1 groups, the Chinese L1 group had significantly greater accuracy in identifying English words when spoken by another Chinese L1 speaker than when spoken by a native English speaker. What might account for this difference in L2 perception accuracy according to the L1 of the input provider among these groups? When we conducted our investigation of listening accuracy, we had also gathered information on a number of other factors. In analyzing the relationship between these factors and the various listening accuracy scores, we could find no connection between listening scores and self-reported factors such as amount of time using English outside class, number of same L1 friends, number of different L1 friends, or number of English L1 friends. Despite their generally having the lowest accuracy scores, the CL1 group reported a greater average number of years studying English (8.29) than the SL1 group (7.34) and the VL1 group (5.13). However, no significant correlation was found between years studying English and any listening accuracy condition. Nor was there any significant correlation between any listening accuracy condition and reported years living in the United States: CL1 (2.17), SL1 (1.66), VL1 (4.58). What did provide a possible clue was the apparent difference in the type of English language experience that the CL1 and VL1 groups had had.

BRIEF REPORTS AND SUMMARIES

521

If we compare the figures for time spent studying English and time spent in the U. S., we might infer that these account for differences in L2 perception accuracy. The general English language learning experience of the Vietnamese group had been in an ESL context (about 90%), but not so for the Chinese group (about 25%). While the VL1 group had learned English in classes with learners from other L1 backgrounds, with English NS teachers, and surrounded by the English language, the CL1 group confirmed that they had had a traditional EFL experience, with very limited English speaking or listening components. Whatever listening experience they had been provided with involved Chinese L1 learners and teachers speaking English. This scenario may provide some insight into the fact that the CL1 group had achieved their highest accuracy scores when listening to other CL1 speakers. This explanation, however, does not seem as strong when we examine the time spent by the SL1 group studying English in the ESL versus the EFL context. These SL1 students had, on average, no more ESL experience than the CL1 group (about 22%), yet had coped better with the English NS input. We might point out, however, that these SL1 learners, from South and Central America, had generally had much greater exposure to spoken U.S. English, both in and out of their EFL classes, prior to arriving in the U.S. Their general familiarity with U.S. cultural references, for example, was observed in class to be much greater than that of the CL1 students. CONCLUSION It is impossible, in this type of study, to measure the effect of previous EFL instruction on performance within the ESL situation of a U.S. university. We should, however, try to remember that the individual learners in our ESL classes may have had qualitatively quite different EFL learning experiences and developed quite different levels of ability in specific skill areas. When we advocate NNS/NNS pairings in spoken interaction tasks in the language classroom, we should be sensitive to the potential differences in ability among different L1 learners to cope with English language input from learners with different Lls. While the present study has been limited to simple perception accuracy in isolation and not within an ongoing interaction, it does provide some grounds for suspecting that what might count as comprehensible input from a Chinese L1 speaker to a Spanish L1 speaker may not count as such in the opposite direction. Finally, it must be clear from our results that claims from research with one specific L1 group of learners may not be accurate with regard to another L1 group even, as in this case, when those groups are currently sharing the same classroom experience. Consequently, when making claims about, for example, the performance of a small group of Spanish L1 ESL learners on some task, in a particular setting, we should be extremely cautious about turning those claims into powerful and unqualified statements about what all learners do. 522 TESOL QUARTERLY

REFERENCES Doughty, C., & Pica, T. (1986). Information gap tasks: Do they facilitate second language acquisition? TESOL Quarterly, 20 (2), 305-325. Gass, S., & Varonis, E. M. (1984). The effect of familiarity on the comprehensibility of nonnative speech. Language Learning, 34, 65-89. Kenworthy, J. (1987). Teaching English pronunciation. Harlow, England: Longman. Long, M. H. (1983). Linguistic and conversational adjustments to non-native speakers. Studies in Second Language Acquisition, 5, 177-193. Long, M. H., & Porter, P. A. (1985). Group work, interlanguage talk, and second language acquisition. TESOL Quarterly, 19 (2), 207-228. Yule, G., Hoffman, P., & Damico, J. (1987). Paying attention to pronunciation: The role of self-monitoring in perception. TESOL Quarterly, 21 (4), 765-768. Authors Address: Linguistics Program, 136 Coates Hall, Louisiana State University, Baton Rouge, LA 70803.

Using Brainstorming and Clustering with LEP Writing to Develop Elaboration Skills
ANDREA B. BERMUDEZ and DORIS L. PRATER University of Houston-Clear Lake The need to develop intervention and prevention programs for at-risk populations has clearly become a national educational priority as dropout levels continue to escalate (United States General Accounting Office, 1987). To date, most educational models used with minority students, particularly the limited English proficient (LEP), have approached instruction from the standpoint of students deficits rather than their strengths. This focus has resulted in temporary and costly solutions to the problem (Fernandez, Bermudez, & Fradd, in press). The development of writing skills in LEP and at-risk students has been largely ignored by educators in the field. In addition, these writers face several challenges in developing composing skills: (a) lack of awareness of critical cognitive processes inherent to good writing, e.g., clustering related ideas and selfdirected memory searching (Englert & Raphael, 1988); (b) inability to transform conversational patterns into writing (Bereiter & Scardamalia, 1982); (c) lack of ability to regulate their own and others comprehension of text and to organize ideas for writing (Englert & Raphael, 1988); (d) dependence on external criteria and resources (Englert & Raphael, 1988); and (e) conception of language as a set of discrete and mutually independent skills (Padron & Bermudez, 1988). In addition, recent studies of LEP writers suggest that these students focus on form to the detriment of content or ideas (Padron & Bermudez, BRIEF REPORTS AND SUMMARIES 523

1988; Widdowson, 1978; Zamel, 1982, 1983). LEP writers are not systematically using strong metacognitive strategies such as planning, brainstorming, and considering the audience; instead, they are focusing on grammatical features and punctuation (Bermudez & Padron, 1988; Raimes, 1980). Furthermore, being able to plan, draft, and revise requires the ability to use recursive processes that tend to overwhelm the LEP writer. Scardamalia and Bereiter (1986) report that lack of sufficient metacognitive awareness necessary for successful writing would jeopardize the quality of the written product. Consequently, developing metacognitive awareness of the writing process is a good starting point for these learners. Additionally, strategy instruction seems to be a promising writing methodology for the LEP student (Chamot & OMalley, 1987). Teaching cognitive strategies has also shown to be effective in helping students make the transition from oral to written language (Bereiter & Scardamalia, 1982; Chamot & O'Malley, 1987). Mapping strategies, for example, have been found to help students establish priorities and focus their writing (Miccinati, 1988). This strategy has been defined in the research literature as a technique for externalizing the individuals cognitive structure by diagraming his/her knowledge base of concepts and the relationships among concepts (Novak & Gowin, 1984). Similarly this type of strategy improves comprehension of text from elementary school to adult levels (Prater & Terry, 1988; Singer & Bean, 1984; Slater, Graves, & Piche, 1985) and retention of information (Miccinati, 1988). In addition, mapping assists in developing more cohesive (Ruddel & Boyle, 1984; Miccinati, 1988) and longer essays (Ruddel & Boyle, 1984). However, the impact of these strategies on ESL writing has not been examined. As a result of the pressing need for additional research in this area, the researchers conducted the present study to investigate the effects of brainstorming and clustering on the development of written language fluency and the elaboration of ideas in LEP students.

METHOD Subjects The sample consisted of two groups of 16 third- and fourth-grade Spanish-speaking LEP students from a low-income urban school in Texas. Procedure Students were matched according to their Individualized Developmental English Activities (IDEA) language proficiency scores. One student from each pair was randomly assigned to a treatment group, one to a control group. The same teacher worked with both groups for 45 minutes per day for 6 days. Three reading selections from the basal reader that was the class text were used (Arnold, Smith, Blood, & Lapp, 1987). The 524 TESOL QUARTERLY

selections were factual/informative reading passages. Two days were spent on each of three selections. The treatment group brainstormed ideas about the topic of the story using the title of the selection and illustrations as a stimulus for their thinking. The teacher served as scribe, placing the ideas on a transparency. The students read the selection silently for the remainder of the class period. The next day the teacher led the group in brainstorming other ideas that they had gathered from their reading. Then, with the assistance of the teacher, the students clustered the ideas by circling like topics in the same color ink pen. Next, the treatment group wrote a paragraph about the story. A 6-item comprehension measure was given to the students at the end of the 2 days spent on each respective selection. One week later a retention measure made up of the three comprehension measures was given. With the control group, the teacher introduced the selections using preliminary questions provided in the basal reader. Then, the students read the selection silently. The next day the teacher led a class discussion based on questions provided in the basal reader and the student wrote a paragraph about the selection. The same comprehension and retention measures were given to the control group. Evaluation Measures Comprehension measure. Three 6-item tests based on the respective stories were developed by the researchers. Items were written at both literal/factual and inferential/interpretive levels. For each student, the three test scores were summed to yield a single comprehension score (range 0-18). Retention measure. The 18 items were administered 1 week after the last class session as a comprehensive retention measure. Writing measures. The essays were analyzed for fluency (number of words, number of idea units, number of main ideas); for elaboration (number of ideas beyond text material); and for organization (number of clusters). Idea units are defined by Gere & Abbott (1985) as a single clause, independent or dependent. The number of main ideas in the reading selections was determined by having two graduate students read each selection and list the main ideas. In most cases these main ideas were topic sentences within each paragraph, stated or implied. The final selection of main ideas was agreed upon by both readers. Student essays were then scored against this list. The number of ideas beyond the text material was determined by two independent readings of the essays. New material added to the essay that was not present in the text was counted. The number of clusters was determined by counting the number of occasions in which two or more sentences with related ideas were expressed consecutively within the essay. BRIEF REPORTS AND SUMMARIES 525

RESULTS The means, standard deviations, and t-values for each of the dependent measures are shown in Table 1. Elaboration was significantly higher for the treatment group. No significant differences between the two groups were found on measures of fluency, organization, comprehension, or retention. Note, however, that the use of multiple t-tests calls into question the significance of the elaboration measure. Thus, conclusions reported here should be interpreted as suggestive only; further research is needed in this area.
TABLE 1 Comparison of Experimental and Control Groups on Each of the Evaluation Measures

Discussion It may be that the graphic representation of concepts enables the LEP writers to expand their discussion of materials presented in the basal reader. The mapping activities and exchange of related ideas may activate prior knowledge and facilitate linkages with new knowledge. The original ideas presented beyond those contained in the text suggest that the learner is actively involved in the process. While not statistically significant, the finding that the treatment group essays were better organized than the control groups is also suggestive. The clustering activities may provide the necessary mechanism for fostering the grouping of related ideas in written products. The maps may facilitate visualization of conceptual relationships among parts and between parts and the whole. These learners may need more practice over an extended period of time. Educational Implications If future research finds these effects to be significant, several 526 TESOL QUARTERLY

implications follow. In both brainstorming and clustering procedures, the goal of the classroom teacher should be to move the student to independent use of such techniques. The comprehension of factual/ informative texts is essential to handling reading in content areas. LEP writers need to be provided a variety of tools to help them cope with demands made upon them in content classes. The use of cognitive strategies to enhance the development of composing skills may well be effective with LEP learners. Strategy-oriented writing instruction enhances the opportunities for the LEP student to use writing as a learning tool as well as a skill to develop other language and thinking skills.

REFERENCES
Arnold, V. A., Smith, C. B., Blood, J., & Lapp, D. (1987). Connections: Level 3.2. Observing. New York: Macmillan. Bereiter, C., & Scardamalia, M. (1982). From conversation to composition: The role of instruction in a developmental process. In R. Glaser (Ed.), Advances in instructional psychology (Vol. 2, pp. 1-64). Hillsdale, NY: Lawrence Erlbaum. Bermudez, A. B., & Padron, Y. N. (1988). Teachers perceptions of errors in second language learning and acquisition. In L. M. Malave (Ed.), NABE 87. Theory Research and Applications: Selected Papers (pp. 112-124). Fall River, MA: National Dissemination Center. Chamot, A. U., & OMalley, J. M. (1987). The cognitive academic language learning approach: A bridge to the mainstream. TESOL Quarterly, 21 (2), 227-249. Englert, C. S., & Raphael, T. E. (1988). Constructing well-formed prose: Process, structure and metacognition in the instruction of expository writing. Exceptional Children, 54 (6), 513-520. Fernandez, M. R., Bermudez, A. B., & Fradd, S. L. (in press). The Hispanic dropout cycle: A proposal for change. Southwest Journal of Educational Research Into Practice. Gere, A. R., & Abbott, R. D. (1985). Talking about writing: The language of writing groups. Research in the Teaching of English, 19, 362-379. Miccinati, J. (1988). Mapping the terrain: Connecting reading with academic writing. Journal of Reading, 31, 542-552. Novak, J. D., & Gowin, D. B. (1984). Learning how to learn. New York: Cambridge University Press. Padron, Y. N., & Bermudez, A. B. (1988, Spring). Promoting effective writing strategies for ESL students. Southwest Journal of Research Into Practice 2, 19-27. Prater, D. L., & Terry, C. A. (1988). Effects of mapping strategies on reading comprehension and writing performance. Reading Psychology, 9 (2), 101-120. Raimes, A. (1980). Composition: Controlled by the teacher, free for the student. In K. Croft (Ed.), Readings on English as a second language: For teachers and teacher trainees (pp. 386-398). Cambridge, MA: Winthrop. Ruddell, R., & Boyle, O. (1984). A study of the effects of cognitive mapping on reading comprehension and written protocols. (Technical Report No. 7). Berkeley: University of California Press. Scardamalia, M., & Bereiter, C. (1986). Research on written composition. In M. Wittrock (Ed.), Handbook of research on teaching (3rd ed, pp. 778-803). New York: Macmillan.

BRIEF REPORTS AND SUMMARIES

527

Singer, H., & Bean, T. (Eds.). (1984). Learning from texts: Selection of friendly text. Proceedings of the Lake Arrowhead Conference on Learning from text. (ERIC Document Reproduction Service No. ED 251 512). Slater, W., Granes, M., & Piche, G. (1985). Effects of structural organizers on ninth grade students comprehension and recall of four patterns of expository text. Reading Research Quarterly, 20, 189-202. United States General Accounting Office. (1987). School dropouts: The extent and nature of the problem. (GAO/HRD Report No. 86-106BR). Washington, DC: Author. Widdowson, H. (1978). Teaching language as communication. London: Oxford University Press. Zamel, V. (1982). Writing: The process of discovering meaning. TESOL Quarterly, 16 (2), 195-209. Zamel, V. (1983). The composing processes of advanced ESL students: Six case studies. TESOL Quarterly, 17 (2), 165-187. Authors Address: University of Houston-Clear Lake, 2700 Bay Area Blvd., Houston, TX 77058.

528

TESOL QUARTERLY

THE FORUM
The TESOL Quarterly invites commentary on current trends or practices in the TESOL profession. It also welcomes responses or rebuttals to any articles or remarks published here in The Forum or elsewhere in the Quarterly.

Comments on James W. Tollefsons Alien Winds: The Reeducation of Americas Indochinese Refugees and Elsa Auerbachs Review
Two Readers React. . .
DONALD A. RANARD and DOUGLAS F. GILZOW
Center for Applied Linguistics

In her recent review of James Tollefsons Alien Winds (1989), Elsa Auerbach (Vol. 24, No. 1, Spring, 1990) finds little to criticize and much to praise, recommending the book as required reading for all ESL educators. It is not surprising that the review was favorable, since Auerbach is singled out for special thanks in the acknowledgments section of Alien Winds, and her own work suggests a sympathy for Tollefsons point of view. It was disturbing, however, that there was apparently no attempt to verify Tollefsons charges against the Overseas Refugee Training Program (ORTP). In fact, Auerbach highlights some of the more sensationalistic accusations, describing raw sewage flowing through refugees living quarters and teachers extorting sexual favors and bribes from students. Had the reviewer taken steps to check the facts, she would have discovered dozens of inaccuracies and distortions in the book. It is for this reason that we wish to provide an alternative view of the book and the program it attacks. Originally planned as a temporary response to the refugee crisis in Southeast Asia, the Overseas Refugee Training Program has evolved into a sophisticated educational program that has continued for over a decade. It has operated under unique constraints, employing Thai and Filipino educators to prepare refugees for the U.S. in camps 10,000 miles away. 529

Alien Winds makes serious charges against the Overseas Refugee Training Program and the living conditions at one of the sites. Our comments focus on the educational program since space does not permit a full discussion of both issues, and because education is the area most relevant to readers of the TESOL Quarterly. However, two points should be noted about the processing centers. First, they are not American run, as Tollefson asserts (p. 16); they are operated by the host country governments and the United Nations High Commissioner for Refugees. Second, conditions in the processing centers are superior to those in any first-asylum camp in Southeast Asia, in terms of housing, sanitation, and securityand those conditions have steadily improved over time. In a recent report (Pihl, 1990), a consultant to Lutheran Immigration and Refugee Service characterizes the Philippine Refugee Processing Center as a country club (p. 8), in comparison to first-asylum camps in the region. According to Tollefson, for the past ten years the U.S. government has spent millions of taxpayers dollars providing English as a second language instruction and cultural orientation to Indochinese refugees as part of a systematic effort to divest these people of their cultures, inculcate them with new values of subservience, and then track them into dead-end menial jobs required by the U.S economic system. This attempt, he maintains, shares the same fundamental purpose as efforts by turn-of-the-century educators to Americanize immigrants, and has led the overseas program into shoddy pedagogy and an alarming number of abuses against the refugees. If Tollefson is correct, he has uncovered an educational conspiracy of unprecedented magnitude, involving thousands of U. S., Thai, Indonesian, and Filipino educators, and organizations such as the Experiment in International Living, Save the Children Federation, World Education, International Catholic Migration Commission (ICMC), World Relief Corporation, and the Center for Applied Linguistics. In his single-minded attachment to his point of view, Tollefson not only ignores contrary evidence but also shapes the facts to fit his thesis. We find half-truths, inaccuracies, misleading examples, and simplistic generalizations throughout Alien Winds in criticisms of the staff, the curriculum, the infrastructure, and other aspects of the Overseas Refugee Training program in Thailand and particularly in the Philippine Refugee Processing Center. Furthermore, the author fails to take into account changes that have taken place in the program since he left it in 1986. 530

TESOL QUARTERLY

EMPLOYMENT In Alien Winds, Tollefson argues that the Overseas Refugee Training Program focuses almost exclusively on employment and that it trains refugees for dead-end, minimum-wage jobs by teaching them the language and behavior of subservience and obedience. This view is a serious misrepresentation of the program and, in our view, reflects a skewed understanding of the realities that refugees face in the U.S. The program does not, as is charged, define successful resettlement solely in terms of employment. While employment is an important topic in the program classes, other topics, such as health, transportation, shopping, directions, and personal information, far outnumber employment-related areas in both the ESL and Cultural Orientation curricula. Interestingly, in the mid-1980s, at a time when almost all federally funded training for refugees in the U.S. had become employment-related, thus restricting eligibility to only employable refugees, the overseas program was committing its resources to better meet the needs of children, adolescents, and homebound mothers. Tollefson does not report these and similar developments that do not support his thesis. (The special curriculum for homebound women is mentioned, but surprisingly is used as proof of the programs failure to force these women into minimum-wage employment.) In Alien Winds, not only is the amount of attention paid to employment exaggerated, but the quality and content of employmentrelated instruction is also misrepresented. Tollefsons charge that the ORTP willfully disregards the employment backgrounds of students, many of whom (he suggests) are well-educated professionals, misrepresents who the refugees are and what the program does. In fact, since the ORTP was established, only a minority of the refugees have been well-educated professionals. In contrast to the Indochinese refugees who arrived in 1975, many since then have had rural backgrounds, little formal education, and no previous contact with people from the U.S. or other Westerners (Caplan, Whitmore, & Bui, 1985; Office of Refugee Resettlement, 1983; Rumbaut, 1985). For example, the average number of years of education for the 1988 arrivals was 4.2, compared to 9.5 years for those arriving in 1975 (Office of Refugee Resettlement, 1983, 1989). While it is true that the Work Orientation component has concentrated its resources on less educated refugees, it is also true that in recent years the component has taken into account the varying backgrounds of refugees. Students in these classes are THE FORUM 531

grouped according to their previous work experience and levels of education. Furthermore, many of the better educated refugees are not students in the program at all, but rather serve as resources in Cultural Orientation classes and in programs for elementary schoolaged children and adolescents. These English-proficient refugees are offered additional classes in TOEFL preparation, advanced English, and career exploration. Tollefsons claim that the program teaches refugees the language and behavior of subservience and obedience likewise is based on a selective use of the facts. Contrary to Tollefsons allegations, refugee students practice giving as well as following orders. At Level C, students learn to give one-step instructions, and at Level E, they practice explaining a technique or procedure (Center for Applied Linguistics, 1985). Nor does Alien Winds mention that Work Orientation devotes several hours to identifying legal rights and responsibilities in the workplace. In these lessons, refugees learn how to spot errors in paychecks, discuss different types of discrimination from which workers are protected, describe the purpose of unions, and list typical worker benefits (International Catholic Migration Commission, 1987). Because of the educational backgrounds of most of the students in the program, Work Orientation does tend to focus on the language and skills needed for entry-level employment. At the same time, the component has always spent considerable time on the strategies needed to move up on the job or to find a job elsewhere. In a series of lessons, refugee students complete skills inventories and interest and preference checklists, and then examine various occupational options before developing a plan for jobs they can do now, jobs they would like to have in the future, and skills they would have to develop through future education and/or training (International Catholic Migration Commission, 1987, p. 6). These lessons are followed by others that explore educational options. Alien Winds, however, does not give serious attention to this aspect of the curriculum; the author apparently believes that the programs treatment of upward mobility not only betrays an unrealistically rosy picture of economic realities in the U. S., but that it is a purposeful part of the programs intent: to lure refugees into accepting entry-level jobs. If the program expresses a positive attitude towards upward mobility, such an attitude is partly because job mobility is an unfamiliar concept to many Indochinese refugees: In their countries, the first job was often the job one kept for life. In the ORTP, refugees learn that, unlike the situation in their own countries, U.S. workers change jobs frequently and seek additional 532 TESOL QUARTERLY

training in order to improve their lot. (The students get the bad news as well. While refugees learn that it is possible to move up the economic ladder, they also learn that it is possible to move down and to fall off it completely. This is also an unfamiliar notion to many refugees from Southeast Asia, where a personal relationship of responsibility between employer and employee makes layoffs and firings much less frequent than in the U. S.) Tollefson ridicules the programs treatment of upward mobility as naive, arguing that for most refugees upward mobility is more myth than reality. He cites studies showing that several years after their arrival in the U.S. many refugees, even those with jobs, are still living at levels of poverty. Thus, by encouraging refugees to take entry-level jobs, Tollefson argues, we are dooming them to poverty and helping to create an underclass. As elsewhere, we find a selective use of the data. The main evidence for the claim that refugees are suffering a permanent economic crisis and long-term poverty (p. 124) is drawn from a 1985 Office of Refugee Resettlement (ORR)-commissioned study that examines the economic achievements of post-1978 arrivals (Caplan, Whitmore, & Bui, 1985). While the study found high rates of unemployment and poverty in the early period of resettlement, it also found a gradual ascent out of poverty over time. After four years in the U.S., 70% of the refugee sample was above the poverty line, a rate that is not very different from rates for other U.S. minorities (Caplan, Whitmore, & Bui, 1985). In a more recent look at the data, the authors predicted the likelihood of continued economic independence and improvement in economic status and called the refugees climb out of poverty a major accomplishment, particularly considering that it occurred at a time of economic recession, when the percentage of households above the poverty line fell for the U.S. population in general, and for other minority groups in particular (Caplan, Whitmore, & Choy, 1989, p. 65). At the same time, it was found that individual incomes did not improve very much over the four- year period. After a year of employment, refugees were earning an average of $5.20 per hour; after two or more years of employment, they earned an average of $5.35 an hour. (Note, however, that both figures are substantially above minimum wage, Throughout his book, Tollefson misleadingly uses the terms entry-level employment and minimum-wage employment interchangeably.) Economic improvements were made less through individual advancements, the study found, than by increases in the number of people working in households. Another study (Baker & North, THE FORUM 533

1984), however, shows significant increases in individual income among 1975 arrivals. Baker and North found that among young male refugees, the median income rose steadily, approaching parity with U.S. workers in 1979, while the median earnings of female refugees actually surpassed those of U.S. female workers. It is unlikely, given the backgrounds of post-1979 arrivals, that these refugees will achieve the same level of success as the bettereducated 1975 arrivals. Still, there is reason for some optimism. In a recent Wall Street Journal article, one of the authors of the 1985 ORR study says that the economic and educational attainments [of post-1979 arrivals] are stunning. To a surprising degree the boat people have achieved a high level of control over their own destiny (Caplan, 1990, A14). While the sensational successes of individual refugees reported in the media are hardly typical, evidence does not support what Tollefson calls the bleak picture of resettlement (p. 122). The truth is somewhere in between, and it is this truth that the Overseas Refugee Training Program tries to convey. ESL AND PROFESSIONALISM When the program was first launched in 1980, there was more concern for the refugees immediate survival needs than for their long-term language development. There was good reason for this concern. The ORTP was established after President Carters decision to accept for resettlement 14,000 Indochinese a month, more in two months than had been resettled in the two previous years combined (Office of Refugee Resettlement, 1981). Having shifted at jetspeed from their familiar rural Southeast Asian surroundings to cities and suburbs across the U. S., these refugees frequently found themselves in linguistic and cross-cultural confusion. There were stories of Hmong hilltribe people from Laos hosing down the living room floor in their Seattle apartments, then industriously sweeping the water down the heat registers; others attempted to cook a whole chicken in an electric toaster (Levine, 1982). A number of stories were more alarming. Sponsors and case workers in resettlement agencies were concerned that many refugees were unable to understand or follow instructions from doctors, could not read warning and danger signs, and had no idea what to do in case of a household emergency except call the sponsor. There were reports of widespread and serious mental health problems, particularly depression, among refugees (Cohon, 1980; United Community Planning Corporation, 1982; Westermeyer, 1985). In one widely reported incident, a refugee man, 534 TESOL QUARTERLY

unable to understand what was happening to him and fearing that he had failed his family, organized the attempted mass suicide of his entire family (Trillin, 1980). Thus, the overseas program began with a specific but critical purpose: to ease the initial shock of entry into an unfamiliar culture. Over time, as the numbers of refugees decreased and the sense of crisis diminished, the program expanded its scope of purpose to pay more attention to refugees longer range needsin particular, their general language development and strategies for cross-cultural coping. Tollefson is apparently unaware of these developments, or has chosen to ignore them. For example, in his recommendations, Tollefson urges the overseas program to reconsider the competency-based approach to ESL (p. 155) because in his view it limits the education students receive. In fact, in 1985 overseas program staff were concerned that there was not enough attention given to general English language skills, particularly literacy skills (Kharde & Corey, 1986). Their solution was not to throw out the competency-based model, but to reduce the number of required competencies, thus allowing teachers and students more time for other areas. Since 1986, ESL classes in the ORTP have devoted more time to developing students reading and writing skills. For example, students in upperlevel classes read authentic essays and newspaper texts and debate current U.S. social issues. In lower-level classes in the Philippines, many instructors employ a whole language approach, in which students and teachers collaboratively choose the topics for study. In these classes, survival competencies are still covered, however. They come up naturally, teachers report, because they reflect areas of basic concern to refugee adults on their way to the U.S. (Snyder, 1990). The program has changed in many other ways over the years, partly in response to constructive criticism from its own staff, as well as from outside the program. For example, in a 1985 TESOL Quarterly article, Tollefson made a number of suggestions: grouping students by gender and age for specific lessons, mixing students of various ethnic backgrounds for ESL instruction, increasing the number of former refugees on the staff, and increasing the focus on communicative language teaching methods. In fact, very similar measures were already being implemented by the time the article was published. Given the programs commitment to constructive change, it is especially distressing to find in Alien Winds the contention that weaknesses in the programs instruction are intentional. Tollefson, who was once a part of that process of change, now asserts that the THE FORUM 535

Overseas Refugee Training Program is rigid and backward in its ESL instruction, purposely maintaining a U.S. staff that is ineffective and less than competent. Such a staff, he says, is a necessity to the programs ill-intentioned administration, which is committed to restricting refugees access to better jobs. This conspiracy theory of poor instruction is as absurd as it is groundless. Tollefson admits that hiring has recently improved (p. 99) but does not mention that since 1985, the coordinators of the ESL components in the Philippines have had PhDs in relevant fields. In fact, nearly all the U.S. ESL supervisors in the program have masters degrees and relevant overseas experience. The typical ESL supervisor is one who has spent several years overseas as a Peace Corps volunteer, returned to the U.S. for a masters degree in TESL, and after some work with U.S. programs (often refugeerelated), accepts a position with the overseas refugee program. It is simply unfair of Tollefson to characterize these professionals as having virtually no previous experience (p. 99) and to imply that they are motivated by greed. The ORTP has a demonstrated commitment to recruiting and maintaining a professionally qualified staff, emphasizing professional development for all instructors and supervisors: Thais and Filipinos, as well as U.S. staff. For 16 months, Tollefson himself was part of a large training department in the Philippine program site offering sessions on ESL techniques, second language acquisition theory, and dozens of other related topics. His own training sessions on Krashen and Terrells natural approach were documented in an article in the ORTP publication Passage: A Journal of Refugee Education (Wachman, 1985). Since 1986, a regular schedule of university courses has been offered in the Philippines Refugee Processing Center. These courses have been of particular benefit to teachers wishing to pursue masters degrees in linguistics and language teaching. In addition, there are excellent, large collections of professional materials in libraries at training sites in both Thailand and the Philippines. These would be rather peculiar endeavors for a program dedicated to poor ESL instruction. From the beginning of instruction in the ORTP in 1981 until budget cuts took their toll in 1988, the program collected teaching ideas from its staff and published them, first in the form of resource manuals and later in the magazine, Passage. The purpose of these collections was to document the instruction in the program and to share expertise among the training sites and with practitioners in the U.S. In Alien Winds, individual lessons and activities in the resource manuals are quoted as if they were mandated procedures that every teacher is compelled to follow. And Passage articles written by 536 TESOL QUARTERLY

teachers and other staff members are cited as though the authors were articulating policy decisions from the State Department. (Similarly, Tollefson cites curricula or activities intended for only beginning-level students and states or implies that the activities are required of all of the refugees, regardless of their previous education or work experience [p. 79]. ) In this way, Tollefson gives the impression that the ORTP has a central, rigid, lesson-by-lesson syllabus that every teacher must follow. In fact, a reader looking over the resource manuals and issues of Passage would be impressed by the diversity of professional viewpoints encouraged within the instruction in the Overseas Refugee Training Program. Whole language approaches for literacy instruction, cooperative learning, peer teacher coaching, problem posing, and the natural approach, as well as more conventional methods, all have their advocates within the programreflecting the lively diversity of opinions in our field in general. Attempting to show the lack of professionalism in the ESL program, Tollefson states that the Overseas Refugee Training Program is negligent for not having involved more members of TESOLS Executive and Editorial Boards. This kind of involvement is not the function of those groups, and the suggestion can mislead the non-ESL professional who might believe that any large ESL program would normally seek the services of board members. Alien Winds fails to document the ORTPS use of highly respected professionals in the field as consultants. Although their consultancies have necessarily been brief (most for 2 or 3 weeks), many of their training sessions have been videotaped so that staff even years later can benefit from taped lectures, discussions, and demonstration classes. Among the consultants to the processing center programs have been John R. Boyd, Mary Ann Boyd, John L. D. Clark, JoAnn Crandall, Carolyn Graham, Else Hamayan, Wayne Haverson, Michael H. Long, Rebecca L. Oxford, K. Lynn Savage, Lydia Stack, Carole Urza, and Nina Wallerstein. If the goal of the overseas program is to keep the staff uninformed and the ESL classes substandard, as is claimed, it has chosen counterproductive means of doing so. CULTURAL ORIENTATION Although involving students in determining course content is a fairly recent development in ESL classes, this participatory approach has always been a part of Cultural Orientation classes. In a 1985 handbook for teacher trainers in the overseas program (Resnich, 1985), the section dealing with Cultural Orientation THE FORUM 537

strongly encourages teachers to find out from their own students what they want and need to learn. The handbook states that such a needs assessment will generally produce a remarkably clear and insightful statement of what [the students] want to learn about. It also serves as an early indication to them that their participation is encouraged and that their interests and opinions are valued (p. 349). In Needs Assessment and Learner-developed Objectives in Cultural Orientation, an article in Passage, Vernon (1985) describes how teachers can implement this approach. Tollefson fails to recognize the learner-centered aspect of Cultural Orientation, claiming that the program not only ignores refugees interests and opinions, but actually seeks to replace their own beliefs with new ones. According to Alien Winds, one of the many similarities between turn-of-the-century Americanization programs and the ORTP is the shared assumption that refugees must give up their cultural traditions (p. 76) as part of the process of becoming American. Those who are familiar with both the Americanization movement and the ORTP, however, are much more likely to be struck by the differences between the two than their similarities. In the ORTP, native language and culture are regarded as sources of strength and bridges to the new language and culture, rather than as impediments. Staff know that resettled refugees who maintain a connection to their own cultures in the U.S. tend to do better in school and at work, and have fewer mental health problems than those who attempt to sever ties with their past (Rumbaut & Ima, 1987; Caplan, Whitmore, & Choy, 1989). Indeed, the generalization that refugees must maintain (rather than give up) their cultural traditions would be closer to the aims of the ORTP. In teacher training, teacher resources, and instructional practices in numerous Cultural Orientation (CO) lessons on preserving your culture, and in native language literacy classes for nonliterate adult refugees, the ORTP treats the refugees native languages and cultures as sources of strength. The program has made a notably strong effort in this area with younger refugees, who are often the first to reject their own cultures and languages, particularly since many have spent years in refugee camps and lack education about their own heritage. Since 1986, the program for 12- to 16-year-old refugees has included a survey course on the cultures of Indochina, using a curriculum developed by a group of concerned refugees (Lambrecht, 1987). In Cultural Orientation classes, not only the content of instruction but the process itself shows a respect for cultural heritage. A guiding principle in ESL as well as CO instruction has always been that 538 TESOL QUARTERLY

lessons should progress from the familiar to the unfamiliar (Center for Applied Linguistics, 1982). In Cultural Orientation classes, this means that teachers generally address a given topic by asking students about their own cultures. It is only after the refugees own cultural viewpoint has been affirmed that U.S. viewpoints are presented as a comparison and contrast. At the same time that the program encourages refugees to preserve their native languages and cultures, it also recognizes that pressures to conform are real. Refugees learn that although there is an ideal of celebrating cultural diversity in the U. S., this ideal is not embraced by all. (Center for Applied Linguistics, 1982; Hixon, 1987). Thus, many Cultural Orientation lessons are similar to those that new Peace Corps volunteers receive to help them understand how their behavior is likely to be understood in a new culture, what the consequences of that behavior may be, and how they can develop strategies to deal with cross-cultural conflicts. What the program hopes will be the final outcome of this process is a bicultural sophistication, an ideal described in a 1986 Passage article: Traditional beliefs and values are preserved, while new values and practices necessary to function in the new society are acquired. With a bicultural approach, the individual is able to function either in American society or within his or her own ethnic group. This type of adjustment is considered quite compatible with a pluralistic society like the U.S. (Corey, 1986, p. 42) Remarkably, Tollefson cites this article as proof of the programs attempts to divest refugees of their cultures. CONCLUSION The unique constraints under which the Overseas Refugee Training Program operates, together with developments in the fields of ESL and cross-cultural training, the changes in backgrounds among various refugee groups, and a range of political, social, and economic forcesdomestic and internationalhave all affected this training program. A scholarly analysis of their impact would make for a thought-provoking, informative study. In fact, an analysis of the strengths and weaknesses of the training program could be a positive contribution to the programs evolution as well as to the field of refugee education. It is disappointing that Alien Winds proves to be a one-sided polemic rather than a balanced assessment. THE FORUM 539

REFERENCES
Baker, R. P., & North, D. S. (1984). The 1975 refugees: Their first five years in America. Washington, DC: New TransCentury Foundation. Caplan, N. (1990). Boat people prove their worth. The Wall Street Journal, August 1, 1990. p. A14. Caplan, N., Whitmore, J. K., & Bui, Q. L. (1985). Southeast Asian selfsufficiency study (Contract No. HHS-100-81-0064). Washington, DC: U.S. Department of Health and Human Services, Office of Refugee Resettlement. Caplan, N., Whitmore, J. K., & Choy, M. H. (1989). The boat people and achievement in America: A study of family life, hard work, and cultural values. Ann Arbor, MI: The University of Michigan Press. Center for Applied Linguistics. (1982). Cultural orientation resource manual (Vol. 1). Manila, Philippines: Author. Center for Applied Linguistics. (1985). English as a second language, revised competencies. Washington, DC: Author. Cohon, J. D. (1980, March). Can TESOL teachers address the mental health concerns of the Indochinese refugees? Paper presented at the 14th Annual TESOL Convention, San Francisco, CA. Corey, K. (1986). The cultural assimilation of Indochinese refugees. Passage: A ]ournal of Refugee Education, 2 (3), 41-43. Hixon, A. (1987). Examining attitudes and stereotypes through video. Passage: A Journal of Refugee Education, 3 (l), 16-18. International Catholic Migration Commission. (1987). Work orientation (level CDE) curriculum. Morong, Bataan, The Philippines: Author. Kharde, L. S., & Corey, K. (1986). Competencies revisited: Revising the overseas ESL curriculum. Passage: A Journal of Refugee Education, 2 (2), 43-49. Lambrecht, R. (1987). Developing a survey course in Indochinese culture for PASS students. Passage: A Journal of Refugee Education, 3 (l), 23-26. Levine, K. (1982). Becoming American [Videotape]. Seattle, WA: Iris Films and Video. Office of Refugee Resettlement. (1981). Report to the Congress: Refugee resettlement program. Washington, DC: U.S. Department of Health and Human Services. Office of Refugee Resettlement. (1983). Report to the Congress: Refugee resettlement program. Washington, DC: U.S. Department of Health and Human Services. Office of Refugee Resettlement. (1989). Report to the Congress: Refugee resettlement program. Washington, DC: U.S. Department of Health and Human Services. Pihl, C. (1990). Report on Southeast Asian refugee camps. For your information, No. 93, 5-10 [Newsletter]. New York: Lutheran Immigration and Refugee Service. Reznich, C. (1985). Teaching teachers: An introduction to supervision and teacher training. Brattleboro, VT: The Experiment in International Living. 540 TESOL QUARTERLY

Rumbaut, R. G. (1985). Mental health and the refugee experience: A comparative study of Southeast Asian refugees. In T. C. Owan (Ed.), Southeast Asian mental health: Treatment, prevention, services, training, and research (pp. 433-486). Washington, DC: National Institute of Mental Health. Rumbaut, R. G., & Ima, K. (1987). The adaptation of Southeast Asian refugee youth: Comparative study. Washington, DC: Government Printing Office. Snyder, S. (1990, March). ESL literacy: Whats working, why and how. In M. Adkins (Chair), Refugee concerns interest section academic session. Colloquium presented at the 24th Annual TESOL Convention, San Francisco, CA. Tollefson, J. W. (1985, December). Research on refugee resettlement: Implications for instructional programs. TESOL Quarterly, 19 (4), 753-764. Tollefson, J. W. (1989). Alien winds: The reeducation of Americas Indochinese refugees. New York: Praeger. Trillin, C. (1980, March 24). U.S. journal: Fairfield, Iowa. The New Yorker, 56, 83-100. United Community Planning Corporation. (1982). Needs assessment of Southeast Asian refugee population in Massachusetts. Boston: Author. Vernon, A. (1985). Needs assessment and learner-developed objectives in Cultural Orientation. Passage: A Journal of Refugee Education, 1 (2), 60-62. Wachman, R. (1985). A quiet revolution in language teaching at Bataan. Passage: A Journal of Refugee Education, l (l), 57-59. Westermeyer, J. (1985). Mental health of Southeast Asian refugees: Observations over two decades from Laos and the United States. In T. C. Owan (Ed.), Southeast Asian mental health: Treatment, prevention, services, training, and research (pp. 433-486). Washington, DC: National Institute of Mental Health.

The Reader Responds. . .

ELSA AUERBACH
University of Massachusetts at Boston

Although Ranard and Gilzows response is primarily directed toward Tollefsons book rather than my review, I would like to make a few remarks. Regarding the issue of verification of documentation: While the task of the reviewer is to evaluate documentation rather than check facts, I did in fact consult a number of experts about the accuracy of Tollefsons claims. These included refugees themselves (students who had lived in the camps), and Southeast Asia scholars (colleagues
THE FORUM 541

at the University of Massachusetts Joiner Center for the Study of War and its Social Consequences, as well as Chuong Hoang Chung, perhaps the leading Vietnamese researcher on issues of language use and language education for Southeast Asians in the U.S.). In each case, the response was overwhelmingly supportive of claims made in Alien Winds. From the refugees perspectives, the accounts of life in the camps were accurate; the scholars were satisfied not only with the documentation, but with the analysis. My sense is that Ranard and Gilzows real concern is not with documentation (if anything Alien Winds is overdocumented), but with the analysis. Regarding the issue of bias: I readily acknowledge that my review was sympathetic, although not wholly uncritical (in fact, Tollefson thanks me because of my critical reading of an earlier version of his book at the request of a publisher); one responsibility of reviewers is to call attention to work they feel makes a contribution to the field. Moreover, as I argued in the review, we all bring our own ideological biases to our work and this is not negative. The only dishonesty comes when we fail to make these perspectives explicit or when we promote a particular perspective under the guise of objectivity. The tone of Ranard and Gilzows response is testimony to the force of their own bias, indicating that they manifest the same subjectivity of which they accuse Tollefson. In fact, showing that education is a terrain of contestation for different ideologies is precisely one of the contributions of Alien Winds. The kind of critique and countercritique of which this exchange is an example underlies the shifts in paradigm that move the field forward. Finally, I want to address informal feedback I have received to the effect that Alien Winds and my review have caused pain among those who have dedicated years of their lives to improving refugee education. I do not believe it was the aim of the book and it was certainly not the aim of the review to condemn or discredit individual efforts and contributions. Perhaps neither the book nor the review went far enough in exploring the relationship between individual acts and their aggregate impact, between intentions and outcomes, and I certainly regret any pain this may have caused. At the same time, I feel that one message of the book is that as educators we need to examine how our work fits into and contributes to a larger picture. To the extent that Alien Winds has caused this kind of critical self-examination, it has made a contribution.

542

TESOL QUARTERLY

Response to Ranard and Gilzow. . .

The Economics and Ideology of Overseas Refugee Education
JAMES W. TOLLEFSON
University of Washington

In her review of my book about the United States overseas refugee processing centers, Alien Winds: The Reeducation of Americas Indochinese Refugees, Auerbach concluded that it should be required reading for all ESL educators. In response to this review, Ranard and Gilzow, two staff members of the Center for Applied Linguistics who have held posts in the refugee program since the early 1980s, outline their criticisms of the book. Ranard and Gilzows decision to limit their comments to the educational component of the Overseas Refugee Training Program (ORTP) indicates the wide gap between their perspective and mine. The fundamental aim of Alien Winds is to analyze the ORTP within its social and political context. Separating pedagogical matters from political and economic issues or from the circumstances of refugees daily lives in the overseas centers presents an incomplete picture of the ORTP and of the analysis presented in Alien Winds. The two major issues that Ranard and Gilzow ignore are (1) the causes of refugee migration, and (2) the ideology of the ORTP. I will argue that the failure to address these issues fundamentally undermines Ranard and Gilzows position that current U.S. refugee policy should be supported; then I will turn to some of their specific criticisms of my book. REFUGEE MIGRATION In the twentieth century, the migration of people for political and economic reasons has become a permanent feature of the global political economy, with structural roots that encourage, even require migration. Immediately after the Second World War, migration to North America, Australia, and Europe had two main functions (see Sassen-Koob, 1988; Skutnabb-Kangas, 1981). One was to provide a large, cheap pool of labor for the many new jobs being created by the rapid industrial expansion and accumulation of capital. The second function of migration was to provide labor for the most difficult and unpleasant industrial and service jobs, which THE FORUM 543

were no longer being filled by native-born workers (with the exception of certain ethnic minorities such as African Americans in the U.S. and Aboriginal people in Australia). In the view of policy makers in the U.S. and elsewhere, immigrants who did not speak English or other dominant languages were particularly suitable for these purposes because the language barrier that separated them from native-born workers made it difficult for them to gain political rights and to benefit from improving economic conditions enjoyed by the rest of the population (Muller & Espenshade, 1985; Skutnabb-Kangas, 1981). During the 1970s, immigration began to have a new function in North America, Western Europe, and Australia (Jiobu, 1988; Skutnabb-Kangas, 1981). Industrial expansion based on the importation of labor had slowed considerably, and jobs were beginning to be exported from industrial countries to the Third World. This process of economic restructuring increased the probability of boom and bust swings in local economies, which threatened the prosperity of the working and middle classes. Therefore a buffer was neededa group that would absorb periodic increases in unemployment and other consequences of economic restructuring. Immigrants were perfectly suited for this purpose. As a largely disenfranchised and politically weak group, their dissatisfaction could not easily be translated into political action, and therefore, unlike working class and middle class people, they would not threaten the dominant power structure. In order for immigrants to continue to fulfill these functions, they must remain politically and economically marginalized. Alien Winds argues that the ORTP contributes to the labor policy objective of marginalizing migrants in order to maintain the important functions they serve in the changing economyas cheap labor for new industries and as an economic buffer for politically more powerful groups. This goal is expressed in federal language policies declaring the official purpose of refugee education to be the teaching of survival English for entry-level employment (Office of Refugee Resettlement, 1984; also see Haines, 1988), which means English that is sufficient only for marginal employment in the peripheral economy (i. e., in temporary and part-time jobs, which have few benefits and little opportunity for advancement, and which are the first to be eliminated in slow economic times). It is not sufficient to argue, as do Ranard and Gilzow, that entrylevel employment is appropriate for refugees with limited skills. It would be entirely possible to offer, for instance, extended ESL 544 TESOL QUARTERLY

and employment training for refugees. This is not done, however, because refugees with additional skills are not needed in the new economy. For the same reason, those refugees who have technical skills do not receive special education and assistance to help them gain jobs in which they may use their skills. Alien Winds does not argue, as Ranard and Gilzow claim, that a great many refugees are highly skilled, but rather that those who have skills are tracked into the same educational program in the overseas centers as those with little or no previous education, due to the dictates of U.S. migration and labor policy. Similarly, some federal and state agencies have made it difficult for professional refugees to be recertified. As noted in Refugee Reports, medical professionals in particular face an array of bureaucratic barriers to their efforts to work in their professions (Staff, 1988). These practices reflect the overall policy of preparing refugees only for limited occupational categories. This policy is also one reason why domestic ESL programs for refugees are chronically underfunded; restricted funding means that programs can offer only a narrow range of curricular options for relatively short periods. In claiming that I hold staff members in the overseas centers responsible for this situation, Ranard and Gilzow inaccurately summarize Alien Winds. The book consistently argues that policy makers use largescale migration and poorly funded educational programs as mechanisms for achieving labor policy objectives. Teachers are not responsible for these policies. Nevertheless, it is crucial that those of us who are ESL professionals examine the function of ESL programs within a political-economic system that creates and sustains massive migration of Southeast Asian refugees and other groups. The fact that Ranard and Gilzow, as well as other supporters of current refugee policies, do not confront the reasons for migration and for current low funding levels for educational programs, perpetuates contemporary ideologies concerning refugees. This is the second fundamental issue addressed in Alien Winds. IDEOLOGY IN THE ORTP ldeology refers to (often unconscious) assumptions about the world that come to be seen as common sense, and thus are typically not the focus of critical discussion and debate. (For a more detailed discussion of ideology, see Giddens, 1987; Tollefson, 1991). Assumptions that become widely accepted as common sense tend to sustain existing power relationships. As ideologies become institutionalized, they tend to reinforce privilege and grant that THE FORUM 545

privilege legitimacy as a natural condition of society. For instance, the policy of requiring everyone to learn a single dominant language is widely seen as a commonsense solution to linguistic inequality. The argument is simple: If refugees and other linguistic minorities learn English, they will not suffer economically and politically. This view grants privilege to those who speak English, it ignores the economic and political forces that deny adequate language education to refugees and immigrants, and it excludes language from the list of structural categories (such as race and gender) that are protected by legislation against discrimination. Similarly, the argument that entry-level jobs are appropriate for refugees precludes increased funding for educational programs, ignores U.S. responsibility for creating and sustaining refugee movements, and provides a rationale for blaming refugees for their plight: When they fail to acquire English despite refugee language programs, or when they fail to get better jobs despite learning the language to a degree defined as satisfactory by policy makers, then they can be held responsible for their own economic circumstances. Thus Alien Winds examines numerous examples of policy makers blaming refugees and the voluntary agencies that assist them for their high rates of unemployment and use of public assistance. This perspective was successfully used during the Reagan administration as a rationale for reducing funding for refugee education and public assistance. Alien Winds argues that the ORTP takes an ideological stance that helps to sustain existing economic inequalities by insisting that it offers refugees a mechanism for upward mobility (despite evidence to the contrary, discussed below). The effect of Ranard and Gilzows claim that refugees who complete the program are climb[ing] out of poverty is to support this ideology. Their commentary perpetuates this ideology in other ways as well. For instance, they repeat the official claim that the purpose of the ORTP is to meet the needs of refugees rather than of the U.S. economy. They support efforts in the ORTP to teach refugees the meaning and value of job mobility, which is often a euphemism for the pattern of employment and unemployment experienced by individuals in the peripheral economy. And they depict the ORTP as a benevolent system designed to help refugees, rather than as part of a larger political-economic system that displaces them from their homes and then provides education suitable only for long-term peripheral employment. Because they do not examine the social and political context of the educational program, Ranard and Gilzow fail to address the 546 TESOL QUARTERLY

central issue in Alien Winds, which is not a conspiracy, but rather the relationship of the educational program to migration and labor policy. In order to understand this relationship, we must ask: Why are there refugees? Whose interests encourage refugee movements in Southeast Asia and refugee resettlement in the U. S.? How does the educational program serve those interests? What is the public image of the program and how is that image created and maintained? What ideological assumptions underlie the content of the educational program? These are the questions Alien Winds seeks to answer. In doing so, the book argues that it is misleading and pedagogically ineffective to ignore the political and economic roots of refugee movements and to seek educational solutions to problems that are fundamentally economic and political. Thus Alien Winds shows that the ORTP is driven by a U.S. foreign policy that creates and sustains large-scale refugee movements in Southeast Asia, and by a labor/migration policy that channels refugees and immigrants into poorly paid jobs in the peripheral economy. Nowhere does Alien Winds suggest a conspiracy of officials in the educational program. In fact, Alien Winds argues exactly the oppositethat the policies and programs adopted for refugees result from institutional structures and ideologies rather than from the preferences of people employed in the ORTP or the interests of refugees. A full analysis of the ideology of the ORTP, presented in Alien Winds, is beyond the scope of this forum. However, an example of its impact on the language used to describe resettlement may be useful. Ranard and Gilzow correctly note that Alien Winds does not always clearly distinguish entry-level from minimum-wage jobs. Yet they ignore the more fundamental issue: Entry level is an ideological term implying upward mobility through the image of an economic system which, once entered, will steadily lead to improved economic circumstances. But in fact certain minorities in the U. S., including many refugees, permanently hold low-paid jobs. For them, entry level is a euphemism for poorly paid. The question of whether entry-level jobs provide an initial step toward improving refugees economic circumstances is the first of several specific disagreements with Alien Winds that Ranard and Gilzow outline. Due to space limitations, I will only deal with their four major criticisms of Alien Winds Ranard and Gilzow claim that the book inaccurately portrays refugees economic circumstances in the U. S.; misleadingly depicts U.S. control and conditions in the overseas camps; misstates the purpose of the ORTP; and unfairly describes the professional role of ORTP staff. THE FORUM 547

ECONOMIC CIRCUMSTANCES OF SOUTHEAST ASIAN REFUGEES IN THE U.S. Ranard and Gilzow dispute the bleak picture of resettlement in which many refugee households suffer long-term economic crisis because members are trapped in dead-end jobs in the peripheral economy. As evidence that refugees are climbing out of poverty in impressive numbers, Ranard and Gilzow cite Caplan, Whitmore, and Choy (1989). This study does indeed report that refugees poverty rate gradually decreases over time. But the study, based upon interviews and questionnaires given to 1,384 refugee households at five locations in the U.S. in 1981 and 1984, also reports other key findings: 1. Most households with incomes above the poverty line ($800 per month for a family of four at the time of the study) reached this level through a combination of wages from two or more adults and cash assistance and other forms of public assistance payments. Over 50% of the adults lived in groups of extended families and unrelated individuals who pooled resources to survive. Only 25% of the households received no form of public assistance. 2. Over 42% of the employable adult refugees were unemployed. 3. Those who were employed improved their average salary by a total of less than 3% over the 4-year period of the study. As Caplan, Whitmore, and Choy point out, In such a limited job market as the refugees found themselves, individual initiative to advance could not move the household ahead economically, in either a comparative or an absolute sense (p. 55). 4. Employed refugees were overwhelmingly in low-paying jobs in the peripheral economy that offered very little opportunity for economic improvement. Caplan, Whitmore, and Choy conclude: As defined by SEI [Socio-Economic Index] scores, the overwhelming majority (71 percent) of those refugees in the labor force held low-status jobs. Slightly more than one-half (55 percent) were also employed in the periphery of the economy rather than in the core economic sector (45 percent). Thus in the main, the refugees tended to hold low-level, lowpaying, dead-end jobs. (pp. 55-56) Alien Winds cites 13 other studies of refugee resettlement that support similar conclusions about refugees economic circumstances in the U.S. Ranard and Gilzows argument that refugees are clirnb[ing] out of poverty, as well as their optimistic view of Caplan, Whitmore, 548 TESOL QUARTERLY

and Choys research, is an example of an unstated ideology. As evidence of refugees economic progress, Ranard and Gilzow state that the 30% rate of poverty among refugees in the U.S. more than four years (found by Caplan, Whitmore, and Bui [1985]) is not very different from rates for other U.S. minorities. In fact, Caplan, Whitmore, and Bui noted that a 30% poverty rate is roughly equal not to minorities generally, but specifically to the rate for African Americans and Latinos, two groups who have served for many years as cheap labor and as economic buffers for politically more powerful groups (see Jaynes and Williams, 1989; Jiobu, 1988). Ranard and Gilzows argument is that a comparison of Southeast Asian and African American/Latino poverty rates is appropriate, and that an equivalent poverty rate for these groups is an indication of moderately successful refugee resettlement. The belief that refugees are doing well when they live in large groups with multiple wage earners and others who share public assistance payments does not apply to members of dominant groups in the U.S. This is precisely the point of Alien Winds: Policy and ideology underlying the ORTP ensure that refugees serve the same economic functions as African Americans and Latinos. U.S. CONTROL AND CONDITIONS IN THE OVERSEAS CENTERS Control of the Overseas Centers Ranard and Gilzow state that the processing centers are operated by host country governments (the Philippines and Thailand) and the United Nations, rather than by the U.S. Indeed, the organizational charts for the centers list UN and local officials as operational directors, a system established when the centers were first created in 1979-80 as international holding centers for refugees awaiting resettlement in many countries. Since that time, however, the main Philippine center has become overwhelmingly a U.S. operation, with tiny programs for a few refugees to be resettled elsewhere vastly outnumbered by the program serving up to 20,000 refugees bound for the U.S. As a result of their overwhelming dominance in financial and staffing matters, U.S. agencies and officials have come to control camp policy. The fiction that the centers are not U.S. run is maintained, however, in part because it helps to obscure funding sources and to provide a mechanism for U.S. officials to deny responsibility for what happens in them. In a 1989 report to Congress on its visit to the Philippine center, the Vietnam Veterans of America Foundation (VVAF) pointed out that the cost of running the PRPC [Philippine Refugee Processing Center] is obscured through the fiction that the camp is an THE FORUM 549

international camp (p. vii). The VVAF called for a Congressional investigation into the funding of the Philippine center and whether U.S. officials are using the requirement that refugees spend 6 months in the center as a mechanism for keeping refugee admissions below Congressionally authorized levels. Conditions in the Overseas Centers Apparently believing that conditions in the U.S. centers are satisfactory, Ranard and Gilzow quote a visitor who called them country-clubs compared to first-asylum camps. And certainly the conditions in the U.S. centers are better than those in the Hong Kong prisons and the Thai-Cambodian border camps, where refugees denied resettlement are subject to particularly brutal treatment. All supporters of human rights should express outrage over the mistreatment and denial of basic standards of human decency, which are institutionalized in those locations (see Amnesty International, 1990; Bui, 1990; U.S. Committee for Refugees, 1985, 1986, 1987). Yet the brutal treatment accorded refugees elsewhere does not absolve U.S. officials of responsibility for providing humane conditions in the U.S. processing centers. Although Ranard and Gilzow claim that Alien Winds does not take into account recent changes, current analyses confirm that conditions remain unacceptable. In the report on its 1989 visit to the Philippine center, the VVAF pointed out that refugees live in atrocious conditions, where there is insufficient food and water, where they are crowded into billets constructed of asbestos, with people unrelated or [un]known to them, and where their daily lives are regulated by coercion and fear (Vietnam Veterans of America Foundation, 1989, p. vi). Macdonald (1990) contrasts the relatively good living conditions for staff with those for refugees, and describes poor sanitation, shortages of water and food, overcrowding and lack of privacy in refugees living areas. In her description of the process of coercion to which refugees are subject in the Philippine camp, Mortland (1987) points out that administrators believe refugees fortunate to be there rather than in the first-asylum camps, and that they should therefore follow precisely the detailed rules and procedures prescribed for them. Ranard and Gilzows commentary participates in the (recreation of this ideology, and does not address the unsanitary and unsafe conditions, the atmosphere of coercion and fear in which refugees must live, the pervasive denial of human rights, and the failure of camp officials to rectify these conditions. 550 TESOL QUARTERLY

THE PURPOSE OF REFUGEE EDUCATION Ranard and Gilzow criticize my claim that the ORTP seeks to transform refugees identities. However, this is the explicit goal repeated to refugees in program materials and by program officials. For instance, the administration building in the Philippine center includes a display board that states: Refugee transformation, the primary goal of the PRPC operations, is achieved through a psycho-social recuperative process involving the critical phases of adaptation, capability building, and disengagement which result in changing a displaced person into an individual wellequipped for life in his country of final destination. Based upon her anthropological study of the Philippine ORTP, Mortland (1987) concluded that the central myth at the [Philippine] processing center is that when refugees finish their stay, they have been transformedthat they will go to the new country and become Americansthat the transformation process will allow them to be successful in the promised land (p. 400). Similarly, the VVAF report concluded that the pervasive philosophy of the PRPC [Philippine Refugee Processing Center] is clear: Indochinese refugees need to be transformed in order to survive in America (Vietnam Veterans of America Foundation, 1989, p. 20; also see Knudsen, 1983). In its claim that the ORTP seeks to transform refugees, Alien Winds is not controversial. The important issue is the impact of Ranard and Gilzows denial of what is obvious to observers and to refugees. The acknowledgment of the ideology of Americanization that underlies the ORTP is the first step toward public discussion and debate of this ideology. Is the current approach to refugee education the most effective, given refugees long-term economic problems in the U. S.? Does the ORTP serve refugees interests? What other approaches might be considered, besides transforming refugees? In addressing these questions, Alien Winds argues that the ORTP ideology of Americanization does not serve refugees long-term economic, cultural, or political interests. THE ROLE OF PROFESSIONALS IN THE OVERSEAS CENTERS Ranard and Gilzow claim that Alien Winds presents a conspiracy theory of poor instruction. Nowhere does Alien Winds present such a view. Rather, its critique of the ORTP staff focuses on the ideology of camp life among the expatriate U.S. administration. In an analysis of the role of expatriate administrators in refugee THE FORUM 551

assistance programs, Cromwell (1988) examines the hierarchy of refugee camps, where expatriates are in charge of host country staff, who themselves are in charge of refugees. By virtue of their high status within the camp, expatriates develop an intense loyalty not to the host country or to the refugees, but instead to the expatriate administration and to an expatriate peer group ideology (p. 299), which views host country nationals and refugees alike as inefficient, backward, ignorant, and corrupt. Cromwell argues that this implicit and unstated ideology blocks initiative and critical analysis by expatriate staff. Alien Winds describes U.S. refugee camps as company towns in which staff members are isolated far from home in an atmosphere of conformity and, for those who may disagree with current practices, the constant threat of isolation. This atmosphere is sustained in part by a rhetoric of diversity, which claims that professional debate is welcomed. Indeed, this is the picture that Ranard and Gilzow present of professional life in the camps. Yet, as McDonald has pointed out in her analysis of the PASS (Preparation for American Secondary Schools) program in the Philippine center: Staff with the confidence to ask questions, or to question policy are generally labeled as trouble makers, in the best bureaucratic tradition, and suffer from intense pressure to conform. The rigid hierarchical structure is a particularly effective device for preventing change from below. (p. 15) Analyses of the ORTP simply do not support Ranard and Gilzows claim that the ORTP has been flexible, innovative, and effective in its educational administration, curriculum, and prescribed teaching practices. CONCLUSION Ranard and Gilzow depict Alien Winds as a distorted view of the ORTP, a one-sided polemic rather than a scholarly analysis. Apart from a federally funded study (RMC Research Corporation, 1984), which found no evidence that the ORTP improves refugees employability in the U. S., there are four other independent analyses of the overseas centers: Knudsen (1983), Mortland (1987), Vietnam Veterans of America Foundation (1989), and Macdonald (1990). Within the context of these analyses, all of which document serious problems in the ORTP, Alien Winds presents mainstream views (for a review of related studies, see Tollefson, 1989). Ranard and Gilzows comments provide no evidence justifying continued support of current policies, which create and sustain 552 TESOL QUARTERLY

refugee movements and lead to long-term economic crisis for hundreds of thousands of refugee households in the U.S. In its failure to examine the full context and impact of the ORTP, Ranard and Gilzows commentary ignores virtually all of Alien Winds, including: its historical analysis of refugee movements and U.S. refugee policy in Southeast Asia since 1954; its detailed examination of the ideology of refugee education; its analysis of U.S. immigrant education since 1880; its examination of the political interests of agencies responsible for refugee resettlement and education; its description of human rights violations in the. U.S. centers, such as arrest and imprisonment of refugees without the right to confront their accusers or to be presumed innocent until proven guilty; and its account of serious health and safety problems, including the continued use of asbestos for walls and roofs in refugee housing and classrooms. Approximately 20,000 refugees remain at the U.S. centers in Thailand and the Philippines. They continue to live in deplorable conditions and to attend an educational program whose purpose is determined by labor/migration policies requiring that refugees be channeled into low-paying jobs in the peripheral economy. These statements do not deny the remarkable individual efforts of staff members in the camps who seek to provide effective instruction within an educational administration that blocks most professional discussion and debate. The solution to the continuing economic, social, and personal challenges refugees face in the U.S. depends upon major shifts in U.S. foreign policy in Southeast Asia, changes in labor and migration policy in the U. S., and increased support for programs that provide refugees and other migrants with the language, education, and employment skills they need. The political effort to change U.S. refugee education policy continues. The foundation for this effort must be a clear-sighted analysis of the causes and consequences of current policies for the lives of nearly 1 million resettled Indochinese living in the United States today.

REFERENCES
Amnesty International. (1990). Memorandum to the Governments of Hong Kong and the United Kingdom regarding the protection of Vietnamese asylum seekers in Hong Kong. New York: Author. Bui, D. D. (1990). Hong Kongthe other story: The situation of Vietnamese women and children in Hong Kongs detention centres. Washington, DC: Indochina Resource Action Center.

THE FORUM

553

Caplan, N., Whitmore, J. K., & Bui, Q. L. (1985). Southeast Asian refugee self-sufficiency study (Report prepared by the Institute for Social Research, University of Michigan). Washington, DC: U.S. Department of Health and Human Services, Office of Refugee Resettlement. Caplan, N., Whitmore, J. K., & Choy, M. H. (1989). The boat people and achievement in America. Ann Arbor: University of Michigan Press. Cromwell, G. (1988). Note on the role of expatriate administration in agency-assisted refugee programmed. Journal of Refugee Studies, 1, 297-307. Giddens, A. (1987). Social theory and modern sociology. Stanford, CA: Stanford University Press. Haines, D. W. (1988). The pursuit of English and self-sufficiency: Dilemmas in assessing refugee programme effects. Journal of Refugee Studies, 1, 195-213. Jaynes, G. D., & Williams, R. M. (Eds.). (1989). A common destiny: Blacks and American society. Washington, DC: National Academy Press. Jiobu, R. M. (1988). Ethnicity and assimilation. Albany: State University of New York Press. Knudsen, J. C. (1983). Boat people in transit: Vietnamese in refugee camps in the Philippines, Hongkong and Japan (Occasional Paper No. 31, Migration Studies Project). Bergen, Norway: University of Bergen. Macdonald, J. (1990). Almost freedom, almost American: An ethnographic study of the Philippine Refugee Processing Center. U n p u b l i s h e d doctoral dissertation, Teachers College, Columbia University, New York. Mortland, C. A. (1987). Transforming refugees in refugee camps. Urban Anthropology, 16, 375-404. Muller, T., & Espenshade, T. J. (1985). The fourth wave: Californias newest immigrants. Washington, DC: Urban Institute Press. Office of Refugee Resettlement. (1984). Statement of program goals, priorities and standards for state administered refugee resettlement programs. Kansas City, MO: Author. RMC Research Corporation. (1984). The effects of pre-entry training on the resettlement of Indochinese refugees (Report prepared for the U.S. Department of State, Bureau for Refugee Programs). Hampton, NH: Author. Sassen-Koob, S. (1988). The new labour demand: Conditions for the absorption of immigrant workers in the United States. In C. Stahl (Ed.), International migration today, (Vol. 2, pp. 81-104). Paris: UNESCO. Skutnabb-Kangas, T. (1981). Bilingualism or not: The education of minorities. Clevedon, Avon, England: Multilingual Matters. Staff. (1988, January 22). Hurdles bar path to continuing medical practice for refugee physicians. Refugee Reports, pp. 1-7. Tollefson, J. W. (1989). Educating for employment in programs for Southeast Asian refugees: A review of research. TESOL Quarterly, 23 (2), 337-343. 554 TESOL QUARTERLY

Tollefson, J. W. (1991). Planning language, planning inequality: Language policy in the community. London: Longman. U.S. Committee for Refugees. (1985). Cambodians in Thailand: People on the edge. Washington, DC: American Council for Nationalities Service. U.S. Committee for Refugees. (1986). Refugees from Laos: In harms way. Washington, DC: American Council for Nationalities Service. U.S. Committee for Refugees. (1987). Uncertain harbors: The plight of Vietnamese boat people. Washington, DC: American Council for Nationalities Service. Vietnam Veterans of America Foundation. (1989). Report on the Amerasian issue. Washington, DC: Author.

Comments on Martha C. Bennington and Aileen L. Youngs Approacbes to Faculty Evaluation for ESL
A Reader Reacts. . .
ALASTAIR PENNYCOOK
Ontario Institute for Studies in Education

I have just received a letter informing me that my employers will be doing an evaluation of [my] class on . . . the second to last class meeting. Someone (whom I do not know) will, I am told, come to [my] class within the first 15 minutes of the start time. He or she will hand out the evaluation slips to [my] students, wait until they are completed, collect the slips and then return them to our office. Fortunately, I am assured that I will receive a report of the evaluation in due course, although I have so far heard nothing from last semesters evaluation. I have been given no curriculum, no set materials, nothing beyond some advice on books I might want to use and the time and location of the classes, and yet I gather the outcome of this evaluation will play a major role in deciding whether I will be rehired next semester. This, I suggest, will ring a familiar bell with many other practicing ESL teachers. The most important question that I wish to raise here is whether, in light of my current situation, I should welcome the recent article by Martha Pennington and Aileen Young (Vol. .23, No. 4, December 1989) with its numerous suggestions for improving ESL faculty evaluation. Despite the fact that the wider range of options that these researchers offer might improve the type of evaluation to
THE FORUM 555

which I am to be subjected, I feel that I must ultimately take exception to their article. While they commendably draw on a broader educational literature than is usually the case for ESL research, they nevertheless fail to explore still broader but more significant questions concerning the politics of education and evaluation. In this brief response to their article, then, I would like to broaden the discussion of teacher evaluation to include issues that are crucially absent from their article. I would also like to try to locate this discussion in a yet broader context. Editors, operating within difficult constraints, work in mysterious waysby juxtaposition. Stephen Gaies placement of Peirces (1989) article close to Davies (1989), and Sandra Silbersteins placement of my own (Pennycook, 1989) next to Pennington and Youngs point to an important division within applied linguistics, which I think needs emphasis. First, however, Pennington and Youngs article. Most disconcerting is the lack of discussion of why we are being evaluated, and who is evaluating whom. The overall concern of the authors appears to be to describe methods for teacher evaluation (p. 619) in order to further the goals of the profession (p. 643). To the extent that the authors leave unexamined questions concerning the development of yet more methods (in this instance for evaluation), the power relationship between evaluator and evaluated, the unspecified goals of evaluation, and the implications of an appeal to the notion of the profession, this description of methods for evaluation runs the risk of becoming reactionary. As I argued in my own article, there has been a growing incursion of technical rationality into all domains of human investigation, an incursion that not only limits the possibilities of other modes of thought, but also has serious implications in terms of social control and regulation. Foucault (1979) has greatly helped our understanding hereof how, in evaluation, modes of societal surveillance combine the techniques of an observing hierarchy and those of normalizing judgment. The belief that improvement can be brought about by the correct application of rational organization is what Marcuse (1964) came to criticize as one dimensional. It is a view that disregards all notions of the political in social life. While my own discussion centred on the implications of the construction and prescription of a concept of Method, my argument that we have witnessed a deskilling of teachers and greater institutional control over classroom practice may be more pertinent to the question of teacher evaluation. It is essential that we explore the cultural politics of teacher education, for, as Popkewitz (1987) puts it, the behaviors, patterns of language, and actions used in 556 TESOL QUARTERLY

teacher education contain codes of culture that have implications for fundamental issues of power in American society (p. 26). Historical analyses of teaching (e.g., Apple, 1986) have clearly shown that, as teaching moved from a predominantly male to a predominantly female occupation, and as teacher education became ensconsed in universities and colleges, an ever greater degree of control came to be exerted over the growing body of women teachers. Evaluation has started to play an increasingly important role in that control, especially within the context of the conservative cries for standardized curricula, accountability, and more educational responsiveness to market forces. Popkewitz (1984) argues that unless we place evaluation within an adequate political theory of context, evaluation remains solely a symbolic canopy that legitimates occupational and institutional authority and control (p. 179). Pennington and Youngs reliance on a normative concept of the goals of the profession (p. 643) also requires comment. It is important to see the concept of professionalism within both the overall context of the increasing specialization and fragmentation of modern, industrial life, and the conservative call for accountability. The ideology of professionalism in teaching constitutes a means of masking the structural basis of class, race, gender, and other inequalities in our society (Ginsberg, 1987). As Densmore (1987) argues, we must challenge any teacher education that encourages the ideology of professionalism. I am not arguing against all teacher evaluation, though I would like to see a greater emphasis on teacher autonomy and greater encouragement of teachers own explorations of the cultural politics of their classrooms and schools. But I think it is dangerous to move towards developing more and better methods of teacher evaluation without exploring the implications. I think Zeichner (1983) has expressed this most usefully: It is hoped that future debate in teacher education will be more concerned with the question of which educational, moral and political commitments ought to guide our work in the field rather than with the practice of merely dwelling on which procedures and organizational arrangements will most effectively help us realize tacit and often unexamined ends. Only after we have begun to resolve some of these necessarily prior questions related to ends should we concentrate on the resolution of more instrumental issues related to effectively accomplishing our goals. (p. 8) Finally I would like to suggest the connections I see between, on the one hand, Davies (1989) and Pennington and Young, and, on the other, Peirce (1989) and my own work (Pennycook, 1989). I do not intend to enter the structuralist/poststructuralist debate because I THE FORUM 557

feel Peirces (1990) own reply to Dubois challenge in the TESOL Quarterly Forum Section has more than adequately covered this ground. Rather, I would like to suggest that what Davies and Pennington and Youngs work shares is an apparent acceptance of society as it is. Neither article problematizes its area of study in social or political terms. One leaves us with methods to make teacher evaluation more effective without questioning the goals and politics of evaluation; the other compares English as an international language with interlanguage (why?) without raising a vast range of cultural and political issues in the global spread of English. Furthermore, both articles operate with many of the standard assumptions of modernism: the stress on evaluation and efficiency, and the dichotomies between language and culture, universality and relativism, and society and the psychological/individual. They seem reluctant to admit to constraints on the rational unity of the individual, implying that we can all make free choices unclouded by societal, political, cultural, or ideological conditions. To suggest that such assumptions are no longer acceptable may seem harsh since the predominant positivist paradigm of applied linguistics has allowed no space for such issues. What Peirce and I have been trying to do, however, through our use of poststructuralist and postmodernist views of language, discourse, and knowledge, is to show the reactionary implications of concepts such as communicative competence and methods, and to find new ways of exploring questions around language teaching that allow us to develop ethical and political stances in our work that reflect our views on an inequitably structured world.

REFERENCES
Apple, M. (1986). Teachers and texts: A political economy of class and gender relations in education. New York Routledge & Kegan Paul. Davies, A. (1989). Is international English an interlanguage? TESOL Quarterly, 23 (3), 447-467. Densmore, K. (1987). Professionalism, proletarianization and teacher work. In T. S. Popkewitz (Ed.), Critical studies in teacher education: Its folklore, theory and practice (pp. 130-160). London: The Falmer Press. Foucault, M. (1979). Discipline and punish: The birth of the prison. N e w York: Vintage.

558

TESOL QUARTERLY

Ginsburg, M. (1987). Reproduction, contradiction and conceptions of professionalism: The case of pre-service teachers. In T. S. Popkewitz (Ed.), Critical studies in teacher education: Its folklore, theory and practice (pp. 86-129). London: The Falmer Press. Marcuse, H. (1964). One-dimensional man. Boston: Beacon Press. Peirce, B. N. (1989). Toward a pedagogy of possibility in the teaching of English internationally. TESOL Quarterly, 23 (3), 401-420. Peirce, B. N. (1990). Comments on Toward a pedagogy of possibility in the teaching of English internationally: Peoples English in South Africa: The author responds. TESOL Quarterly, 24 (l), 105-111. Pennycook, A. (1989). The concept of method, interested knowledge, and the politics of language teaching. TESOL Quarterly, 23 (4), 589-618. Popkewitz, T. S. (1984). Paradigm and ideology in educational research. London: The Falmer Press. Popkewitz, T. S. (1987). Ideology and social formation in teacher education. In T. S. Popkewitz (Ed.), Critical studies in teacher education: Its folklore, theory and practice (pp. 2-33). London: The Falmer Press. Zeichner, K. (1983). Alternative paradigms of teacher education. Journal of Teacher Education, 34 (l), 3-9.

Response to Pennycook. . . The Political Economy of Information in TESOL

MARTHA C. PENNINGTON
University of Hawaii at Manoa

In The Political Economy of Information, Schiller (1988) argues that the value of information derives not from its inherent attributes as a resource, but rather stems uniquely from its transformation into a commoditya resource socially revalued and redefined through progressive historical application of wage labor and the market to its production and exchange (p. 41). Through this capitalizing process, not only English, but also discourse about English, about the teaching of English, and about a variety of attendant matters involving teachers and learners have become salable commodities whose value is determined by economic forces such as the law of supply and demand. As in the case of other commodities; scarcity or inaccessibility of information to the average consumer drives up its value and, by projection, the value of anyone who possesses it. Moreover, as all those who work on
THE FORUM 559

Madison Avenue are aware, ones stock can also be enhanced by trading in a well-promoted commodity that accrues a certain glamour value as a result of its perception as chic or stylish. According to this economic line of reasoning, one can assume that Pennycooks fashionably postmodern response to Pennington and Young will be accorded a reasonable market value by all those scholars who perceive his ideas to be new or unique. However attractive or important those ideas may be, Pennycook errs in assuming that he is in possession of all of the valid information and the only (politically correct) point of view on matters that he addresses. These assumptions cause him to draw unwarranted conclusions and to state his case quite boldly: In his response, he employs the word dangerously in connection with the Pennington and Young article on evaluation, implying that our work could be damaging to the field of TESOL. Considering this highly negative implication of Pennycooks response, I feel that it is imperative for me to clarify our intentions and to respond to his charges point by point. I take exception to Pennycooks inference that Aileen Young and I are on the wrong side of an important division within applied linguistics, that we accept many of the standard assumptions of modernism: the stress on evaluation and efficiency, and the dichotomies between language and culture, universality and relativism, and society and the psychological/individual, and are reluctant to admit to restraints on the rational unity of the individual, on the ability of the individual to make free choices unclouded by societal, political, cultural or ideological conditions. I do not think that Pennycook can infer our acceptance of certain assumptions of modernism or our reluctance to admit the restraints placed on individuals based on our text. Though these concepts and issues were not in fact the terms of the discussion in the original TESOL Quarterly article, I welcome the chance to expound upon them as the terms of Pennycooks response. First, it seems that Pennycook takes a much narrower view of evaluation than we have taken in our article, conceptualizing evaluation as a form of top-down social control of teachers by outside authorities. While this is a common view of what the term evaluation means in an educational contextand indeed, the view that we expected many readers of our article to have at the outset we clearly were trying to broaden this perspective to one that incorporates peer review and self-evaluation and in which evaluation is seen as an essentially formative and long-term process of developing the potentials of individual teachers. We are strong advocates of reflective practice and, as we stated in our article, 560 TESOL QUARTERLY

believe that training in self-evaluation should be considered an essential component of professional education that will help to ensure the long-term career development of confident and responsible faculty members who are able and willing to (a) evaluate input on their professional skills and behavior, and (b) expand competencies and alter teaching approach as circumstances dictate (p. 640; for further elaboration, see Pennington, 1990a). Secondly, it is not accurate to claim that there is a lack of discussion of why we are being evaluated, and who is evaluating whom. The general purposes of faculty evaluation are addressed at the beginning and the end of the article, and the specific purposes of each form of evaluation reviewed are explicitly addressed. There is, moreover, considerable discussion of who is, and in our view ought to be, evaluating whom. To clarify this point, we strongly advocate an evaluation system for every language program in which teachers are themselves centrally involved in developing and implementing the standards, the criteria, and the mechanisms by which evaluation is conducted. Indeed, I would myself claim, on the basis of my own experience and the related experience described by Sashkin (1986), that teacher participation in all aspects of the management of a language program is essential not only to the health of the program, but also, quite literally, to the psychological and physical health of the teachers themselves. I would therefore go beyond a recommendation of teacher participation to an insistence on such participation, following Sashkin (1986), as an ethical imperative. At the same time, in my view it must be recognized that the faculty, though central to the functioning of a language program, is not the only constituency whose values must be considered in assessing a programs degree of success in achieving its goals: In evaluating the worth or success of a language program, value must be defined relative to the needs and desires of all of the groups who make up or interact with the program. These groups include the administration of the school or other body in which the program is housed; the programs own administration; its faculty its students; and the parents, sponsors, and external agencies which are concerned with the success of the program and its students. (Brown & Pennington, in press) Where a successful evaluation system is defined as one that effects positive change, the degree of involvement in the evaluation process of parties affected by it will in large measure determine the success of the evaluation system (Brown & Pennington, in press). As noted by Darling-Hammond, Wise, and Pease (1983): Effective change requires a process of mutual adaptation in which [participants] at all levels can shape policies to meet their needsone in THE FORUM 561

which both the participants and the policy are transformed by the convergence of internal and external reference points (p. 17). In his paper on method, Pennycook (1989) argues that there is no such thing as disinterested information and that education is always situated in a social and political context. As argued by Schiller (1988, p. 41), information itself is conditioned and structured by the social institutions and relations in which it is embedded. Moreover, as noted by Schaef and Fassel (1988): Full personal participation results in a totally different kind of knowledge and information from that which is gathered abstractly and objectively by someone else. This kind of information in turn affects the organization differently from the information of nonparticipatory management. (p. 16) For these reasons, I have long been of the opinion that it is critical for teachers to take active responsibility for the development of the information, and the social institutions and relations within which it is embedded, that underlie all aspects of the educational process from designing curriculum and materials to deciding the terms of evaluation of students, of language programs, and of the teachers who work in them (see Pennington, 1989a, 1989b; Brown & Pennington, in press). Thus, Pennycook and I are in complete agreement with Girouxs vision of teachers as transformative intellectuals, that is, as Pennycook (1989, p. 613) quotes Giroux and McLaren (1989, p. xxiii): as professionals who are able and willing to reflect upon the ideological principles that inform [our] practice, who connect pedagogical theory and practice to wider social issues, and who work together to share ideas, exercise power over the conditions of [our] labor, and embody in [our] teaching a vision of a better and more humane life. Contrary to Pennycooks implication, the Pennington and Young article did not in fact stress evaluation and efficiency; nowhere did we state or imply that the purpose of faculty evaluation was or should be tied to efficiency. In fact, I personally believe that efficiency is overemphasized by many ESL administrators, who focus too much on completing paperwork and not enough on the much more important but inefficient interpersonal side of their job (Pennington, 1985). For example, Reasor (1981) found that most ESL administrators assessed their own style as separatedan inappropriately isolative orientation that I believe stems from insecurity and lack of knowledge about how to function as a program director. Unfortunately, some ESL administrators assume a bureaucratic rather than a facilitative role for themselves: rather than focusing on the welfare of their employees, they focus on 562 TESOL QUARTERLY

getting the job done; rather than focusing on building an organization, they focus on the bottom line. Disturbingly, when former teachers become program directors, the learner-centered orientation for managing a classroom does not always transfer to an employee-centered orientation for managing a program. The suggestion by Pennycook of TESOL as a field dominated by male administrators controlling female instructors may be less the case than it was in the past, at least in the United States. While over two thirds of the respondents to two recent surveys of ESL teachersin which 80% of the responses came from U.S. programswere female (Pennington & Riley, in press-a, in press-b), another recent survey of ESL program directors in the U.S. (Pennington & Xiao, 1990) derives a profile of the typical ESL program director as likely to be female (58%), relatively young and inexperienced in administration, as compared with a group of university department chairs and non-ESL program directors. As to the other dichotomies that Pennycook mentions in his second-to-last paragraph, assuming that he meant these to apply to the Pennington and Young article and not to the Davies (1989) article, which he lumps together with ours on philosophical grounds, I believe, as a matter of fact, that much too much has been made in linguistics of the false dichotomization of universal and relativistic principles, and that second language theorizing has gotten a lot of air play out of faddish dichotomies such as learning versus acquisition or transfer versus developmentterms that are at best unclear and at worst conceptually vacuous (for a discussion, see Pennington, 1988). Moreover, I do not at all believe in dichotomizing language and culture, or language and society; in fact, it seems to me entirely unremarkable to say that language is a sociocultural phenomenon and indeed cannot be defined except in those terms (Pennington, 1990b). Finally, far from believing in any dichotomy between society and the individual, or the social and the psychological, I share the view that whatever is reflected on the individual level in the way of functional or dysfunctional behavior is reflected in the society, and vice versa (Schaef, 1987), and that psychological theories of personality and of the behavior of individuals can be applied quite directly to cultures and to the behavior of groups (Kets de Vries & Miller, 1984), from departments or institutions to much larger social groups. For this reason, I believe strongly in the importance of individual action as the impetus for social change and in freedom of choice as a basic principle. However, I do not agree that anyone can do anything in a classroom and that I must accept it as ESL teaching. I believe that THE FORUM 563

there are certain approaches to instruction that achieve more valid results in a given context than others and that there are certain background characteristics that qualify one to be a member of the ESL teaching profession. Moreover, as Pennycook would presumably agree, the students have a right to a voice in what happens in the classroom. In other words, the students have rights and claims on information that may not be consistent with what the person charged with teaching them makes available, and they also have the right to exercise freedom of choice and individual action to effect change to satisfy their own perceived needs. Pennycook has maintained-and this is not at all a new ideathat the construct of language teaching method is essentially incoherent. At the same time, it seems clear that we do not yet have recognized methods for teacher preparation or evaluation in ESL. This is not to say that we should not keep trying to develop such methods. All recognized academic pursuits are based on method; without method, all research and practice within education is reduced to muddling through. Pennycook, working within a deconstructionist framework, argues that the very concept of method is reactionary and outmoded, ignoring the fact that deconstructionism is itself a highly disciplined method, as it must be to have gained the credibility that it has within academia. While Aileen Young and I may agree with Pennycook about the need to empower ESL teachers, we may not agree with him that professionalism. is basically a dirty word (though in fact Pennycook is somewhat inconsistent here, in that he allows the term professional to occur in a positive, empowering sense in the Giroux and McLaren quotation above). While some denounce professionalism as promoting the informational sterility of groupthink and the withholding of privileged information from the public (Ginsberg, 1988), the attribution of professional also carries implications of service, and of practitioner self-regulation (Crookes, 1989, p. 45). One of the biggest problems in our fieldand one of the reasons that TESOL still cannot command competitive salariesis that it is not in fact perceived as a coherent field or profession, with a coherent set of practices and standards for those practices. As Blaber and Tobash (1989) remark in connection with the TESOL Committee on Professional Standards employment concerns survey: The consensus is that 1) until the field of TESOL is viewed as a profession with unique characteristics, and 2) until TESOL professionals are viewed as having comparable worth to peers and colleagues, it will be difficult to resolve or even address many salary, security, and benefit issues. (p. 4) 564 TESOL QUARTERLY

I submit that TESOL has a very long way to go towards being recognized as a viable field rather than as an auxiliary service or stop-gap form of employment. TESOL is still very much plagued by the problem of untrained or minimally trained nonprofessionals billing themselves as ESL teachers: there are still college students going overseas and teaching ESL to help support themselves; still community volunteers offering ESL instruction for free. On the other side, there are still many people with PhDs and even MBAs billing themselves as ESL program directors who know little if anything about TESOL or about directing programs. As long as we are willing to accept large numbers of such unqualified people in the field, we will continue to have problems being respected and rewarded for the specialized information we possess. One of the areas of widespread dissatisfaction among teachers within TESOL is low pay (Day, 1984; Blaber & Tobash, 1989; Pennington & Riley, in press-a, in press-b). Three other often mentioned areas of dissatisfaction are (a) lack of professional recognition (Blaber & Tobash, 1989), (b) lack of opportunities for advancement, and (c) dissatisfaction with administration in the areas of supervision and implementation of organizational policies and practices (Pennington & Riley, in press-a, in press-b). These areas of dissatisfaction confirm the perception of ESL teaching as having relatively low status within academia and even within the TESOL field itself. The low status of ESL teaching is possibly one of the reasons for the rationalization of the field that Pennycook (1989) decries. It is also one of the reasons that Aileen Young and I decided to do the faculty evaluation article. Our purpose was to raise the level of awareness among TESOL administrators and teachers of the formal and informal pitfalls inherent in the evaluation of faculty, with the anticipated result that teachers and administrators in functioning ESL programs would initiate an examination and revision of the policies and procedures that affect the lives and livelihood of teachers. In publishing our article in the TESOL Quarterly, Aileen Young and I hoped that the information it contained would empower those in the field of TESOL to take more responsibility for faculty evaluation in their own individual contexts, to improve the accountability and the ethics of the profession on a local, program-specific level. Since it does not appear to be a realistic option to dispense with faculty evaluation in ESL, I believe that every ESL practitioner must (a) learn about faculty evaluation; (b) contribute to ensuring that the faculty evaluation process in their own context is conducted in a professional manner, that is, competently and fairly; and (c) be accountable for the resultsas teachers in their classrooms, and as THE FORUM 565

administrators in their evaluations of teachers. Hawkins (1990) contends that accountability to students and to curricular goals is a central aspect of responsible teaching, while Fox (in press) maintains that accountability, while rare in TESOL, is the hallmark of responsible administrative behavior. In closing this response to Pennycook, I would make a general point about information and its power to effect change. One way to confront the problems in TESOL is to attempt a radical reconstruction of its constructs by examining it from a new perspective, applying new concepts and styles of argumentation, which, because of the nature of the enterprise, are likely to be unfamiliar to those inside the field. While such an approach can provide a valuable new perspective for identification and description of problems, its potential impactand especially its utility in actually effecting changewithin the field is limited when the writer opts for correct ideology over clarity of message and for abstraction over pragmatism, offering no immediate alternatives to the repudiated practices, or solutions to the pressing problems that practitioners are facing. A different and more moderate approach is to offer an informed perspective on practices in the attempt to work to improve those practices from within and to develop solutions to existing problems. In the latter way of proceeding, which is the one adopted in the Pennington and Young article, one attempts to open up the field to scrutiny, to demystify it, and to share with practitioners knowledge that was formerly reserved for academics. I submit that this latter way of operating, far from being a danger to TESOL, produces more direct effects and, particularly, more immediate improvements, to the field than the former approach, which is informationally opaque or irrelevant to the majority of practitioners. In attempting to clarify the nature of faculty evaluation and to detail its inherent problems, Aileen Young and I hope that we have succeeded in opening up dialogue with the TESOL membership towards improved practices in TESOL, thereby demonstrating the power and the value of the commodity of shared information for building our profession.

ACKNOWLEDGMENT
This response was written by me and then read by Aileen Young, who provided helpful editorial feedback and who is in complete agreement with all points. Ms. Youngs affiliation was incorrectly identified in our original article. Aileen Youngs affiliation should have been listed as the Hawaiian Mission Academy.

566

TESOL QUARTERLY

REFERENCES
Blaber, M., & Tobash, L. (1989). Report of employment concerns survey. TESOL Newsletter, 23 (2), 4-5. Brown, J. D., & Pennington, M. C. (in press). Developing effective evaluation systems for language programs. In M. C. Pennington, Building better language programs: Perspectives on evaluation in ESL. Washington, DC: National Association for Foreign Student Affairs. Crookes, G. (1989). Grassroots action to improve ESL programs. University of Hawaii Working Papers in ESL, 8 (2), 45-61. Darling-Hammond, L., Wise, A. E., & Pease, S. R. (1983). Teacher evaluation in the organizational context: A review of the literature. Review of Educational Research, 53, 285-328. Day, R. R. (1984). Career aspects of graduate training in ESL. TESOL Quarterly, 18 (1), 109-127. Davies, A. (1989). Is international English an interlanguage? TESOL Quarterly, 23 (3), 447-467. Fox, R. P. (in press). Evaluating the ESL program director. In M. C. Pennington, Building better language programs: Evaluation in ESL. Washington, DC: National Association for Foreign Student Affairs. Ginsberg, M. (1988). Contradictions in teacher education and society: A critical analysis. London: The Falmer Press. Giroux, H. A., & McLaren, P. (1989). Introduction to H. A. Giroux & P. McLaren (Eds.), Critical pedagogy, the state, and cultural struggle (pp. xi-xxxv). Albany, NY: SUNY Press. Hawkins, B. W. (1990, March). Lesson planning as hypothesis formulation, teaching as hypothesis testing. Paper presented at the 24th Annual Conference of Teachers of English to Speakers of Other Languages. San Francisco, CA. Kets de Vries, M. F. R., & Miller, D. (1984). The neurotic organization. San Francisco, CA: Jossey-Bass. Pennington, M. C. (1985). Effective administration of an ESL program. In P. Larson, E. L. Judd, & D. S. Messerschmitt, (Eds.), On TESOL 84 (pp. 301-316). Washington, DC: Teachers of English to Speakers of Other Languages. Pennington, M. C. (1988). In search of explanations for interlanguage phenomena. University of Hawaii Working Papers in English as a Second Language, 7 (2), 41-74. Pennington, M. C. (1989a). Directions for faculty evaluation in language education. Language Culture and Curriculum, 2 (3), 167-193. Pennington, M. C. (1989b). Faculty development for language programs. In R. K. Johnson (Ed.), The second language curriculum (pp. 91-110). Cambridge: Cambridge University Press. Pennington, M. C. (1990a). A professional development focus for the language teaching practicum. In J. C. Richards & D. Nunan (Eds. ), Second language teacher education (pp. 132-151). Cambridge: Cambridge University Press. THE FORUM 567

Pennington, M. C. (1990b). American English phonology: A course for language teachers. Unpublished manuscript. University of Hawaii at Manoa, Department of English as a Second Language, Honolulu. Pennington, M. C., & Riley, P. V. (in press-a). Measuring job satisfaction in ESL using the Job Descriptive Index. Perspectives: Working Papers of the Department of English, City Polytechnic of Hong Kong. Pennington, M. C., & Riley, P. V. (in press-b). A survey of job satisfaction in ESL: TESOL members respond to the Minnesota Satisfaction Questionnaire. University of Hawaii Working Paper in English as a Second Language. Pennington, M. C., & Xiao, Y. (1990). Defining the job of the ESL program director: Results of a national survey. Unpublished manuscript. University of Hawaii at Manoa, Department of English as a Second Language, Honolulu. Pennycook, A. (1989). The concept of method, interested knowledge, and the politics of language teaching. TESOL Quarterly, 23 (4), 589-618. Reasor, A. (1981). Administrative styles of English-as-second-language administrators. Unpublished doctoral dissertation. American University, Washington, DC. Sashkin, M. (1986, Spring). Participative management remains an ethical imperative. Organizational Dynamics, pp. 62-75. Schaef, A. W. (1987). When society becomes an addict. San Francisco: Harper & Row. Schaef, A. W., & Fassel, D. (1988). The addictive organization. S a n Francisco: Harper & Row. Schiller, D. (1988). How to think about information. In V. Mosco & J. Wasko (Eds.), The political economy of information. Madison, WI: University of Wisconsin Press.

568

TESOL QUARTERLY