The past year or so has brought a virtual avalanche of dramatic events in the field of bilingual education, portending a significant shift in how English language learners are taught in the United States.
In April 1998, Secretary of Education Richard Riley announced a major shift in policy, calling for a goal of English language proficiency in three years for virtually all English language learners. Riley asserted that "new immigrants have a passion to learn English, and they want the best for their children" (p. 2). A survey of 420 randomly selected members of the Association of Texas Educators (both inside and outside the field of bilingual education) found that the majority agreed with the secretary. They believed that children spend too much time in native language instruction (Tanamachi, 1998). Traub (1999) also argues that Latino students spend far too much time in native language instruction, concluding that, in its current form, "bilingual education seems to be hurting" Latino students the most—"the one group it was initially designed to help" (p. 33).
This view stands in stark contrast to the position of several noted scholars in the field, who feel that English language learners should be taught all academic subjects in their native language for no fewer than five, and preferably seven, years (for example, Cummins, 1994). These scholars believe that extensive academic instruction in the native language is necessary for students to benefit from mainstream classrooms.
Recent events indicate that some large school districts (for example, New York and Denver) and some states (for example, California) are seriously rethinking how they educate English language learners. Invariably, the initiatives call for students to enter English language academic instruction at a much earlier age, and they propose a significant reduction in academic instruction in native languages. An article in the New York Times reports that "in response to years of criticism of the city's bilingual education programs . . . New York City plans to dramatically increase the amount of time devoted to English language development" (Archibold, 1998). The article concludes with a summary of major lawsuits. Lawsuits or threatened litigation in Sacramento, Denver, and Albuquerque convey the emotional tenor of the debate.
Increasingly, parents and teachers (most notably Jaime Escalante and Gloria Tuchman) have begun to question the small amount of time devoted to English language development in many bilingual education programs in the primary grades. Advocacy groups have consistently raised such issues as parental choice in the amount of English language instruction each child receives, how early a child is introduced to substantive English language instruction, and when a child should exit classrooms that use a great deal of native language instruction.
- At what age is it best to introduce academic instruction in English to young students?
- To what extent—if any—does native language instruction benefit students' cognitive and academic growth?
- Which are the best instructional methods for developing English language proficiency?
Unfortunately, research findings have stubbornly failed to provide answers to the first two questions. Ironically, we have more research-based information on the third—and least emotional—of these guiding research questions.
Searching for Answers
An unbiased review of research addressing the first question indicates that we do not have adequate information to determine the optimal time for a child to be taught academic content in English.
This is not to say that researchers have not passionately debated the issue or that they have not developed and disseminated a vast array of complex theories. This issue has been debated extensively and serves as the basis of some of the aforementioned lawsuits.
The cornerstone of most contemporary models of bilingual education is that content knowledge and skills learned in a student's primary language will transfer to English once the student has experienced between five and seven years of native language instruction. Yet absolutely no empirical research supports this proposition. Methodological problems so severe that the question cannot be adequately answered plague the research on the subject (August & Hakuta, 1997). These problems appear to be most severe in some of the larger studies intended to "answer" major policy questions.
The recent report released by the National Academy of Sciences, Improving Schooling for Language Minority Children(August & Hakuta, 1997), offers a laundry list of complaints concerning these studies: The major national-level program evaluations suffer from design limitations, lack of documentation of study objectives, poorly articulated goals, lack of fit between goals and research designs and excessive use of elaborate statistical designs to overcome shortcomings. (P. 138)
In addition, the report concludes that "it is difficult to synthesize the program evaluations of bilingual education because of the extreme politicization of the process" (p. 138). The report makes clear that the prevalence of writings by "advocates who are convinced of the absolute correctness of their positions" (p. 138) presents serious barriers to attempts to improve the quality of instruction for English language learners.
Trying to unravel the issues behind these conflicts and debates can be frustrating. Even the National Academy of Sciences report is of little immediate help. It is as filled with contradictions as most other writing in the field. For example, the authors savagely critique the research on effective schooling and classroom processes, yet report the findings from these seriously flawed studies as if they represented solid facts. Similarly, the authors indicate that there is no empirical support for the effectiveness of native language instruction in the early grades, yet still advocate its use. However, the report also demonstrates an awareness of the contradictory nature of the database by noting: It is clear that many children first learn to read in a second language without serious negative consequences. These include children who successfully go through early-immersion, two-way, and English as a second language (ESL)-based programs in North America. (P. 23)
Michael Kirst of Stanford University (Schnaiberg, 1998) recently provided some valuable insight into the problems within the bilingual education knowledge base. In discussing California, he noted: From its inception . . . in the 1970s, bilingual education has been oriented toward inputs, process and compliance. . . . The assumption was if you have this input, the outputs would take care of themselves. So . . . [we monitor] . . . whether you mounted the program, and not its results. (P. 16, emphasis added)
Although Kirst was discussing California, similar problems have been noted in states such as Texas and Massachusetts. This concern with compliance as opposed to learning outcomes helps explain why the bilingual education knowledge base is so inadequate—which in turn contributes to many of the current problems in the field.
Increasingly, researchers argue that we need to focus on aspects of instruction that lead to improved learning outcomes as opposed to political labels that at best crudely describe complex instructional interventions. Several years ago, my colleagues and I received support from the U.S. Department of Education to begin to articulate these components. Our charge was to synthesize the knowledge base on effective classroom practices that simultaneously promote English language development and academic learning. We intentionally eschewed the ongoing political debates. Our goal was to delineate specific techniques that teachers could use to simultaneously promote learning and English language development.
English Language Development
Although questions about optimal age remain unanswered, at some point all English language learners begin academic instruction in English. The initial transition is often called "content area ESOL," "structured immersion," or "sheltered content instruction." The common feature is teachers' use of English designed for students who are not proficient in the language. In sheltered instruction, teachers modulate their use of English so that it is comprehensible to the student and base their degree of support on their knowledge of that student. In some cases, teachers use native language to help a child complete a task, to clarify a point, or to respond to a question.
Almost invariably, sheltered content instruction is coupled with instruction geared toward building the student's knowledge of the English language. In years past, this component has been referred to as ESL or ESOL. Increasingly, educators are using "English language development" (ELD). Historically, teachers focused on the formal structure of language (for example, grammar and mechanics). Critics routinely attacked this approach, however, because it failed to capitalize on the communication function of language, did not generate student interest, and resulted in very limited generalization.
The 1980s brought more "natural" conversational approaches to teaching English. These also attracted criticism, primarily because they did not necessarily help students learn the highly abstract, often decontextualized language of academic discourse. A movement began about 10 years ago to merge English language learning with content acquisition. The rationale is that students can learn English while learning academic content and that this type of learning will build academic language (Cummins, 1994)—that is, the abstract language of scientific, mathematical, or literary discourse. However, too often teachers merely "hope that language occurs [during lessons]. There is a risk during content instruction of neglecting language development" (Gersten & Baker, in press).
The erratic quality of ELD instruction is at the root of the growing dissatisfaction with current practice. Inadequate attention has been devoted to curriculum development, pragmatic teacher training and professional development, and applied research. In a recent professional work group that I conducted in California for the U.S. Department of Education (Gersten & Baker, in press), an educator from the district bilingual education office articulated the problem: "It's important for teachers to be clear about objectives and goals . . . yet an explicit statement of goals does not exist [in district or state curricular materials]."
I would argue, however, that we have made definite progress in understanding what instructional goals are feasible for this group of students and what specific classroom practices are likely to help meet these goals. In our two-year research synthesis project (Gersten & Baker, in press), we concluded that the beginning of an empirical knowledge base on effective instructional practices for English language learners exists. It is important to emphasize, however, that this knowledge base is emerging and should be the topic of controlled, high-quality classroom research.
Principles for Merging ELD with Content Area Learning
- Avoid oversimplifying with contrived, intellectually insulting material when teaching acadmic content in English. Subjects such as science and math can be excellent venues for merging English language development because all students are learning a new technical vocabulary and there is great potential to use concrete objects (Chamot, 1998).
- Use visuals to reinforce verbal content when teaching in English (Saunders et al., 1998; Reyes & Bos, 1998).
- Use both oral and written modalities frequently (Saunders et al., 1998).
- Employ strategic use of synonyms. Word choice and sentence structure need to be consistent and concise during second language learning. Pay attention to use of metaphors and similes and other highly culture-specific phrases and expressions (Gersten & Jiménez, 1994).
- Focus on approximately five to eight core vocabulary words in each lesson. Some strategies include (1) carefully selecting words (evocative words that stimulate instruction, key words for understanding a story), (2) linking words or concepts to words known in the native language, (3) showing new words in print, and (4) using visuals (for example, concept maps) to depict concepts or word meanings (Saunders et al., 1998).
- Use native language during ELD strategically. At times, it might be helpful to use both native language and English during instruction; however, be aware of the risk of overreliance on simultaneous translations.
- During the early phases of language learning, modulate and be sensitive when providing feedback and correcting language usage; however, during later stages, identify errors and provide specific feedback to students (Reyes, 1992).
To date, much has evolved from grassroots experimentation and attempts by researchers to describe practices that appear to engage students and enhance their learning. Nonetheless, it is becoming increasingly clear that a set of practices exists that teachers can use to persistently, but sensitively, encourage students to learn content while expressing their ideas in a new language. This information may be particularly helpful as the shift toward greater emphasis on instruction in English takes effect.
