Language learning in school: the promise of two-way immersion (1).

Consider the following excerpt from an elementary school class. Mr. Hernandez (2) is teaching a lesson about migration to his class of eight-year-olds. Half of the class are fluent speakers of English (including one student already bilingual in Chinese and English); the other half of the students are native Spanish speakers, many of whom come from families whose heritage is in Central America or Mexico. The lesson is in Spanish, and the students break into small, heterogeneous groups to work on research on the migratory patterns of different animals and birds (one group investigates whales, another butterflies, another geese, and so on). The students mostly talk to each other in Spanish, although some English can be heard. The native Spanish speakers often help the Spanish learners on how to say or spell words or phrases in that language. The teacher moves from group to group, speaking only in Spanish, guiding the research and the writing of the findings. A bit later, the groups share their work with each oth er and the teacher helps them form generalizations about migration.

Later on in the same day, during English language time with Mrs. Gilbert, students talk about the migration of people from one country to another and what issues arise. In this session, the native English speakers serve as resources for the English learners. The students spontaneously offer comparisons with what they learned earlier about animal migration and bring up similarities and differences. The teacher uses only English, but occasional interchanges in Spanish can be heard among the students (mostly quick clarifications). These students spend about 60% of their instructional time learning through Spanish, about 40% through English, and most of them have participated in the two-way program for three years, since kindergarten.

This composite description illustrates a typical day in a Spanish/English two-way immersion program, an increasingly popular educational approach in the United States that holds great potential for contributing to effective pedagogy in multilingual contexts elsewhere. In "two-way" classrooms, students from two different language backgrounds study together and receive content instruction in both groups' native languages. As a result, students have access to native speaker peer resources for second language learning and for interaction in and out of the classroom.

The two-way bilingual immersion approach blends the goals and methodologies of maintenance bilingual education for minority language speakers with language immersion for majority language speakers (in the United States, English language speakers). In the U.S., the largest number of programs operate in Spanish and English in elementary schools (grades kindergarten through six). The students are integrated for most, if not all, of their instruction. The goals of such programs include high levels of academic achievement and high levels of proficiency in both languages for all students as well as positive intergroup relations.

Differences between two-way immersion and bilingual education stem primarily from the class composition and language goals. As they have evolved in the United States, bilingual programs do not typically include native speakers of English, while two-way programs do, making available peer language models and resources during English-medium instruction for language minority students. Heterogeneous classes also help prevent the isolation of language minority students from others in their school and community. Most bilingual programs are transitional; the aim is generally to move into all-English instruction as soon as possible, and native language development is not emphasized. Two-way programs, on the other hand, continue to provide instruction in the native language, even after proficiency in English has developed. In other societies where the majority language is other than English, bilingual programs operate in similar ways, offering a transition from a minority language to the majority Language of instructio n in the schools.

In "one-way" immersion programs, all (or nearly all) of the students are immersed in a language that they do not speak natively, and the primary model for the new language being learned is the teacher. This approach is one type of "foreign" language learning in schools offered to primarily monolingual English speakers in the U.S. In contrast, two-way immersion brings in students who are native speakers of each of the languages of instruction, so they can be models and resources for each other during instruction through the target language. An important consideration for programs, then, is how to maximize the use of peer resources for language learning.

Thus, two-way programs bring potential advantages for language learning when compared with other forms of language instruction contexts, particularly in light of recent evidence (Pica et al. 1996) suggesting that some features of interaction may only be available with native speakers of a language. In regions where there are large numbers of native speakers of two or more major languages, this educational approach aligns well with both the sociolinguistic context and the value of bilingualism.

Two-way bilingual immersion is increasingly popular in schools in the United States. A directory, compiled as part of our research and updated most recently in 1997, gives an overview of programs in over 200 schools (Montone -- Christian 1997, Christian -- Whitcher 1995). From the overview, we have found that:

-- there has been a rapid growth in such programs since 1987, when 30 two-way programs were found;

-- most two-way programs operate in Spanish and English (181 out of 204), but others pair English with Korean, French, Navajo, Cantonese, Arabic, Japanese, Russian, or Portuguese;

-- most programs operate at elementary grade levels, kindergarten through sixth grade (172 out of 204);

-- programs vary in their design, language allocations, and student integration, as well as on a number of other characteristics;

-- reviews of program evaluation reports show that, where comparisons are possible, students in two-way programs on the whole have done as well or better academically than their counterparts in other programs.

While there are certain core characteristics shared by two-way programs, there is a great deal of variation in the conditions and manner of implementation. For example, design variations abound. The two major model variations are termed "90/10" (comparable to total immersion) and "50/50" (comparable to partial immersion), based on the proportion of time spent in each language of instruction at the beginning of the program. In the "90/10" model, a large part of the instruction in the early grades is given in Spanish (or other target language), and the amount of English instruction is increased at each grade level until the balance of the two languages is about equal. The allocation of languages of instruction by grade level in a 90/10 program might look as follows:

Grades K-1: 90% Spanish, 10% English

Grades 2-3: 80% Spanish, 20% English

Grades 4-5: 60% Spanish, 40% English

Grade 6: 50% Spanish, 50% English

In the "50/50" model, approximately half the instruction is provided in each language at every grade level, the proportions of use of the two languages of instruction remaining relatively constant over time.

While these two models describe in general terms the major types of two-way programs, there are a number of other variations that occur as well. The distribution of languages for instruction, for example, follows diverse paths. Some programs define language of instruction by time of day (e.g., mornings in one language, afternoons in the other language). Other programs link language of instruction to particular subjects or teachers. Many use a combination of two or more bases for distribution.

In Christian et al. (1997), a study of three elementary schools documents the characteristics and outcomes of two-way programs in action, highlighting what has been learned and what issues are emerging. The programs that we examined exist in diverse geographic and sociocultural settings in the U.S. They represent both total and partial immersion program models and had all been in operation eight or more years. Brief overviews of the three schools provide a picture of their similarities and differences:

One program in the eastern United States includes kindergarten through fifth grade students at an elementary school site (with extensions at a middle and a high school). Students from Spanish-speaking and English-speaking backgrounds come together in a modified magnet arrangement (about half of the students live in the school's neighborhood, the rest come in from other neighborhoods). The program started in 1986 with a single first grade class; in 1994-95, 318 students participated, out of 600 total in the school; and in 1995-96, the school became entirely two-way. Throughout the program, all students receive half of their instruction in Spanish and half in English, on a daily basis. The language switch occurs at lunch time, when classes move from one language medium to the other. For example, a second grade class may learn language arts, reading, and mathematics in the morning in English, and then deal with language arts, social studies, and science in the afternoon in Spanish. Since much of the subject matt er instruction is organized thematically, there is opportunity to treat all disciplines in both languages on a regular basis.

A second program, located in the western U.S., operates as a magnet school for 380 students in preschool through grade 6. Like the previous example, the program was founded in 1986, bringing together Spanish and English language background students. It follows the "90/10" model: in kindergarten and first grade, all students have 90% of their instruction in Spanish, with 10% of the time devoted to oral English language development. All students learn initial literacy in Spanish. During English time, students develop oral English skills through literature, poetry, and music. In second grade, the amount of English time is increased to 15%, and in third grade, to 20%, when English literacy is introduced. Fourth and fifth grades are 60% Spanish/40% English and in sixth grade, instruction is evenly divided between the two languages.

The third school we studied is a magnet school for kindergarten through grade 8 in a large city in the midwestern United States. It serves over 600 students, 60% from low income families; This school-wide two-way program was established in 1975 using a "50/50" model, but the instructional design was later changed because it was felt that students were not gaining enough Spanish proficiency. Since then, the program has followed an "80/20" allocation of languages, as follows: in kindergarten through third grade, 80% of instruction is given in Spanish. Students are grouped by language background for language arts instruction (the only time they are separated), and all students learn to read in their native language. At fourth grade through sixth grade, Spanish is the language of instruction 60% of the time, and in seventh and eighth grade, instruction is evenly divided by language.

Our findings from school-based research on these programs (across model types) include:

-- teaching strategies for language and content. Teachers in two-way immersion classes tend to use hands-on learning, thematic units, and visual/graphic representations to teach content. Cooperative learning is used frequently as an effective way for students to act as linguistic, cultural, and academic resources for each other. Language arts in both languages emphasize meaningful experiences, and many opportunities for reading and writing in both languages are provided. Strategies such as rephrasing, modeling, paraphrasing, and repetition are incorporated regularly to promote language development.

-- separation of languages by teachers and students. Teachers in two-way immersion programs teach for extended periods of time in one designated language. Students are encouraged to use the designated language as well, but flexibility in language choice for students is often allowed, particularly for younger students. Some code-switching occurs, but the amount is quite variable by program and teacher.

-- integration of students. Students from both language backgrounds learn together for significant portions of the day. Some programs separate students by language background for some language arts instruction in the native and/or second language, while others maintain student integration at all times during the day.

-- programmatic factors. Two-way programs often emphasize the importance of support from families and the community at large, and serious efforts are made to ensure that both languages and cultures are valued equally (Craig 1996). Within the school, library resources and computer programs in both languages are available and utilized.

-- initial literacy development. Two-way programs vary in their approach to initial literacy. In some programs, children learn to read in both languages of instruction at the same time. In others, children learn to read in their native language (so they are grouped by language for early literacy instruction), while some programs teach initial literacy to all students in the target language only (for example, all students learn to read in Spanish). All of these approaches seemed to work well in the programs studied.

Implications for planning and implementing two-way programs were also drawn from the findings of the study, along with many issues that remain open questions. Some issues that merit further attention include:

-- What are the consequences of various program models and variations and how do they relate to features of local contexts to optimize student learning?

-- How can language proficiency be developed and maintained in both languages, given the power and influence of English (in the United States)? In general, how should the language proficiency goals be adjusted to take account of the sociolinguistic profile of the broader society?

-- How can two-way bilingual immersion be carried beyond the elementary school most effectively? What articulation issues need to be addressed? Once children have bilingual abilities, what approaches will provide the necessary and sufficient support to continue language development in both languages for older students?

-- What do teachers need to know in order to be most effective in two-way immersion and how can that teacher development happen?

-- What forms of academic and language assessment are appropriate and how can they be developed and used? If comparable assessment instruments are not available in both languages of instruction, what course of action should be followed?

-- Can two-way programs provide a truly additive bilingual environment for language minority students, given the power differences inherent in situations with majority and minority language speakers (Valdes 1997)?

Implementation of two-way bilingual immersion programs is a complex process that reflects not only policies at many levels, but also the local sociocultural context and resources available. The decisions made at each stage of planning and implementation have consequences that need to be recognized and considered so that the optimum results for students and their families will be obtained. Results that are emerging at schools in many communities in the United States indicate the high levels of academic achievement and second language proficiency can be obtained in programs that are well-implemented (Collier 1995; Lindholm 1993). We still have much to learn about two-way bilingual immersion education, but our base of experience and research is growing steadily. As it does, we can see that the approach has potential as an effective model to promote academic achievement and language learning for students in multilingual communities around the world where rich language resources are available and the benefits of b ilingualism are recognized.

(1.) I am indebted to many colleagues who have worked on our research team exploring two-way immersion education, especially Chris Montone, Liz Howard, Kathryn Lindholm, and Fred Genesee. This work has also benefited greatly from discussions with Nancy Rhodes and Deborah Short. Their contributions are reflected throughout the paper, but I take responsibility for all shortcomings.

(2.) This example is a composite, illustrative description of a hypothetical classroom, based on numerous observations in schools throughout the United States; the names and events cited are fietitious.

REFERENCES

Alatis, James E. (ed.)

1990 Georgetown University Round Table on Languages and Linguistics. Washington, D.C.: Georgetown University Press.

Christian, Donna

1996 "Two-way immersion education: Students learning through two languages", The Modern Language Journal 80, 1: 66-76.

Christian, Donna -- Christopher Montone -- Kathryn Lindholm -- Isolda Carranza

1997 Profiles of two-way immersion education. McHenry, IL: Center for Applied Linguistics and Delta Systems.

Christian, Donna -- Anna Whitcher

1995 Directory of two-way bilingual programs in the United States. (Revised edition.) Santa Cruz, CA -- Washington, D.C.: National Center for Research on Cultural Diversity and Second Language Learning (also: http://www.cal.org).

Collier, Virginia

1995 Acquiring a second language for school. (Directions in Language and Education 1, 4.) Washington, D.C.: National Clearinghouse for Bilingual Education.

Craig, Barbara

1996 "Parental attitudes toward bilingualism in a local two-way immersion program", The Bilingual Research Journal 20, 3-4: 383-410.

Lindholm, Kathryn J.

1993 "Un modelo educacional de exito para estudiantes inmigrantes de origen latino", Revista de Psicologia Social y Personalid 9: 85-94.

Montone, Chris - Donna Christian

1997 Directory of two-way bilingual programs in the United States. 1997 Supplement. Santa Cruz, CA -- Washington, D.C.: Center for Research on Education, Diversity and Excellence (also: http://www.cal.org).

Pica, Teresa -- Felicia Lincoln-Porter -- Diana Paninos -- Julian Linnell

1996 "Language learners' interaction: How does it address the input, output, and feedback needs of L2 learners?", TESOL Quarterly 30, 1: 59-84.

Tucker, G. Richard

1990 "Cognitive and social correlates of additive bilinguality", in: James E. Alatis (ed.), 90-101.

Valdes, Guadalupe

1997 "Dual language immersion programs: A cautionary note concerning the education of language-minority students", Harvard Educational Review 67, 3: 391-429.