Dual Language Immersion Program

Patrick H. Smith University of Arizona Universidad de las Américas-Puebla

Elizabeth Arnot-Hopffer Davis Bilingual Magnet School

Abstract

 

In this article, the authors describe how teachers in a Spanish/English dual language elementary school in Tucson, Arizona promote Spanish literacy using a school-designed program, Exito Bilingüe. Based on ongoing work and participant observation with dual language students and teachers, the authors show how dual language schooling has evolved at the school, and how the model currently in use compares to case studies in the literature in terms of program goals, type, and distribution of languages of instruction. The components of Exito Bilingüe, a school wide, multi-age, non-scripted Spanish literacy program, and its implementation are described and preliminary results are discussed. The authors find support for the transfer of reading skills from Spanish to English and for the inclusion of exceptional education students in dual language schooling. They argue that, contrary to the promises of commercially prepared, scripted reading curricula, dual language readers are best served by teachers working together to design literacy instruction to meet local conditions and learner needs.

The dual language (DL) model is widely cited as an exemplary means of educating language minority children while promoting second language acquisition among children of all backgrounds (Christian, 1994; Collier, 1995; Wink, 2000). DL programs are sites for examining possible tensions between the need for all students to develop high levels of literacy in English and the fact that language minority students have historically been underserved in U.S. schools (Valdés, 1997). There is considerable evidence that learning through the native language has many advantages for language minority students. It facilitates the development of both basic and advanced literacy in Spanish and English; it allows Spanish dominant students to gain important content knowledge that will make the English they encounter more comprehensible; and it enhances overall cognitive and social development. Many schools, even those with bilingual education programs, often treat the native language of minority students as a problem to be overcome, adopting a remedial attitude, with its attendant negative connotations.

Research on literacy development and the connections between minority language communities and dual language schooling are particularly important in the "quasi-border" community of Tucson, Arizona (Jaramillo, 1995). The historic role of local educators in the development of the modern bilingual education movement (National Education Association, 1966), and current efforts to restrict or ban outright bilingual education in Arizona are central issues to present research. This paper describes how language minority and language majority students at Davis Bilingual Magnet School in Tucson are becoming biliterate via a school designed Spanish literacy program, Exito Bilingüe. We trace the development of Exito Bilingüe as it has evolved in response to the unique needs of students and to the strengths and concerns of Davis educators.

Describing Davis Bilingual Magnet School

Located only a few blocks from downtown Tucson, Davis Bilingual Magnet School sits on a narrow piece of land between the tracks of the Southern Pacific Railroad and Interstate-10 in one of the city's first Hispanic neighborhoods, Barrio Anita. Davis is one of the oldest schools in Tucson (founded in 1901), and for most of its history the school has served the families and children of Barrio Anita and nearby Hispanic barrios. Following a desegregation order by a federal court in 1978 and faced with dwindling enrollment, the district planned to close the school, but parents and residents of Barrio Anita convinced the district to rebuild it instead (Pérez, 1998). As a result, Davis became the district's bilingual magnet school in 1981.

Students from Barrio Anita continue to attend Davis, as do magnet students from other parts of the city. Barrio students, who make up about 35% of the student population, are mostly from Spanish-dominant homes, although almost all begin school with considerable knowledge of English and Spanish. Magnet students constitute approximately 65% of the student population, and most begin school as monolingual or dominant speakers of English. About 70% of Davis students are of Latino heritage, approximately 20% are European-American, 6% are African-American, and 4% are of Native American heritage. The largest single group of students at Davis is third and fourth generation Mexican-Americans whose families have chosen the school's dual language program hoping their children will (re) gain Spanish and revitalize their Latino heritage. Besides differences in ethnic background and linguistic dominance, there are also social class differences between the barrio and magnet populations. About half of Davis students qualify for free or reduced school lunch programs, most of them from barrio families.

The Davis faculty consists of 12 classroom teachers (two per grade level), plus three subject area specialists for art, music, and physical education. There are full-time instructional aides in each classroom, as well as a full-time exceptional education teacher, librarian, curriculum specialist, and counselor on staff. All faculty, staff, and administration are bilingual. Like most magnet schools, the student body and faculty are quite stable. In fact, there is a school joke that teachers leave Davis only on their way to the mortuary.

Another feature of dual language education at Davis is close links with Spanish speakers and Latino culture in Tucson. The school is home to a "guitarristas" group and the state's first performing mariachi group for elementary school students. The "Escuela Nocturna" (night school) features Spanish as a Second Language, English as a Second Language, and other classes for adults. There is also an extended day program, which is particularly useful to working parents. Many dual language programs attempt to include parents and families of all backgrounds in the life of the school; these mechanisms are some of the ways Davis is structured to involve language and cultural minority families.

Program Goals and Curriculum Model

In this section we describe dual language schooling at Davis more closely in terms of (a) goals; (b) whole program vs. strand-within-a-school; and (c) distribution of languages of instruction across grade levels. For purposes of comparison, we show how the program at Davis is similar to and different from well-known dual language programs.

According to Lindholm and Fairchild (1990), dual language programs typically share four goals: academic achievement in two languages; development of bilingualism and biliteracy; high levels of self-competence; and positive cross-cultural attitudes. These goals are shared by Davis educators, along with the additional goals of promotion of inquiry and critical thinking, and participation in community service projects (Arizona Department of Education, 1998). At Davis, Spanish literacy is important for all students, for its contribution to full bilingualism, and as a vehicle for content learning. Consistent with other approaches to DL education (viz., Freeman, 1996), this enrichment approach does not emphasize language development over academic and social development; the goal is balanced development in all areas.

Dual language programs can be distinguished by the way they structure curriculum in response to two fundamental questions: (a) will all students participate in the dual language program? and (b) for the DL students, how will the two target languages be distributed across the curriculum? Leaving aside the discussion of exceptional education students for now, there are two basic approaches to the first question. The first approach can be called a strand-within-a-school, in which parents choose to enroll children in the dual language component or in another instructional program offered at the school. In Arizona, the Valley View School in Phoenix offers a strand-type DL program along with a more traditional transitional bilingual model. In Tucson, Nash Elementary is another example of a strand-type program, with the alternative choices to DL being ESL and English immersion.

The second approach, which is the option practiced at Davis, is a whole-school approach, meaning that all students are enrolled in the DL program. Table 1 shows how some well-known programs fit into the strand vs. whole-school approach to DL schooling.

Table 1 Two Approaches to DL Education

The question of how to distribute the majority language and minority language across grade levels is a complex one because it requires consideration of teacher strengths and linguistic abilities, availability of appropriate materials, and scheduling demands, as well as articulation with the language offerings at subsequent levels of schooling (i.e., middle school and high school). Two basic approaches have been identified in the literature on dual language education (Christian, Montone, Lindholm, & Carranza, 1997; Ovando & Collier, 1998). The simpler version, certainly in terms of scheduling, is the so-called 50-50 model, in which instructional time is equally divided between the minority and majority languages. The Key School in Arlington, Virginia (Christian, et al., 1997; Craig, 1993), the Amigos Program in Cambridge, Massachusetts (Cazabon, Nicoladis, & Lambert, 1998), and the Oyster Bilingual School (Freeman, 1996; 1998) are well-known examples of programs which use the 50-50 model. The other approach is the 90-10 model, used by the River Glen School in San José (Christian, et al., 1997) and other programs in California. In this model, instruction in Kindergarten uses the minority language 90% of the time and the majority language the remaining10%. As children pass through the program, the percentages of minority and majority language use gradually converge until, by the end of elementary school, language of instruction is equally divided between the two languages. By year five, the language-of-instruction equation is typically the same for 50-50 and 90-10 programs.

Table 2

Distribution of Minority and Majority Languages in Three Models of Dual Language Schooling

Grade	50-50 Model	90-10 Model	Davis School
K	50-50	90-10	100-0
1st	50-50	90-10	100-0
2nd	50-50	80-20	85-15
3rd	50-50	70-30	70-30
4th	50-50	60-40	70-30
5th	50-50	50-50	70-30

As shown in Table 2, the dual language model at Davis is similar to the 90-10 model in that the use of English language instruction increases in the higher grades. Indeed, comparison of grades two and three shows nearly identical distribution of languages. There are important differences, however. Teachers at Davis have elected to use Spanish as the sole language of instruction in Kindergarten and first grade after October 1st because they want to use the first weeks of the school year to allow very young learners to adapt more readily to the school culture and classroom practices. Also, the Davis model does not aim for an equal balance by the end of elementary schooling. Instead, the ratio of Spanish-English instruction remains at 70-30 in the final three elementary grades. Thus, a student who begins Kindergarten at Davis and continues through fifth grade experiences a marked increase in the number of contact hours in the minority language as compared to students in the 50-50 and 90-10 models.

Rather than static models, the distribution of languages in DL programs should be seen as a dynamic response to local conditions and as an index of the school/community commitment to the minority language. In Chicago, one of the oldest dual language programs in the country, the Inter-American Magnet School recently moved to a 90-10 model following more than 20 years of using the 50-50 model (Christian, et al., 1997; Zucker, 1995). At Davis, the ratio of minority-to-majority language use has steadily moved toward the increased use of Spanish. When the school became a magnet in 1981, a maintenance bilingual education program was implemented in which teachers were expected to use Spanish as a vehicle for instruction 50% of the time and English 50% of the time. By the mid-1980s, however, the school began to question this model as it became obvious that students who entered the program dominant in Spanish exited in fifth grade bilingual and biliterate, but that students who entered the program dominant in English made little progress toward becoming productive bilinguals, although most could understand Spanish quite well. In the late 1980s, under the direction of a teacher leader and the curriculum specialist, a teacher study group was formed to examine different bilingual education program models and goals (Murphy, 1999). During the 1993-94 school year the school decided to implement a new model currently referred to as the dual language immersion program, "el Programa de Inmersión en Dos Idiomas." Initially, there was reluctance on the part of the faculty and the administration to use the word "immersion" in describing the model. In fact, during those first years the program was referred to as "Spanish enrichment," and there was some confusion among faculty about the amount of Spanish to be used at each grade level. In March 1997 the faculty and administration met to formalize the program and agreed to begin to refer to it publicly as dual language immersion.

The question of how to translate the DL model shown in Table 2 into effective practice at the classroom level is regularly discussed at Davis. Unlike some DL programs where native speakers of Spanish and English "team teach" at Davis the same teacher uses both Spanish and English for instruction in each (2-5) classroom. Decisions about language distribution across the subject areas are made by classroom teachers. Successful dual language programs often mark language of instruction by day of the week or by content (Christian, et al., 1997). Some teachers at Davis have decided to alternate language use in certain subjects depending on the unit of study. For example, in social studies a unit on Brazil might be taught in Spanish, while students learn about Italy via English. Because second grade classrooms use English for instructional purposes 15% of the time, one second grade teacher decided to use that language on Wednesdays only using Spanish exclusively the other days of the week (Wednesday is a short day at Davis. Students are dismissed early so that faculty can participate in staff development). Teacher commitment to the DL model, teacher autonomy in making decisions about how to implement the model, and the administration's respect for teacher professionalism all contribute to the success of the dual language immersion program.

Because teachers use the minority language across the curriculum, it has been important for the school to be able to purchase Spanish materials for all subject areas. Since the district has a long-standing tradition of maintenance bilingual education, equitable materials in both languages are usually available for consideration. In the fall of 1997 the Davis faculty implemented a new math curriculum school wide (the Investigations curriculum developed at TERC, formerly Technical Education Research Centers, and published by Dale Seymour Publications, 1998). Kindergarten and first grade teachers requested materials exclusively in Spanish since, according to the program model, they use Spanish 100% of the time for instructional purposes. At other grade levels, materials are purchased in both languages so that teachers can make informed decisions about which parts of the new math curriculum to teach in Spanish or in English. During a recent district adoption of a social studies curriculum, the primary teachers (K-2) at Davis ordered the new materials in Spanish while the intermediate teachers (3-5) requested that half of their materials be in Spanish and half in English (McGraw-Hill Social Studies Curriculum). Similarly, the district science materials (Full Option Science System, University of California, 1992) are available in Spanish and English. Even with subject area materials available in both languages, Davis teachers, like dual language educators across the country, continue to make many of their own instructional materials (Howard & Loeb, 1998).

The approach to dual language schooling currently used at Davis—a whole—school program which gives preference to Spanish as the instructional language through all grade levels—has been developed in order to promote high levels of oral bilingualism and biliteracy. However, as recent dialogue on electronic discussion lists (BILING, 1999, February-March) suggests, many dual language educators and researchers are concerned with the question of what languages dual language learners and teachers actually use in classrooms and other school domains. Although reports remain almost entirely anecdotal (see Freeman, 1996; 1998), there is a sense that even in well-designed and implemented programs, students tend to use the majority language increasingly over the minority language. Carmichael (1998) has documented the language ideologies of dual language students to show that students have identified and internalized the social values associated with Spanish and English as early as Kindergarten, and that attempts to foster student-student use of the minority language face considerable barriers. As in other dual language programs Davis teachers are becoming more aware of the role of student-student discourse, particularly outside the classroom, as factors in overall attainment of Spanish. In the following section, we discuss the motivation for the development of a new approach to Spanish literacy instruction at Davis, Exito Bilingüe.

Like most U.S. public school educators, students and teachers at Davis work in the context of increased public demand for accountability, typically associated with high scores on standardized measures of academic achievement. Although many Davis teachers are critical of the validity of standardized measures for minority students and for second language learners in particular, they recognize that continued parental and district support for the DL program is at least partially contingent on high test scores. Therefore, although teachers employ more authentic methods for gauging literacy at Davis (including portfolios of student work and a writing assessment in both Spanish and English scored using a rubric scale) the school has been pleased with recent results of the Stanford 9 Achievement Test. The scores of the cohort of 1997-98 fifth graders are especially suggestive since most of these students had received five years of dual language instruction at Davis. Although they had received most of their schooling in Spanish, they scored above the district and national averages in English language arts and reading in English (Reyes, 1999, p. 27). The first generation of Davis students to complete Kindergarten through fifth grade under the dual language model took the Stanford 9 test in spring 1999, and analysis of the results is pending.

"Success" on standardized measures of reading presents an interesting situation for educators who do not believe in them. As several teachers observed, even if the results indicate that reading instruction at the school is generally effective, this is of little benefit in planning. Thus, even with these encouraging preliminary results, Davis teachers felt instruction could be improved. In particular, they expressed frustration with the wide range of Spanish literacy skills reflected in each classroom. In first grade, for example, some students were still learning color words in Spanish, while others were already reading Spanish chapter books. Another problem identified by teachers was placement of English dominant "newcomers," students without previous bilingual education or Spanish literacy entering the program in the intermediate grades. A third issue involved articulation with district-wide plans for a new "balanced literacy" approach to reading and writing instruction. Teachers asked how Davis, as the only dual language school in the district, would incorporate an approach to literacy instruction designed for children reading in their first language, and to what extent?

Creating Exito Bilingüe

Together, the faculty began to reflect and brainstorm to develop ways to effectively deal with the variety of strengths and needs in Spanish reading and writing at each grade level. Teachers decided to create a school wide multi-age Spanish language arts program to better meet the needs of students at all levels of Spanish literacy development. Support from the district was critical to this effort, as faculty were paid to participate in a week-long training program aimed at understanding the components of a balanced literacy program and implementation of interactive literacy strategies. With leadership from the curriculum specialist, teachers decided which of the components to implement and when. Thus, Exito Bilingüe was born. It is important to emphasize that this is a school wide program designed for ALL students from first through fifth grade. Kindergarten teachers decided that their students would be included in Exito Bilingüe on an individual basis so that reading and writing did not become the overriding focus of the Kindergarten experience. Exceptional education students participate, as does all Davis faculty, including instructional aides and specialists.

Guiding Principals of Exito Bilingüe

In a strategy she refers to as "identification by negation," Maria Montaño-Harmon (1988) suggests that sometimes things are best defined by a description of what they are not. This is the format we chose in describing Exito Bilingüe. Exito Bilingüe is not tracking. Although Davis students are grouped by their ability to read in Spanish for Exito Bilingüe, the Exito Bilingüe groups are flexible, with students changing groups based on the results of ongoing assessment via running records. Exito Bilingüe is not remedial instruction. Students and teachers at Davis work at their cognitive maximum as the components of a balanced literacy program are implemented using interactive literacy strategies in Spanish, the second language of most Davis students. Exito Bilingüe is not bilingual instruction. Spanish is the language of instruction for all Exito Bilingüe groups. Exito Bilingüe is also not a scripted program. It is a negotiated curriculum that integrates the professional expertise of faculty and the interests and strengths of students. Exito Bilingüe is a school wide, multi-age interactive literacy program in Spanish that is implemented three days a week for 60-minute periods.

In this section, we describe the steps Davis teachers took in implementing this new Spanish literacy program, beginning with assessment of student literacy proficiency and continuing with the components of the district's requirements for balanced literacy. The first step in implementing Exito Bilingüe at Davis was the assessment of Spanish literacy for all first through fifth grade students. Determined to focus on the strengths of students (Genesee & Nicoladis, 1995), teachers borrowed from the work of Marie Clay (1975; 1993) and others in reading recovery, and decided to assess students in Exito Bilingüe

through running records. First developed by Clay, running records are "tools for coding, scoring, and analyzing a child's precise reading behaviors"(Fountas & Pinnell, 1996, p. 89). Each running record documents a reader's actual reading of a target text, "providing quantitative and qualitative information" (Fountas & Pinnell, 1996, p. 78) useful in decisions about placement and individual instruction. During the first six weeks of school, the authors listened to each first through fifth grader read in Spanish. And because the faculty had agreed to define reading as meaning making, students were asked to retell the stories they had read in order to evaluate reading comprehension. Many English-dominant students were able to decode and comprehend the stories they read in Spanish very well, even though some of them chose to re-tell the story in English, the language in which they feel most comfortable speaking. Although colleagues at other schools have reported using running records in Spanish with Spanish-dominant readers, the use of running records in Spanish with English-dominant readers is a pioneering effort that Davis teachers are proud of. Teachers continued reading with students until their instructional level was found, with accuracy between 90-94%. Since this was their first experience with running records this was often times a rather lengthy process. One first grader, for example, began with a level-8 book and, over the course of three different sessions, showed that her instructional level was actually level 40, the highest for which we had materials.

With the results of the running records 14 relatively homogenous multi-age Spanish literacy groups were formed for instruction during Exito Bilingüe. New students are assessed using a running record in Spanish and are then placed at their appropriate instructional level. The first phase of Exito Bilingüe instruction began at the end of September, nearly nine months after the initial idea was conceived. Students were re-assessed by their Exito Bilingüe teacher in mid-December and again in early March, with new group assignments made after each phase of assessment.

Components of Balanced Literacy

Developers of interactive literacy have identified eight components of a balanced literacy program. The following list, taken from Fountas & Pinnell (1996, pp. 22-23), has served as a reference for Davis teachers in the development of Exito Bilingüe:

Reading Aloud to Children

The teacher reads aloud to the whole class or small groups. A carefully selected body of children's literature is used; the collection contains a variety of genres and represents our diverse society. Favorite texts, selected for special features, are reread many times.

Shared Reading

Using an enlarged text that all children can see, the teacher involves children in reading together following a pointer. The process includes:

• Re-reading big books

• Re-reading retellings

• Re-reading alternative texts

• Re-reading the products of interactive writing

Guided Reading

The teacher works with a small group who are at about the same level in reading ability. The teacher selects and introduces new books and supports children reading the whole text to themselves, making teaching points during and after the reading.

Independent Reading

Children read on their own or with partners from a wide range of materials. Some reading is from a special collection at their reading level.

Shared Writing

Teacher and children work together to compose messages and stories; teacher supports process as scribe.

Interactive Writing

As in shared writing, teacher and children compose messages and stories which are written using a "shared pen" technique that involves children in the writing.

Guided Writing and Writers' Workshop

Children engage in writing a variety of texts. Teacher guides the process and provides instruction through mini-lessons.

Independent Writing

Children write on their own, including (in addition to stories and informational pieces) re-tellings, labeling, speech balloons, lists, etc.

Davis teachers chose to begin Exito Bilingüe by implementing three of these eight components and to add more over time. They began with (1) Independent Reading, which is called TEMAL (Todo el Mundo a Leer); this is a 15-minute interval devoted to sustained silent reading in Spanish for everyone at Davis;

(2) Reading Aloud, and (3) Shared Writing (which Davis teachers call Write Aloud). Next, Shared Reading and Interactive Writing were added and, most recently, Guided Reading, "the heart of interactive literacy instruction" (Fountas & Pinnell, 1996, p. 1). Many teachers are already including guided writing and independent writing, and the writing components of Exito Bilingüe are currently the focus of our bi-weekly staff development meetings, which are called "Reflexiones de Exito Bilingüe."

Results of Exito Bilingüe

In addition to teacher discussion of student progress in reading proficiency, the school, in connection with researchers from the University of Arizona, has compiled a database of students' running records. Table 3 shows preliminary results by grade level, based on comparison of Spanish running records taken at the beginning (Etapa 1) and near the end (Etapa 3) of the 1998-99 school year.

Table 3

Average Gains in Running Record Levels by Grade-level

Grade	All Davis Students	Exceptional Education Student	Non-Exceptional Education Students
1st	3.62 levels N=37)	(N=1)	3.69 levels (N=36)
2nd	4.43 levels N=28)	(N=1)	4.55 levels (N=27)
3rd	3.54 levels N=28)	2.0 levels (N=6)	3.95 levels (N=22)
4th	3.03 levels (N=6)	3.33 levels (N=6)	2.96 levels (N=23)
5th	3.54 levels (N=26)	(N=2)	3.59 levels (N=24)
School	3.63 levels (N=148)	2.5 levels (N=16)	3.76 levels (N=132)

These results suggest that all students at Davis are improving in Spanish literacy skills. Exceptional education students are included in these gains, although perhaps at a slower rate of advancement than non-exceptional education students. These early results appear to be consistent with the results of the annual districtwide Spanish writing assessment showing literacy development in the target language. Furthermore, student performance on a districtwide English writing assessment, as well as favorable results on (English) reading and language arts portions of the Stanford 9 Achievement Test, suggest that the development of English literacy by language minority or language majority students is not hindered by primary literacy instruction in Spanish. These results are consistent with studies of early biliteracy via schooling from a wide range of international contexts (cf. Modiano, 1968; Rosier, 1980; Williams, 1996), and can be interpreted as evidence in support of transfer of reading skills from one language to another (Krashen, 1985; Collier, 1995). To better understand these effects, we are currently in the process of assessing Davis students using running records in English. A longitudinal study currently being conducted at Davis will allow us to better understand how the results of standardized tests for both language majority and language minority students relate to their progress in Exito Bilingüe.

Reflexiones de Exito Bilingüe

In this section we discuss how Davis educators have used a study group, Reflexiones de Exito Bilingüe, to reflect on progress toward the goal of Spanish literacy for all students. Structurally, the agenda is set and the meeting is run by the school's curriculum specialist. Teachers and teacher assistants are required to attend the bi-weekly meetings as part of on-going in-service training, and participants also receive credit toward re-certification. Despite their obligatory nature, Reflexiones sessions are perceived by teachers as an opportunity to affect the conditions and outcomes of teaching in their own classroom and throughout the school, and discussion is generally critical and engaged. The nature of the intellectual work accomplished by teachers during Exito Bilingüe is quite remarkable compared to teacher study groups we have experienced in public schools, either as teachers or teacher trainers. Some of the most important features include (a) Peer feedback on implementing balanced literacy; (b) Teacher-to-teacher communication; (c) Reflecting on the school's position within the district; (d) Asking key questions about bilingualism and biliteracy.

Peer Feedback on Implementing Balanced Literacy

Teachers routinely share with the group—by telling, demonstrating techniques and materials, and by use of video—the ways they are implementing the components of Balanced Literacy in their own Exito Bilingüe groups. Although the format allows for the strengths of individual teachers to emerge, we have also been impressed with the willingness of teachers to risk admitting when they do not know, and when their understanding is different from that of others. Comments about HOW teachers want to continue learning indicate that participants are conscious of themselves as learners in this process.

Teacher-to-Teacher Communication

Although the meetings are generally run by the curriculum specialist, discussion does not depend on her holding the floor. Instead, teachers frequently pose public questions to presenters and other colleagues about Spanish literacy. In this way, many important issues are addressed which could not have been envisioned by a single group leader. In recent sessions teachers have developed policy for including Kindergarten students ready for success in Exito Bilingüe groups and for postponing participation for those who are ready academically but not socially. Similarly, the school-wide decision to pool funds for book acquisitions in order to purchase more non-fiction books in Spanish emerged from the study group's concern that all readers need more experience with non-fiction.

Reflecting on the School's Position within the District

As the district's bilingual magnet school and the only dual language program, teachers are aware that district policies do not always fit the needs of Davis students. In agreeing to implement the district's policy of "Balanced Literacy," for example, Davis teachers opted to incorporate components of the model gradually and as a faculty rather than overnight. The school's identity as a somewhat autonomous entity responsible for meeting student needs was evident in a recent Reflexiones session in which teachers used a discourse analysis technique to assess the value of an external evaluation performed by the district. In this meeting, teachers questioned the evaluation team's understanding of Spanish literacy (as participants pointed out, several of the evaluators who observed Exito Bilingüe sessions were English monolinguals!). Rather than focusing on how to implement suggestions they saw as unfounded, participants suggested ways future evaluations could provide feedback useful in improving literacy instruction to benefit dual language students at Davis.

Asking Key Questions about Bilingualism and Biliteracy

Perhaps most importantly, Reflexiones de Exito Bilingüe is a forum for critical thinking and for continuing to ask central questions about how the school promotes bilingualism and biliteracy. Teaching requires making multiple, rapid decisions on a wide range of questions and issues, on a daily basis. Reflexiones provides teachers with an opportunity to reflect on their teaching that is seldom possible in the context of the classroom. It is also an opportunity to consider and reconsider fundamental questions about the nature of bilingualism and biliteracy. Here, Reflexiones serves as a resource not only for teachers but also for researchers involved in a longitudinal study of dual language schooling at Davis. For example, when researchers asked teachers to identify children in their classes as "Spanish dominant," "English dominant," or "bilingual," a second-grade teacher responded by asking the researchers to clarify their criteria for bilingualism. What labels, she asked, should be given to those students who, although clearly dominant in one language, demonstrate that they are capable of learning via written and spoken Spanish and English? Consistent with their understanding of learner proficiency in two languages, several teachers created separate categories for students who they regard as bilingual but dominant in one language or the other. Other teachers distinguished between oral and written proficiency and receptive vs. productive competencies in Spanish and English. These are issues at the heart of dual language education and the formation of bilinguals via schooling.

Implications for Promoting Minority Language Literacy

The successful promotion of Spanish literacy and biliteracy via Exito Bilingüe offers several lessons for dual language educators and those considering implementing dual language programs. First and most broadly, the success of English- and Spanish-dominant students on standardized measures of English reading following initial and sustained reading instruction in the minority language supports claims (viz., Collier, 1995; Cummins, 1992; Krashen, 1985) that reading proficiency developed in one language transfers to the ability to read in the other language. Second, the fact that exceptional education students at Davis are included in these gains in reading is consistent with findings from French immersion programs in Canada (Genesee, 1995/1987), suggesting that such students can benefit from participation in dual language schooling. The multi-age format of Exito Bilingüe groups, which can be viewed as an extension of mixed age music and art groups, challenges exceptional education students and all students at Davis to work at their cognitive maximum as they develop as readers. Third, in the current climate of growing public support for commercially prepared and sometimes highly scripted reading programs, Exito Bilingüe demonstrates that schools and teachers are quite capable of researching, designing, and implementing effective literacy programs which meet district and state guidelines without imposing packaged curricula, materials, or invasive evaluation procedures which devalue what learners and teachers know. Finally, since such programs are typically aimed at schools serving high concentrations of language minority speakers, students from low-income homes, and others who tend to fare poorly on standardized measures of English literacy, it is important to emphasize that barrio students are included in the academic gains Davis students make during Exito Bilingüe.

This research also identifies challenges facing dual language educators who wish to develop a Spanish literacy program that does not rely exclusively on standardized testing. Based on the philosophy that reading is making meaning of written text, Davis teachers chose to use running records as the primary means of assessing Spanish literacy. However, many Davis students, including some as young as first and second grade, are capable of reading Spanish at difficulty levels for which few graded reading materials are available. Exito Bilingüe teachers are working to identify field tested, graded reading materials that reflect the advanced literacy skills of their students. The lack of such materials for running records in Spanish suggests that publishers of Spanish language materials remain focused on the literacy assessment needs of more numerous transitional programs for which high levels of Spanish literacy is seldom a goal. By adding to the demand for advanced reading materials in minority languages, dual language programs such as the one at Davis may contribute to a change in the current textbook publishers market.

Visitors to Davis often point out that the school enjoys resources atypical of U.S. bilingual programs and public schools generally. Indeed, the school's magnet status, stable student population, small school and class size, number of highly bilingual teachers and staff, strong support from the principal, and parent commitment to bilingualism are all critical factors underlying the successful development of biliteracy. None of these conditions is easily created, and at least some are largely or wholly beyond the school's influence. In this sense, Davis offers a further example of a common lesson in public schooling_that schools with resources tend to be successful (as Troike puts it, "them as has, gits," (1984, p. 49). At the risk of stating the obvious, neither the dual language model nor the Spanish literacy program we have described should be regarded as a panacea, especially not for schools without the resources found at Davis. Instead, the best way to view such resources is as necessary but insufficient means of promoting Spanish literacy. Perhaps the most important factors in the success of Exito Bilingüe to date have been the focus of teacher and staff energies on the common goal of biliteracy and the willingness to take risks as a faculty to create the best possible curriculum given local conditions and the needs of Davis students. A resource which rich schools cannot buy and to which poor schools can have equal access, this level of commitment is both powerful and well within the power of schools.

References

Arizona Department of Education. (1998). School profile: Davis Bilingual Magnet School, 1997-98.

BILING: Forum for Discussion of Research on Bilingual Education, Arizona State University. [On-line]. Available e-mail: biling@asuvm.inre.asu.edu

Carmichael, C. (1998). Hablar dos veces: Talking twice: Language ideologies in a dual language Kindergarten. Unpublished manuscript, University of Arizona, Tucson.

Cazabon, M. T., Nicoladis, E., & Lambert, W. E. (1998). Becoming bilingual in the Amigos two-way immersion program. Research Report Number 3, Center for Research on Education, Diversity & Excellence: University of California, Santa Cruz.

Christian, D. (1994). Two-way bilingual education: Students learning through two languages. Research Report Number 12, National Center for Research on Cultural Diversity and Second Language Learning: University of California, Santa Cruz.

Christian, D., Montone, C. L., Lindholm, K. J., & Carranza, I. (1997). Profiles in two-way immersion education. Washington, DC: ERIC/Center for Applied Linguistics.

Clay, M. M. (1975). What did I write? Portsmouth, NH: Heinemann.

Clay, M. M. (1993). An observation survey of early literacy achievement. Reading Teacher, 45, 264-273.

Collier, V. P. (1995). Promoting academic success for ESL students: Understanding second language acquisition for school. Elizabeth, NJ: Teachers of English to Speakers of Other Languages-Bilingual Educators.

Craig, B. A. (1993). The public as language planner: Promoting grassroots bilingualism. Unpublished doctoral dissertation, Georgetown University.

Cummins, J. (1992). Empowerment through biliteracy. In J. V. Tinajero, & A. F. Ada (Eds.), The power of two languages: Literacy and biliteracy for Spanish-speaking students (pp. 2-18). New York: MacMillan/McGraw-Hill.

Fountas, I. C., & Pinnell, G. S. (1996). Guided reading: Good first teaching for all children. Portsmouth, NH: Heinemann.

Freeman, R. D. (1996) Dual-language planning at Oyster Bilingual School: "It's much more than language." TESOL Quarterly, 30 (3), 557-582.

Freeman, R. D. (1998). Bilingual education and social change. Clevedon, England: Multilingual Matters.

Genesee, F. (1995/1987). The Canadian second language immersion program. In O. García, & C. Baker. (Eds.), Policy and practice in bilingual education: A reader extending the foundations (pp. 118-133). Clevedon, England: Multilingual Matters.

Genesee, F., & Nicoladis, E. (1995). Language development in bilingual pre-school children. In E. E. García, & B. McLaughlin (Eds.), Meeting the challenge of linguistic and cultural diversity in early childhood education (pp. 18-33). New York: Teachers College Press.

Howard, E. R., & Loeb. M. I. (1998). In their own words: Two-way immersion teachers talk about their professional experiences. (ERIC Digest, December 1998) Washington, DC: Center for Applied Linguistics

Jaramillo, J. (1995). The passive legitimization of Spanish: A macrosociolinguistic study of a quasi-border: Tucson, Arizona. International Journal of the Sociology of Language, 114, 67-91.

Krashen, S. D. (1985). Inquiries and insights: Second language teaching, immersion and bilingual education, literacy. Englewood Cliffs, NJ: Alemany Press.

Lindholm, K. J., & Fairchild, H. H. (1990). Evaluation of an elementary school bilingual immersion program. In A. M. Padilla, H. H. Fairchild, & C. M. Valadez (Eds.), Bilingual education: Issues and strategies (pp. 126-136). Newbury Park, CA: Sage Publications.

Modiano, N. (1968). National or mother language in beginning reading: A comparative study. Research in the Teaching of English, 2, 32-43.

Montaño-Harmon, M. R. (1988). Discourse features in the compositions of Mexican, English-as-a-Second-Language, Mexican-American/Chicano, and Anglo-American high school students: Considerations for the formulation of educational policies. Unpublished doctoral dissertation, University of Southern California.

Murphy, E. (1999, April). Where do we go from here? Participatory learning and action in a teacher study group. Paper presented at the meeting of the Society for Applied Anthropology, Tucson, AZ.

National Education Association. (1966). The invisible minority: Pero no vencibles. Report of the NEA-Tucson survey on the teaching of Spanish to the Spanish-speaking. Washington, DC.

Ovando, C. J., & Collier, V. P. (1998). Bilingual and ESL classrooms: Teaching in multicultural contexts. Boston: McGraw Hill.

Pérez, A. (Ed.) (1998). Visions of Barrio Anita: Looking through the eyes of different generations. Tucson, AZ: Tucson/Pima Arts Council.

Reyes, E. I. (1999). Stanford 9 Achievement Test Analysis - 1998: Davis Bilingual Magnet School. Unpublished manuscript, University of Arizona, Tucson.

Rosier, P. (1980). A comparative study of two approaches of introducing initial literacy to Navajo children: The direct method and the native-language method. Outstanding dissertations in bilingual education, 1979. (pp. 97-108). Rosslyn, VA: National Clearinghouse for Bilingual Education.

Troike, R. C. (1984). SCALP: Social and cultural aspects of language proficiency. In C. Rivera (Ed.), Language proficiency and academic achievement (pp. 44-54). Clevedon, England: Multilingual Matters.

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

Williams, E. (1996). Reading in two languages at year five in African primary schools. Applied Linguistics, 17 (2), 182-209.

Wink, J. (2000). Critical pedagogy: Notes from the real world (2^nd ed.). New York: Addison-Wesley Longman.

Zucker (1995). The role of ESL in a dual language program. Bilingual Research Journal, 19 (3 &4), 513-523.

EndNotes

¹The research reported in this article was made possible in part by a grant from the Spencer Foundation. An earlier version of this paper was presented at the 1999 Conference of the Arizona Association for Bilingual Education. We are grateful to Davis teachers, Dr. Guadalupe Romero, Davis Principal, and to anonymous reviewers of the Journal for insightful comments on earlier drafts. The views expressed here are solely the responsibility of the authors.

²These figures reflect only those students from whom running records were collected for each of three testing periods during the 1998-99 school year, Newcomers, school leavers, and students who achieved the maximum possible score on their first or second running record are thus excluded from this analysis, as are Kindergarten students.

³Exceptional education students are fully integrated in Éxito Bilingüe reading groups. Due to their low numbers at first, second, and fifth grades, we have omitted figures for gains by exceptional education students at these levels.