The debate around bilingual education continues to spark controversy between its detractors and its supporters. The education of linguistic minority students in the United States is a complex issue, involving contrasting theories of education itself, the values of American society, and the extent to which cross-culturalism can be maintained effectively. Although proponents of bilingual education argue that it increases students' academic success, opponents argue that it leads to academic failure (see, for example, Crawford, 1989; Hakuta, 1986; Porter, 1990; Wong Fillmore, 1991).

Success or failure of bilingual education cannot necessarily be addressed as a whole. Several different kinds of bilingual programs are available to the non-English-speaking student in the United States (see Note). These programs differ in the degree to which they promote and/or use English and the home language of the students in the classroom. Thus, the value of bilingualism is seen differently in the different programs. For example, transitional bilingual education is designed so that use of the two languages in the classroom is a temporary phase during transition to English mastery. In contrast, in two-way bilingual programs, in which instruction is given in both languages throughout the program, bilingualism is seen as the ultimate goal - the mastery both of English and of the home language.

While these differences in programs may seem to be purely ideological, the psychological impact on the students is enormous. Lambert (1974) distinguished between "additive" and "subtractive" bilingualism. The additive case implies that an individual suffers no loss of the primary language and the associated culture, while the subtractive case implies that an individual undergoes a loss of primary language skills and general academic performance. Lambert also drew attention to the roles played by attitudes, aptitudes, and motivation in second language learning. He believes that the degree of language mastery influences an individual's self-concept and sense of attainment of proficiency.

There are few studies of students' attitudes toward their own bilingualism, particularly in two-way programs (Christian, Montone, Lindholm, & Carranza, 1997). Griego-Jones (1994), in a small study of 10 Latino kindergarten students in a two-way program, found that the students actually preferred English over Spanish, because English was perceived to be the language of high status and achievement. Looking at older (fourth grade) students, Hayashi (1998) found that students in a two-way bilingual program and in a transitional bilingual program were equally enthusiastic about their bilingualism, as reported on questionnaires. In individual interviews, however, the students in thetransitional program reported that they thought they did not need instruction in Spanish, because they already spoke Spanish. In contrast, the students in the two-way program all thought the time spent in Spanish instruction was valuable and necessary to their achievement in both languages.

Although neither of the studies mentioned above examined data on achievement, research on the most effective forms of bilingual education (usually in terms of English achievement) suggests that two-way programs may be the best. Two-way bilingual education has been described in a national study as "the program with the highest long-term academic success" (Thomas & Collier, 1997, p. 52). The students' success in these programs is undoubtedly due to a number of factors. These include opportunities for linguistic minority students to assume strong peer leadership roles in the classroom, an emphasis on grade-level academic instruction in both languages, sustained support for and use of multicultural curricula, and opportunities for non-English-speaking parents to form close partnerships with the school staff as well as with other parents. The purpose of the present report is to examine students' development in a two-way bilingual program by focusing on both their attitudes toward becoming bilingual (and possibly bicultural) and their school achievement in both languages. Although we do not have data to examine causal links between attitudes and achievement, we see this study as a first step toward showing the relationship between the two.