bilingual education

For many Americans, bilingual education seems to defy common sense – not to mention the Melting Pot tradition. They ask:
If non-English-speaking students are isolated in foreign-language classrooms, how are they ever going to learn English, the key to upward mobility?
What was wrong with the old "sink or swim" method that worked for generations of earlier immigrants?
Isn't bilingual education just another example of "political correctness" run amok – the inability to say no to a vociferous ethnic lobby? Some English Only advocates go further, arguing that even if bilingual education is effective – which they doubt – it's still a bad idea for the country because bilingualism threatens to sap our sense of national identity and divide us along ethnic lines. They fear that any government recognition of minority languages "sends the wrong message" to immigrants, encouraging them to believe they can live in the U.S.A. without learning English or conforming to "American" ways.
Such complaints have made bilingual education a target of political attacks. One of the most serious to date is now under way in California, a ballot initiative that would mandate English-only instruction for all children until they become fully proficient in English.
No doubt many of the objections to bilingual education are lodged in good faith. Others reflect ethnic stereotypes or class biases. Sad to say, they all reflect a pervasive ignorance about how bilingual education works, how second languages are acquired, and how the nation has responded to non-English-speaking groups in the past.
Reinforcing popular fallacies requires less space than deconstructing them. That's why my writing on these issues grew from a handful of newspaper articles into a 310-page book, Bilingual Education: History, Politics, Theory, and Practice (3rd ed., 1995). Nevertheless, a few points:
Science is often counterintuitive. Its breakthroughs tend to upset common-sense notions, not to mention cherished myths. Linguistics is no exception. In fact, it invites more than its share of opposition from nonspecialists – witness the "Ebonics" controversy – because most people feel like experts when it comes to language. Our reactions are often visceral. Perhaps that's because our speech defines us ethnically, socially, and intellectually. It's tied up with a sense of who we are – and who we are not – evoking some of our deepest emotions.
What once seemed obvious about bilingualism – for example, that it handicaps children's cognitive growth – has usually proved unfounded. Since the 1960s, research has shown that multiple language skills do not confuse the mind. Quite the contrary: when well-developed, they seem to provide cognitive advantages, although such effects are complex and difficult to measure (Hakuta 1986).
Another discredited notion is that children will "pick up" a second language rapidly if "totally immersed" in it. For generations, this philosophy served to justify policies of educational neglect – assigning minority students to regular classrooms, with no special help in overcoming language barriers. Disproportionate numbers failed and dropped out of school as a result. The sink-or-swim approach was ruled illegal by the U.S. Supreme Court in Lau v. Nichols (1974).
Research has shown that the quality – not the quantity – of English exposure is the major factor in English acquisition. That is, the second-language input must be comprehensible(Krashen 1996). Otherwise, it's just noise.
English as a second language (ESL) is best taught in natural situations, with the second language used in meaningful contexts rather than in repetitious drills of grammar and vocabulary. One variant of ESL, known as "sheltered subject-matter instruction," adapts lessons to students' level of English proficiency. This approach is common in bilingual education programs, coordinated with lessons in students' native language.
Native-language instruction also helps to make English comprehensible, by providing contextual knowledge that aids in understanding. When children already know something about dinosaurs, a lesson on the subject will make more sense when instruction shifts to English. Not only will they learn more about dinosaurs; they will also acquire more English.
The same principle applies when it comes to acquiring literacy. Teaching in the native language can facilitate the process, as the linguist Stephen Krashen (1996) explains:
We learn to read by reading, by making sense of what we see on the page. ...
If we learn to read by reading, it will be much easier to learn to read in a language we already understand.
Once you can read, you can read. The ability to read transfers across languages. "Language is not a unitary skill, but a complex configuration of abilities" (Hakuta and Snow 1986). Social communication skills – a.k.a. playground English – should not be confused with academic English, the cognitively demanding language that children need to succeed in school. While playground English tends to be acquired rapidly by most children, academic English is typically acquired over a period of five to seven years (Cummins 1989).
Research on the effectiveness of bilingual education remains in dispute, because program evaluation studies – featuring appropriate comparison groups and random assignment of subjects or controls for pre-existing differences – are extremely difficult to design. Moreover, there is considerable variation among the pedagogies, schools, students, and communities being compared. While numerous studies have documented the benefits of bilingual programs, much of this research has faced methodological criticisms – as noted by an expert panel of the National Research Council (August and Hakuta 1997a).
Certain critics of bilingual education have interpreted the NRC report to mean that, despite a generation of research, "there is no evidence that there will be long-term advantages or disadvantages to teaching limited-English students in the native language" (Glenn 1997). This conclusion – widely circulated by the so-called READ Institute – has been rejected by the NRC study directors. To the contrary, they say, the expert panel concluded that "a great deal has been learned from the research that has been conducted on English language learners." Moreover, there are "empirical results . . . support[ing] the theory underlying native language instruction" (August and Hakuta 1997b). According to the panel's chairman, the "attempt by READ to place its own political spin" on the report hardly advances the cause of objective research (Hakuta 1997).
Other critics continue to deny that such empirical support exists. A recent "review of the literature" (Rossell and Baker 1996) reports that bilingual education is inferior to English-only programs of all kinds, including sink-or-swim. Yet these conclusions owe more to the manipulation of program labels than to student performance in the classroom. Critiques of Rossell and Baker by Cummins (1998) and Krashen (1996) show that, among other distortions, the researchers rely heavily on studies of French immersion in Canada – bilingual or trilingual approaches that they portray as monolingual "immersion" or "submersion" models. Meanwhile, a meta-analysis of the same body of research reviewed by the critics, but using a more rigorous methodology, found quite different results: a significant edge for bilingual education (Greene 1998).
The most sophisticated evaluation study to date – a four-year, longitudinal study of 2,000 Spanish-speaking students in five states – found that "late-exit," developmental bilingual programs proved superior to "early-exit," transitional bilingual and English-only immersionprograms (Ramírez et al. 1991). That is, in programs that stressed native-language skills, students' growth in English reading and mathematics continued to increase long after it had leveled off among their peers in the other programs. While this study has been praised by many, others have rejected the comparison as invalid because all three programs were not tested in the same school districts.
Nevertheless, a consensus of applied linguists recognizes that the following propositions have strong empirical support:
Native-language instruction does not retard the acquisition of English.
Well-developed skills in the native language are associated with high levels of academic achievement.
Bilingualism is a valuable skill, for individuals and for the country. Bilingual education was adopted by many local school districts in the 1960s and 1970s to remedy practices that had denied language minorities an equal educational opportunity. Yet it was hardly a new invention designed to replace the Melting Pot with the Salad Bowl or some other model of ethnic pluralism. There is a long bilingual tradition in the U.S.A., in which minority-language schooling has played a central, albeit largely forgotten, role.