I, Lily Wong Fillmore, declare as follows:

1. I am a Professor of Education at the University of California at Berkeley. I have been on the Berkeley faculty since 1974, and have studied, since 1972, the processes by which children acquire English as a second language under various conditions in the public schools. The studies I have conducted over the past two and a half decades have been longitudinal in design, and I have direct evidence on how children acquire a second language over time. In particular, my research has examined the ways in which instructional practices and classroom structure affect language learning, and how English language proficiency affects children's academic progress, both in English only and in bilingual programs. The children whose language acquisition I have tracked over the years have been Asians and Latinos in elementary school programs in California. I have published articles and reports on second language learning in children, and I know the research conducted by others on how children learn English as a second language at school. A copy of my vita is attached.

2. I am familiar with the provisions of Proposition 227, and have been asked by counsel to provide an expert opinion on how Proposition 227 will affect the education and the learning of English by children.

3. Proposition 227 requires that "nearly all" instruction be conducted in English, and provides children one year of instructional support for the learning of English in "intensive sheltered English immersion programs." I will address two questions in this declaration: (1) How much progress can children make in English with one year of instructional support? And, (2) will the English children are able to learn in one school year give them access to the school's curriculum if this is made available only in English?

4. Proposition 227 is based on the assumption that most children will learn enough English to function effectively in school after one year in so-called "intensive sheltered English immersion" classes. By this the authors of Proposition 227 are apparently blending several terms which refer to programs that are conducted exclusively in English for LEP students. These programs, variously called "sheltered English, "structured immersion," and "intensive English" attempt to make the schools' curriculum available to LEP students by teaching content in English, and by so doing, to promote the learning of English as a second language. There are numerous such classes in schools throughout California, in which all of the students are limited in their knowledge of English. They are usually found where either the number of students or the availability of bilingual teachers or materials are insufficient to justify a program using native language instruction. In some instances they are used for students with advanced English skills who are on the verge of transitioning to an English mainstream situation. From a review of these classrooms, we can determine what is likely to occur in the Proposition 227 classroom and whether children are likely to leave such a classroom after one year able to fully participate in a mainstream classroom. 

How are teachers to teach the school's curriculum to children who do not understand the medium of instruction? The usual solution is to reduce the content ordinarily covered and to simplify the language used to impart the curriculum to the barest essentials. Hence, neither the content nor the language used in such classes are at appropriate levels for the students. The resultant "dumbing down" of the language and content means that students do not get access either to the language or to the curriculum they should be learning in school, whatever their grade level. While a sheltered program with primary language support may approximate equal access for children with substantial but imperfect English language skills, there is no way that it can provide equal or indeed any meaningful access for the non-English proficient student.

Can students base their learning of English on the instructional support provided in such classes? The answer is this: the English they learn in such classes in a year is not enough to allow them to survive, much less compete, in school without substantial further specialized instructional assistance. Moreover, students in such situations often learn an imperfect pidgin-like variety of English rather than the standard variety of English required for school. These learner varieties of English, if not corrected, can cause lasting problems for students, since they are difficult to overcome. This phenomenon has been documented both in the Canadian French immersion programs (Selinker, Swain and Dumas, 1975; Harley and Swain, 1984) and in immersion programs in California (Campbell, 1984) and in sheltered English programs (Wong Fillmore, 1991; Schmida, 1996). There are two reasons why this happens in these "structured immersion" classes. The first is that the students in these classes are all language learners, and the second is that they are compelled to communicate with one another in a language none of them initially know. In the sheltered English programs, the students speak a learner's imperfect English - it is not the standard variety of English the teacher wants them to learn and to use. The only model for standard English is the teacher, but since the language learners outnumber the teacher, they are far more likely to hear and learn the imperfect English being used by classmates than they are the English spoken by the teacher. It takes children far longer to learn standard English in such situations than it would otherwise take. When children get off to such a start in the learning of English, it is often difficult for them to start over again, even after they have been exited from the program. There are numerous LEP students in California who have been through similar such programs and who find it difficult to overcome the learner varieties of English they initially learned. All of them eventually learn survival English, but substantial numbers of them, having gotten their start initially in programs like the proposed sheltered English immersion classes, learn forms of English which do not support progress in school.

5. How much English can children learn in a year? Can they gain access to the school's curriculum with the English they learn in a year? I can say without hesitation children do not learn enough English in a year to make schooling provided primarily in that language meaningful. In the types of classrooms Proposition 227 prescribes, children eventually learn enough English to interpret everyday classroom management language. ("It's time for reading. Group A first. Bring your books with you.") They may not know what the words in those statements mean, but by paying attention to non-verbal cues, they can figure out what the teacher wants them to do. However, they rarely get anything approximating the material they are supposed to be learning in school. It is clear that for most children, learning to read is a cognitively demanding task. Learning to read in a language they do not understand is an exponentially more demanding task. In fact, without a solid foundation of the language used in class, children cannot easily deal with any part of the curriculum which is dependent on language. For example, social studies, science, and even math are subjects that require a solid understanding of the language in which they are taught. It is sometimes believed that math is a subject which can be taught to students without much language, but that is true only if one is referring to the most rudimentary and basic levels of math. The mathematics taught at upper grade levels requires knowledge of mathematical relationships and concepts. Thus, even mathematics knowledge at these levels is limited unless and until a student learns academic English. In school, what children must learn each year serves as the foundation for what they must learn in subsequent years. If they do not understand the language well enough to learn in the second grade, what is lost is more than the curricular content of the second grade. They will not have the background knowledge presupposed by the content of the third grade, and so on.

6. To answer the question of whether or not children can have access to the school's curriculum with the English they are able to learn in a year, we must consider the complex nature of language proficiency. The linguistic knowledge required for academic learning is much more complex than that required for everyday social discourse (Cummins, 1982; 1989). For children, more often than not, everyday social discourse involves the here-and-now. Children can generally make themselves understood in ordinary social interactions whether or not they are fully competent speakers of a language, by counting on their interlocutors' ability to figure out what they are trying to say, given the situation at hand. They themselves can more or less understand what others are saying in such interactions by attending to cues that are provided for them by the situation or that are available in the environment. Even then, the linguistic knowledge required for everyday social discourse takes longer than a year to acquire. Few children can gain more than a shaky command even over this type of language in less than two years. For non-English speaking children to participate on a more or less equal academic and social footing with English speaking peers in the classroom, they must be competent in addition to this basic level of language, with language which is required for academic learning - the abstract and conceptually complex aspects of linguistic knowledge involved in literacy development, logical reasoning and problem solving. This level of linguistic knowledge and functioning takes substantially longer to acquire even under the best of circumstances. (See Cummins, op.cit., Collier and Thomas, Ramirez, et al.) This is true whatever the grade level, including those at the primary level. Children must have the prerequisite linguistic foundation to understand the concepts they are expected to acquire through the school's curriculum. This becomes increasingly necessary with each grade level. The materials covered in each successive grade level become increasingly complex conceptually, and thus students require higher and higher levels of language proficiency to make any sense of the materials they are expected to learn. For these reasons, the research has shown that proficiency in academic English can not be attained by most children in less than 4 to 5 years under the best of conditions. It should also be noted that academic English requires not only proficiency in speaking and understanding, but also in reading and writing at the levels required for each grade level. Absent this, the child will fail in a mainstream classroom that is taught exclusively or nearly exclusively in English. That is what one can predict, on the basis of the available research, for these children.

I declare under penalty of perjury under the laws of the State of California that the foregoing is true and correct.

Executed this day of May, 1998, at _________________________, California.

Lily Wong Fillmore, Ph.D.

Supplemental Declaration of Lily Wong Fillmore