The third-grade classroom experiences of Mary, a child from Liberia with a limited educational and English language background, are viewed in relation to the beliefs that her teacher held about language, acculturation, natural cognitive processes, and cooperative learning. In the case of teaching linguistic minority children in mainstream classrooms, it is argued that a set of powerful ideas from a sociocultural perspective may have more influence on teachers than those deriving from an input-output model. Such ideas rest on a view of the development of language and cognition in context, as opposed to a modular view of the development of linguistic structure.IN MANY SCHOOL DISTRICTS THROUGHout the U S , there is a decreasing number of linguistic minority students whose needs to learn English are being addressed in exclusively pull-out English as a second language (ESL) environments, as inclusion models are being implemented in many K-12 settings.' In Florida, for example, a 1990 lawsuit filed on behalf of linguistic minority community groups against the Florida Department of Education resulted in a Consent Decree mandating that all teachers serving limited English proficient (LEP) students should use English to speakers of other languages (ESOL) strategies to adapt content instruction for those students (LULAC vs. Florida Board of Education, 1990). Subsequently, a massive staff-development effort has been undertaken in the state, giving teachers 60 hours of ESOL strategy training, and inclusion models are, in many cases, replacing the ESL pullout model.* It is assumed, albeit wrongly in many cases (Harper, 1996), that regular teachers are now actually using these strategies for the LEP students in their programs and are ~ ~ The Modern LanguagcJburnal, 81, i (1997) a1997 The Modern L o n e Jrmrnal 0026-7902/97/28-49 @.50/0making their courses comprehensible to these students. It is further assumed that students are also learning English through course content, a belief following from the strong promotion of &ashen's (1985) input model for second language acquisition (SLA) in Florida. The gap between classroom practice and the findings of research about second language (L2) acquisition is of immediate concern, both in district-level staff development and in teacher preparation programs to prepare a wider range of public school instructors to teach learners whose first language is not English. Compo