of the dominant majorityis then examined. The final section reviews findings on educational conditions for second-language learning.
Bilingualism is pervasive throughout the world, but it varies according to (1) the conditions under which people become bilingual, (2) the uses they have for their various languages, and (3) the societal status of the languages. For example, in postcolonial Africa, students may be educated in English or French while another language is spoken in the home, and yet another (e.g., Swahili in eastern Africa) may be used in public encounters and institutional settings, such as the courts (Fishman, 1978). In officially bilingual countries such as Switzerland, children use one language at home and for most schooling, but, at least if middle class, are expected to acquire competence in at least one other official language; French and German are of equivalent social status and importance to success. Yet another set of conditions in created in bilingual households, where parents who are native speakers of two different languages choose to use both in the home. Finally, bilingualism is often the product of migration. Immigrants frequently continue to use their native languagewhich may be of low status and not institutionally supportedat home, and learn the dominant language of their new society only as required for work, public encounters, or schooling. The children of such families, for whom school is the primary social context, may end up fully bilingual, bilingual with the new language dominant, or having little knowledge of the parental language. They are the children of particular interest in this report.
A number of typologies of bilingualism have been offered. A major distinction among these typologies is that some focus their explanation at the individual and others at the societal level.
Weinreich (1953) distinguishes among compound, coordinate, and subordinate bilinguals, who differ in the way words in their languages relate to underlying concepts. In the compound form, the two languages represent the same concept, whereas in the coordinate form, the concepts themselves are independent and parallel. In the subordinate form, the weaker language is represented through the stronger language. These different forms are clearly related to the social circumstances in which the two languages are learned, but the distinction also reflects an individual's mental makeup. Weinreich's distinction led to a number of studies seeking behavioral differences reflecting this typology (e.g., Lambert et al., 1958). Though such attempts were essentially abandoned because of the difficulty of operationalizing the distinction, speculation that different bilingual experiences result in different cognitive and neural organization persisted.