Later in this report we discuss whether different types of education programs might result in qualitatively different types of individual bilinguals. Findings suggest, by and large, that bilingualism attained under different conditions of exposure will not differ in the ways language is organized with respect to cognitive structures.

Typologies of bilingualism based on societal variables have focused mainly on the prestige and status of the languages involved. Fishman et al. (1966) draw a distinction between "folk" and "elite" bilingualism, referring to the social status of the bilingual group. The "folk" are immigrants and linguistic minorities who exist within the milieu of a dominant language that is not their own and whose own language is not held in high esteem within the society. The "elite" are those who speak the dominant language and whose societal status is enhanced through the mastery of additional languages. As Fishman observes, "Many Americans have long been of the opinion that bilingualism is 'a good thing' if it was acquired via travel (preferably to Paris) or via formal education (preferably at Harvard) but that it is a 'bad thing' if it was acquired from one's immigrant parents or grandparents" (pp. 122-123).

Similarly, Lambert (1975) distinguishes "additive" from "subtractive" bilingualism. This distinction relates to the effect of learning a second language on the retention of the native language. In additive bilingualism, the native language is secure, and the second language serves as an enrichment. Canadian French immersion programs for the English-speaking majority are a prime example of additive bilingualism since native English speakers maintain their English and add French. In subtractive bilingualism, the native language is less robust; society assumes that it will be used only temporarily until replaced by the dominant language as the group assimilates. Most immigrants to the United States, Canada, and Australia experience subtractive bilingualism; their skills in their native languages erode over time, and English becomes their dominant language (see also the discussion of language shift later in this chapter).

These broader social distinctions help explain why programs that seem quite similar can have such divergent effects in different social settings—for example, why an immersion program in Canada succeeds in teaching French to English-speaking students who continue to maintain full proficiency in English and to function at a high academic level, while an immersion program to teach English to Spanish-speaking immigrants in