The position taken in this paper is that the basic debate concerning the effectiveness of bilingual education has been totally "socially naive." A study investigated the concept that the ethnolinguistic vitality of a community determines the quality and quantity of linguistic contacts with one's own linguistic group and with other ethnolinguistic groups, which in turn strongly influence linguistic proficiency, ethnolinguistic identity, and desire to integrate first-language (L1) and second-language (L2) communities. A model to that effect is proposed and applied to about 1,500 grade 12 anglophone and francophone students in seven Canadian provinces. Effects of the degree of L1 schooling and those of the strength of the L1 network of linguistic contacts in the social milieu were analyzed. It was found that the latter were stronger than the former for these variables: desire to integrate L1 and L2 communities, ethnolinguistic identity in L1 and L2, L1 and L2 self-rated oral proficiency, and L2 cognitive-academic proficiency. L1 schooling had the strongest effect on cognitive-academic proficiency. Results support the hypothesis that additive bilingualism is best promoted by immersion in L2 for high-vitality groups and by L1 schooling for low-vitality groups. It is concluded that the effects of bilingual education cannot be understood without taking account of the strong influences of the students' sociolinguistic environment. (MSE)