Women use linguistic forms associated with the prestige standard more frequently than men. One reason for this is that working-class speech has favourable connotations for male speakers. Favourable attitudes to non-standard speech are not normally expressed, however, and emerge only in inaccurate self-evaluation test responses. Patterns of sex differentiation deviating from the norm indicate that a linguistic change is taking place: standard forms are introduced by middle-class women, non-standard forms by working-class men. (Sociolinguistic variation; linguistic change; women's and men's speech; contextual styles; social class; British English.)

The paper addresses itself to the study of the sociological correlates of speech behaviour among bilingual speakers of English and Swahili in Tanzania. Factors influencing language maintenance, code-switching and code-mixing are discussed. Four main phases of language acquisition are considered: the pre-primary school phase, the primary school phase, the secondary-school phase, and the post-secondary school phase.

Three languages with both varying and overlapping roles interact, creating a triglossia situation: first the vernacular or mother-tongue of each particular ethno-cultural group; secondly Swahili, the local lingua franca and national language; thirdly English, the predominant language of higher learning and to a certain extent of official and commercial business.

The paper also discusses the diglossia relationship between the vernacular and Swahili on the one hand and Swahili and English on the other. The developmental state of the languages is dealt with in terms of socially ‘restricted’ and ‘elaborated’ codes.

Urban life tends to impose its own socio-cultural influences on the bilinguals. There is free shifting and mixing between Swahili and English interlocutors, topics and setting.

Lastly the paper raises questions of the sociological and linguistic consequences of the multilingual situation. (Multilingualism, diglossia; code-switching; code-mixing; Swahili; English, national language problems.)

Work of the Language Survey of Ethiopia is drawn on to illustrate that social surveys can contribute to scientific theories and methods as well as to policy making. Findings with regard to four general problems are indicated: a metric of linguistic distance (considered here in terms of selection of diagnostic traits for linguistic areas, subgrouping of related languages, and dialect variation within a language; measurement of basic cognate frequencies; and testing of mutual intelligibility, a typology of the processes of language spread (lingua franca, mother tongue, superposed variety); verbal deference behavior (in use of Amharic second person pronouns); and pidginization (Simplified Italian of Ethiopia). Some comparisons to findings in sociolinguistics research elsewhere are made. (Language surveys; measurement of linguistic distance; language spread; social meaning and mode of address; pidginization.)

Socialization is defined as the incorporation of a child into society on the basis of social and linguistic interaction. Central to this process is a child's advancement in communicative skills. As a child progresses in linguistic competence, the structure of interaction changes to accommodate his newfound skills, and underlying these changes are the Luo attitudes and beliefs on child-rearing, language acquisition, and the position of children in society. (Socialization; speech acquisition; social interaction; language attitudes.)