Language shift from German–Hungarian bilingualism to the exclusive use of German is occurring in the community discussed. Young women are further along in the direction of this change than older people and young men. The linguistic contrast between German and Hungarian is shown to represent the social dichotomy between a newly available worker status and traditional peasant status; thus the choice of language in interaction is part of a speaker's presentation of self. Young women's stated preferences concerning this social dichotomy and their changing marriage strategies indicate that their greater use of German in interaction is one aspect of their general preference for the worker's way of life it symbolizes. Rather than simply isolating a linguistic correlate of sex, the present study suggests that women's speech choices must be explained within the context of their social position, their strategic life choices and the symbolic values of the available linguistic alternatives (language and sex roles; interactional analysis; social determinants of language shift; European bilingualism).

Over the past ten years the study of language in its social context has become a mature field with a substantial body of method and empirical results. As a result of this work we are arriving at new insights into such classical problems as the origin and diffusion of linguistic change, the nature of stylistic variation in language use, and the effect of class structure on linguistic variation within a speech community. Advances in sociolinguistics have been most evident in the study of co-variation between social context and the sound pattern of speech. The results reported in numerous monographs have laid the basis for substantial theoretical progress in our understanding of the factors that govern dialect variation in stratified communities, at least in its phonological aspect. The formulation of theories of the causes of phonological variation that go beyond guesswork and vague generalities appears at last to be possible. Therefore, we offer the following discussion, based on the material that is now available, as a contribution to the development of an explanatory theory of the mechanisms underlying social dialect variation. Although we shall state our views strongly, we know that they are far from definitive. We present them, not as positions to be defended at all costs, but as stimuli to further theoretical reflection in a field that has been, thus far, descriptively oriented.

Ongoing change in Wolof noun classification is traced by comparing nineteenth-century linguistic evidence with modern sociolinguistic data. On the rural scene, the noun class system today displays competing generalizations, associated with different social groups. Upwardly mobile middle-aged men of high caste tend to reduce the system, whereas other speakers tend to elaborate it. The source of both tendencies can be found in Wolof cultural assumptions relating social rank to norms of verbal conduct. In particular, an idea that linguistic ‘error’ can be appropriate for high-ranking persons underlies the reductionist tendencies of upwardly mobile speakers. Although the two tendencies compete on one level, on another they combine to form a more general systemic trend toward the incorporation of social features into Wolof syntax. (Sociolinguistic change, noun classification, norms for performance, ‘appropriate error’, criteria for inference of change; Wolof language of Senegal.)

Black English is increasingly recognized as a systemic, ‘rule-governed’, form of language (lect), like any other, at least among scholars. There is less adequate understanding of the meaning of Black English to those who use it. Such understanding requires ethnographic study of the place of Black English in the verbal repertoire of the community in which it is used. There has been some work in this area, but much of it has focused on aspects of Black English of interest for linguistic theory, e.g. the possibility of creole origins as a constraint on variability. Such studies have tended not to recognize the full scope and complexity of Black English. The aspects given attention have been those most exotic and remote from conventional English (for criticism on this point, see Abrahams (1972) and Wright (1972) ). While such studies have increased recognition and acceptance of Black English, the model in terms of which they have done so has been implicitly the ‘other culture’ model. There has been an appeal to the acceptance of cultural difference associated in our minds with Ruth Benedict's studies of Patterns of Culture, Margaret Mead's studies in Samoa and New Guinea, etc. This ‘other culture’ model is inadequate to the situation of the Black Community. There is partial truth to it; respect for the ways in which Black Americans may differ from ‘mainstream’ expectations is necessary. But social policy and action, especially in regard to education, cannot be based on so partial a truth. This became clear when well-meaning scholars attempted to introduce elementary school readers that used vernacular Black English. There was great protest from many in the Black community itself.

Sociolinguistic studies over the past ten years have demonstrated time and again how linguistic behavior changes as a person's social position changes, and how language is thus an excellent indicator of social status and social change.

Linguists and psychologists have noted the potential value of studying speech errors since the 1890s (Meringer & Mayer 1895; Freud 1938; Fromkin 1973; MacKay 1972). The reasoning has been that involuntary errors may lay bare certain aspects of the speech production system which are hidden in normal, errorless speech. Today we are closer than ever before to realizing this hope, because (a) we have more complete samples and analyses of spontaneous errors (Fromkin 1973; MacKay 1970; Garrett 1975) and (b) because of considerable success in recent years in attempts to elicit errors of varying complexity in the laboratory (MacKay 1971; Baars & Motley 1974; Motley & Baars 1976). This paper reports some extensions of the experimental approach, extensions which apparently enable us to elicit almost any arbitrary error at any level of complexity.

We regret several errors in Volume 6, Number 2. In addition to several errors of transliteration of the Russian in Krysin's review article:

p. 225 incorrectly cited the translator as Dr Philip Suelsdorff when his name is in fact Luelsdorff.

pp. 238 and 239, the second Kopylenko reference, the last Krysin reference, and the Krysin & Shemelev reference were all published in 1976.

p. 238, the following reference under Nikolsky should have been included:

— (1976). Sinxronnaja sociolinguistike. Moscow. [Synchronous sociolinguistics.]

p. 241, Roxencvejg, V. Ju. O jaxykovyx kontaktax was published in 1962. The Shweitzer reference and the volume in which it appeared were published in 1976. Also, the following reference under Shweitzer should have been included:

— (1976b). Sovremennaja sociolinguistike. Teorija, problemy, methody. Moscow. [Comptemporary sociolinguistics. Theory, problems, methods.]