My label of ‘social psychologist’ requires that I have some knowledge about the perception of other people, and some recent studies of judgements about others have been set in the wider context of examining how far and in what ways people take the needs, wishes, and expectations of others into account. Since this seminar is organized for the immediate benefit of those engaged in surveys of languages and language usage in East Africa and the ultimate benefit of the African countries concerned, it seems most appropriate to attempt to re-organize our work in England with these considerations in mind. One result of this is that I have eschewed an exposition of Bernstein’s theoretical framework and relevant evidence. Instead I have selected two main topics, one methodological, the other educational, which may bear directly upon the African surveys. The methodological section begins with a short history, but develops the theme of the usefulness of an approach to sociolinguistics which never forgets that language is used to serve many functions and always remembers that it is only one vehicle for communication among others. Even within the limited range of speech samples we have examined, we can show that the speech used varies across situations and persons, while the detection of these differences requires a prior consideration of functions and semantics, and an examination of grammar at 76appropriate levels of analysis. The differences between persons show, not only that performance is a function of social variables, but also that the relationships are not necessarily monotonic and the influences not necessarily simple. Although we cannot yet provide a general description of these phenomena, our examples may be informative guidelines for others.