Sociolinguistics studies the ways in which language interacts with society. It is the study of the way in which language's structure changes in response to its different social functions, and the definition of what these functions are. 'Society' here is used in its broadest sense, to cover a spectrum of phenomena to do with race, nationality, more restricted regional, social and political groups, and the interactions of individuals within groups. Different labels have sometimes been applied to various parts of this spectrum. 'Ethno-linguistics' is sometimes distinguished from the rest, referring to the linguistic correlates and problems of ethnic groups - illustrated at a practical level by the linguistic consequences of immigration; there is a language side to race relations, as anyone working in this field is all too readily aware. The term 'anthropological linguistics' is sometimes distinguished from 'sociological linguistics', depending on one's particular views as to the validity or otherwise of a distinction between anthropology and sociology in the first place (e.g. the former studying primitive cultures, the latter studying more 'advanced' political units). Usage of British and American scholars differs considerably in this respect. 'Stylistics' is another label which is sometimes distinguished, referring to the study of the distinctive linguistic characteristics of smaller social groupings (such as those due to occupational or class differences). More usually, however, stylistics refers to the study of the literary expression of a community, using linguistic methods. None of these labels has any absolute basis: the subject-matter of ethnolinguistics gradually merges into that of anthropological linguistics, that into sociological linguistics, and that into stylistics, and the subject-matter of social psychology. The kinds of problem which turn up are many and various, and some have been illustrated in Chapter i, which was very much concerned with the role of language in society. They include: the problems of communities which develop a standard language, and the reactions of minority groups to this (as in Belgium, India, or Wales); the problems of people who have to be educated to a linguistic level where they can cope with the demands of a variety of social situations; the problems of communication which exist between nations or groups using a different language, which affects their 'world-view'; the problems caused by linguistic change in response to social factors; the problems caused (and solved) by bilingualism or multilingualism; the problems caused by the need for individuals to interact with others in specific linguistic ways (language as an index of intimacy or distance, of solidarity, of prestige or power, of pathology, and so on). I am not arguing that sociolinguistics by itself can solve problems such as these; but it can identify precisely what the problems are (this is sometimes a major task in itself), and obtain information about the particular manifestation of a problem in a given area, so that possible solutions can thereby be hastened.
One thing is clear. There is little chance of solving any of these problems until certain basic principles about the relationship of language to society have been established, and accurate techniques of study developed. And so far, there are many basic issues about which there is much controversy - for example, the extent to which our social background determines our linguistic abilities, or the rationale on which multilingual individuals use their different languages for different social purposes. There are of course innumerable facts to be discovered, even about a language as well investigated as English, concerning the nature of the different kinds of English we use in different situations - when we are talking to equals, superiors or subordinates; when we are 'on the job'; when we are old or young, upper class or lower class, male or female; when we are trying to persuade, inform or bargain; and so on. An informal definition of sociolinguistics highlights this concern to get even the most elementary of descriptive information down on paper: 'Who can say what, how, using what means, to whom, when, and why?' If we knew all these factors, we would know a great deal about social problems. These days sociolinguistics has progressed far in accumulating its own data in order to answer these questions.
To analyse a problem sociolinguistically implies being able to analyse it linguistically. Sociolinguistics makes use of the findings of linguistic theory and description in its work; and in one sense its success is dependent on success in 'pure linguistics'. On the other hand, the nature of its subject-matter means that there will arise a great deal which will be both theoretically and methodologically novel - explanatory constructs of one kind or another which are not constructs of either linguistics or sociology, but a derivative of both. One example of this is the notion of 'interference', that is, linguistic disturbance which results from two languages (or dialects) coming into contact in a specific situation. The problem of interference is not something which linguistics, or any other subject, on its own, could handle. There has been some debate as to whether the existence of uniquely sociolinguistic problems of this kind requires the establishment of a quite independent discipline, with a theoretical identity and methodology of its own, or whether the dependence on linguistics in its general sense is so fundamental that such a prospect is impossible. This is an issue which will doubtless continue to be discussed for some time. Meanwhile, it is the case that for practical purposes (as in teaching linguistics) most courses would not make a clear-cut distinction, but would consider the study of sociolinguistics to be an essential part of the explanation of the subject as a whole.
An even stronger link is argued these days for my second example of interdisciplinary overlap, psycholinguistics. The relation of linguistics to psychology has been the source of some heated discussion of late, largely due to Chomsky's particular emphasis on this question. His view of linguistics, as outlined for instance in his book Language and Mind, is that the most important contribution linguistics can make is to the study of the human mind; and that linguistics is accordingly best seen as a branch of cognitive psychology. This is not an altogether surprising thing in view of the mentalistic claims of parts of his theory (cf. p. 103) and his particular views on the nature of language acquisition in children. But it is an extreme view, which most linguists at the present time do not share. On the other hand, no one would want to deny the existence of strong mutual bonds of interest operating between psychology and linguistics. The extent to which language mediates or structures thinking, the extent to which talk about language 'simplicity' or 'complexity' can be given any meaningful psychological basis, the extent to which language is influenced by and itself influences such things as memory, attention, recall and constraints on perception, and the extent to which language has a central role to play in the understanding of human development are broad illustrations of such bonds.
Psycholinguistics as a distinct area of interest developed in the early sixties, and in its early form covered the psychological implications of an extremely broad area, from acoustic phonetics to language pathology. Nowadays, certain areas of language and linguistic theory tend to be concentrated on by those who call themselves psycholinguists, and most of them have been influenced by the development of generative theory. The most important area is the investigation of the acquisition of language by children. Here, there have been many studies of both a theoretical and a descriptive kind. The descriptive need is prompted by the fact that until recently hardly anything was known about the actual facts of language acquisition in children, in particular about the order in which grammatical structures were acquired. Even elementary questions such as when and how children develop their ability to ask questions syntactically, or when they learn the inflectional systems of their language, went unanswered. And a great deal of work has gone on recently into the methodological and descriptive problems involved in obtaining and analysing information of this kind.
The theoretical questions have focused on the issue of how we can account for the phenomenon of language development in children at all. Normal children have mastered most of the structure of their language by the age of five. The generative approach argued against the earlier behaviourist assumptions that it was possible to explain language development largely in terms of imitation and selective reinforcement. It asserted that it was impossible to explain the rapidity or the complexity of language development solely in terms of children imitating the language used by the people around them. And as a result of the arguments supporting this assertion, it would now be generally agreed that imitation alone is not enough. Imitation is an important factor in the development of language (cf. p. 46), but it cannot be the major one, and thus the basis of any theory of language acquisition, because there is too much of central importance in language which is not amenable to direct observation, and thus not imitatable - the various meaning-relations between sentences or parts of sentences, for instance, or, more generally, the abstract knowledge of the grammatical rules of their language which adults have as part of their competence. All normal children come to develop this abstract knowledge for themselves; and the generative approach argues that such a process is only explicable if one postulates that certain features of this competence are present in the brains of children right from the beginning. In other words, what is being claimed is that children's brains contain certain innate characteristics which 'pre-structure' them in the direction of language learning. To enable these innate features to develop into adult competence, children must be exposed to human language, i.e. they must be stimulated in order to respond. But the basis on which they develop their linguistic abilities is not describable in behaviourist terms.