Paradoxically, American sociology, unlike its European counterpart, has been marked by an individualistic (psychological) orientation, even though early sociologists fought to establish a discipline distinct from psychology. Most specialized research in American sociology still uses the individual as the unit of analysis. The standard practice is to collect data from or about individuals, categorize their social characteristics into “groups,” and relate them to other categories of individuals such as income classes, occupations, and age groups. These intergroup relations are often examined with complex statistical tools. This practice is not generally recognized as social-psychological in nature, yet neither is it regarded as social structural analysis. (See social structure.) Only a minority of sociologists in fields such as demography, human ecology, and historical or comparative institutional study use actual groups, organizations, and social structures as units of analysis.

As the field developed in the United States, many early 20th-century sociologists rejected instinctivist psychology and the classical behaviourism of John B. Watson. One group, however, emphasized the study of individuals in an approach called symbolic interaction, which took root at the University of Chicago early in the 20th century and remains prominent in contemporary sociology. John Dewey, George H. Mead, and Charles H. Cooley argued that the self is the individual’s internalization of the wider society as revealed through interaction, the accumulated perceptions of how others see them. In other words, the mind and human self are not innate human equipment but constructions of the “person” (the socialized individual) derived from experience and intimate interpersonal interaction in small groups. This constructed self, however changing, functions as a guide to social behaviour. Social reality is thus made up of constructed symbols and meanings that are exchanged with others through daily interaction.

William I. Thomas and Ellsworth Faris used symbolic interaction theory to guide their empirical research in the tradition of Robert E. Park and Ernest W. Burgess by using personal documents, life histories, and autobiographies. The two revealed how people attach meanings to their experience and to the broader social world. This research tradition was enriched after 1960 by several innovations. The most sophisticated small-group research was devised by Erving Goffman in The Presentation of Self in Everyday Life (1959). Goffman insisted that the most meaningful individual behaviour occurs in the chance, intimate encounters of each day. These encounters include greeting people, appearing in public, and reacting to the physical appearance of others. Such encounters have structures of their own that can be researched by carefully constructing the “frames” (points of reference) people use to interpret and “stage” interactions. The structures are thought to represent true reality as opposed to the artificially constructed concepts that sociologists impose on the subjects they study.

In Studies in Ethnomethodology (1967), Harold Garfinkel coined the term ethnomethodology to designate the methods individuals use in daily life to construct their reality, primarily through intimate exchanges of meanings in conversation. These constructions are available through new methods of conversational analysis, detailed or “thick” descriptions of behaviour, “interpretive frames,” and other devices. Proponents of this view have favoured the work of earlier European phenomenology, Verstehen (historical understanding), and interpretive sociology. More recently, qualitative sociologists have drawn on French structuralism, poststructuralism, and postmodernism to emphasize ways the “deeper” sources of hidden meanings in culture and language can affect the behaviour of individuals or of whole societies.

Since World War II, sociology has exported much of its theory, methodology, and findings to other divisions of the university, sometimes to its disadvantage. The study of human relations and formal organizations was transferred to business schools. The study of socialization, institutions, and stratification was absorbed by departments of education. Outside the university, the empirical methods and sociological theory prompted government agencies to adopt a behavioral perspective. Economists widened the scope of their research by introducing social variables to the analysis of economic behaviour. In short, although contemporary sociology is divided, it remains a vibrant field whose innovations contribute to its own development and that of social science in general.

Methodological considerations in sociology

Much 19th-century sociology had no system for gathering and analyzing data, but over time the inadequacies of speculative methods became increasingly evident, as did the need for obtaining reliable and verifiable knowledge. Like his contemporaries, Herbert Spencer assembled vast stores of observations made by others and used these to illustrate and support generalizations he had already formulated. Early social surveys like those conducted by Charles Booth in a monumental series on the social problems of London produced masses of data without regard to their theoretical relevance or reliability. Frédéric Le Play made similar use of the French case studies he drew on for his extensive investigations of family budgets.

Early exploitation of statistical materials, such as official records of birth, death, crime, and suicide, provided only moderate advances in knowledge. Data were easily manipulated, often to support preconceived ideas (the status quo). Among the most successful of such studies was that on suicide rates by Durkheim in Le Suicide (1897). Moreover, his Rules of Sociological Method (1895) had begun to meet the standards of scientific inquiry. In gathering data on suicides, Durkheim considered the social characteristics of individuals (e.g., religious affiliation, rural-urban residence) that reflected the degree of their social integration in the community, and he related these variables statistically.