Although the use of the concepts of function and functionalism is usually associated with the work of Talcott Parsons in modern sociology, there is a long tradition of functional explanation in studying societies, and a form of modified functionalism is now undergoing a revival. Among the founders of sociology, Émile Durkheim is most closely associated with functionalism, since he often employs analogies with biology. The most prominent of these is an organic analogy, in which society is seen as an organic whole, each of its constituent parts working to maintain the others, just as the parts of the body also work to maintain each other and the body as a whole. This idea is basic to his conception of organic solidarity. Durkheim did distinguish between functional and historical explanations and recognized the need for both. A functional explanation accounts for the existence of a phenomenon or the carrying out of an action in terms of its consequences—its contribution to maintaining a stable social whole. For example, a functional explanation of the existence of crime is that it serves to mark out and reinforce (through punishment) the boundaries of socially acceptable behaviour, so that crime is therefore a normal feature of social life. Similarly, religious institutions serve to generate and maintain social solidarity. Historical explanations are an account of the chronological development of the same phenomena or actions. Modern functionalism, through the work of Robert Merton, distinguishes between manifest functions (intended consequences or consequences of which the participants are aware) and latent functions (unintended consequences of which the participants are unaware). The latter may or may not be generally beneficial.
There has been a strong and often explicit functionalism present in sociology and social anthropology throughout most of this century. There has also been an implicit functionalism in the more determinist forms of Marxist theory, where so-called surface features of the social formation (such as political systems, ideologies, and trade unions), are seen as produced by, in order to maintain, the underlying relations of production. However, probably the most famous functionalist analysis in sociology is the so-called functional theory of social stratification offered by Kingsley Davis and Wilbert Moore, although Davis also wrote a functionalist textbook, Human Society (1949), and made a spirited defence of functionalism in his Presidential Address to the American Sociological Association in 1959 (see ‘The Myth of Functional Analysis as a Special Method in Sociology and Anthropology’, American Sociological Review, 1959). Herbert J. Gans 's celebrated essay on ‘The Positive Functions of Poverty’ (American Journal of Sociology, 1972), said by some to have been written as a parody of structural-functionalism, is actually a superb example of ideologically neutral functional analysis.
In the late 1960s functionalism came under sustained attack from various sources. It was argued that this approach could not account for social change, or for structural contradictions and conflict in societies, and that its reliance on stability and on the organic analogy rendered it ideologically conservative: it became fashionable to refer to functionalism as consensus theory. This particular group of criticisms is not entirely accurate. Parsons's evolutionary theory, seeing historical development in terms of the differentiation and reintegration of systems and sub-systems, can account for change and at least for temporary conflict until the reintegration takes place. The existence of functional explanations in Marxism indicates that they can exist alongside a recognition of contradictions in social systems. Durkheim himself was able to combine functionalist explanations with a sometimes radical form of guild socialism.