Broadly speaking, ‘functionalism’ refers to a range of theories in the human sciences, all of which provide explanations of phenomena in terms of the function, or purpose, they purportedly serve. In the period spanning the last decades of the nineteenth century and the first half of the twentieth, virtually every human science generated a school that identified itself as functionalist, and in nearly every instance that school dominated its discipline for a time. Darwinian evolutionary theory provided the initial impetus to functionalist reasoning. But Darwin’s multifarious argument admits of variable interpretation, so different constructions of his model yielded varieties of functionalism. The earliest schemes, those of psychology and economics, were promulgated at the turn of the century. These were not equally important, however; functionalist psychology was extremely influential, while functionalist economics was nearly inconsequential. But in both, the individual was the basic unit of analysis, and individual action was conceptualized in terms of recursive processes of evolutionary adaptation. That is, both functionalist psychology and economics relied on an interpretation of the inherent nature of the human organism, and constituted fundamentally historical approaches to explanation.
In the late 1920s, a rather different type of functionalism became the dominant paradigm among British social anthropologists, thereafter diffusing to anthropologists elsewhere, as well as to sociologists – who judged that observation of non-Western societies had revealed the fundamental constituents of human sociability, unconfounded by the workings of ‘organized state machinery’ (Parsons 1934: 230). (After World War II, this type was embraced by practitioners of other disciplines, perhaps most notably political scientists studying colonial societies then making the transition to independence, but these contributed little to its elaboration.) In its original formulation if not necessarily its subsequent permutations, this scheme was informed by Darwinian reasoning, but differed from its precursors in psychology and economics in its basic unit of analysis: functionalist anthropology and sociology considered the group, not the individual, judging that it was in groups that humans withstood processes of natural selection. And the group had to be analysed as a social, not a biological, entity. Because all groups possessed roughly equivalent human resources -individuals differing in talent and temperament -variable natural endowments did not explain given groups’ survival. Human adaptation was effected through social organization. Thus, notwithstanding some functionalists’ professed concern with individual personality structure and volitional action, in analysis of this type individuals were judged derivative creatures of their social orders – practically epiphenomenal as individuals (see Radcliffe-Brown 1949; Wrong 1961).
For anthropologists and sociologists, the point of functionalist investigation was to identify the standardized habits that maintained the social organism in a condition of dynamic equilibrium – the ‘more or less stable social structures’ regulating individuals’ relations ‘to one another, and providing such external adaptation to the physical environment, and such internal adaptation between the component individuals or groups, as to make possible an ordered social life’ (Radcliffe-Brown 1932: 152). The historical antecedents of any given social order were of no interest: they did not explain either the meaning of its practices for those who sustained them in the present or the roles these practices played in maintaining the social organization as a whole, roles which exhibited the general properties of social life the functionalist sought. Moreover, scientific inquiry by definition entailed direct observation. Thus, the very designation ‘ethnology’, which denoted efforts to reconstruct the histories of peoples who left no written records, became a term of opprobrium for functionalists: ethnological findings were at best descriptions of probable pasts (and more likely conjectural ones), yielding accounts of idiosyncratic experiences rather than identification of scientific regularities (see Radcliffe-Brown 1932: 144-8).
The politics of knowledge
Functionalism became the predominant analytic mode in anthropology and sociology following fierce disputes during the 1920s and 1930s; and when functionalism came under concerted attack in the late 1950s, it ceased to be the discipline’s reigning theory only after extraordinarily acrimonious debate. Indeed, the issues raised in the latter controversy have yet to be thoroughly resolved. The social sciences’ current theoretical irresolution may be variously construed. Perhaps it indicates that the disciplines have abandoned naive scientism, achieving intellectual maturity; if, as E.E. Evans-Pritchard argued as early as 1950 (1951), the study of social behaviour has closer affinities with the humanities than with the natural sciences, then the human sciences should share with the humanities a tolerance for interpretative pluralism. Or perhaps the lack of paradigmatic unity is lamentable, the product of the crisis of identity the disciplines have been suffering since roughly 1960; if the mission of the human sciences is to replicate the pattern of inquiry characteristic of the natural sciences (as authoritatively defined by Kuhn 1962), then their research must be guided by some (almost any) dominant paradigm so that their findings may be cumulative. Thus, recent theoretical controversy has embroiled practitioners in consideration of the fundamental constitution of the human sciences.
Moreover, debates over the merits of functionalist analysis have been significant because they have constituted sites for convergence of the disciplines’ academic politics and intellectual conflicts. Both the functionalists of the 1920s and 1930s and the anti-functionalists of the 1950s and 1960s were younger practitioners determined to discredit their elders’ theoretical commitments. And at stake in both controversies were also the disciplines’ material resources — positions in university departments, as well as the funds for research dispensed by government agencies and private philanthropies (see Kuklick 1992: 208—16). We cannot overestimate the importance of academic political considerations in prolonging theoretical dispute, for in the 1950s and 1960s members of the social scientific establishment were able to frame the debate over functionalism in terms that still persist. Their position, perhaps most clearly articulated by the American sociologist Kingsley Davis, was that non-functionalist scholars were not really social scientists of any sort, for the functionalist creed entailed only the explanatory habit of recognizing interdependence among any given social order’s constituent parts (1959). That Davis’s formulation was widely acknowledged as authoritative may be judged from its reverential treatment in such definitive texts as the International topic of the Social Sciences, which reported that functionalism as he defined it was ‘the most widely used type’ of functionalist analysis, and was so pervasive in anthropology and sociology that ‘it is misleading to distinguish it by a special name’ (Cancian 1968: 29—30).
Davis’s definition of functionalism was an invitation to ignore differences of social scientific opinion far more significant than those of the 1950s and 1960s. If concern to establish patterns of correlation among elements of a social order denotes functionalist analysis, then such figures as the British evolutionist anthropologist E.B. Tylor should be counted among its proponents -since he developed a method for identifying such patterns – although his was one of the very approaches functionalism was intended to supplant. Moreover, Davis’s own analyses provided compelling illustrations of the very interpretative defects functionalism’s critics identified: the tendency to confound the actual and the optimal, as well as to presume consensus and stability in the absence of incontrovertible evidence of societal breakdown. Unless we disaggregate his (re)inter-pretation of the model from its characteristics during its era of paramountcy, we will be unable to appreciate the distinctive properties of functionalist theory. And from the points raised in the theoretical debate of the 1960s, we may deduce that any useful definition of functional-ism has to specify at least two properties. One, it is one example of a theoretical type often termed ‘grand theory’ — a category that admits of Tylor’s evolutionism among other theories — a scheme that is intended to comprehend the behaviour of peoples at all times and places. Two, it is holistic analysis of a sort that Tylor’s most certainly was not. That is, it asks a specific question about every habitual practice: How does this contribute to the maintenance of the whole?
Since the 1960s, no grand theory has compelled collective effort among either anthropologists or sociologists. Indeed, it has become a commonplace that their fields no longer exist as such, having dissolved into specialized sects devoted to specific subject areas and/or documentation of particular perspectives. But many still insist that the social sciences ought to be guided by some dominant paradigm — which must represent a comprehensive scheme; their view accounts for the disciplines’ reluctance to repudiate functionalism definitively, and, indeed, for some significant recent efforts to rehabilitate it. It is important to stress, however, that many disciplines were seized by paradigmatic crisis at the same time as anthropology and sociology were — indicating that irresolution within the social sciences does not necessarily derive from distinctive problems raised by the study of human behaviour; perhaps in consequence of the embattled state of the university in the 1980s and 1990s, and particularly of the politics of research funding, practitioners of such fields as biology and chemistry also see their enterprises as fragmented. Furthermore, the social scientific consensus of the era of functionalist dominance may have been more apparent than real, an artefact of prosperous conditions. Until 1960, British social anthropologists were beneficiaries of relatively generous funds from the Colonial Social Science Research Council, and researchers of somewhat disparate views were able to co-exist peaceably; and the incompatibility of functionalist principles with the sociological nominalism of survey research went unremarked among their sociologist contemporaries (see Platt 1986) because these were able to appeal to financial patrons with the claim that sociology was truly rigorous science, its surveys providing empirical documentation for its theory.
An origin story
Virtually all functionalists have invoked Emile Durkheim’s The Rules of Sociological Method (1938 [1895]) as their first programmatic manifesto, placing its author at the head of their intellectual lineage, if not necessarily identifying him as their sole progenitor. In no small degree, functionalists’ equation of Durkheimian sociology with their scheme constitutes historical revisionism. That is, as his contemporaries knew, although we have been wont to forget, Durkheim enjoyed an extraordinarily diverse readership during his lifetime, and the proto-functionalists of the turn of the century represented only a fraction of his audience. Moreover, in its mature form func-tionalism represented the confluence of various theoretical and methodological trends: anthropologists and sociologists differed from one another in the routes they took to reach the functionalist position they eventually came to share; and the intellectual trajectories that led to functionalism differed from country to country as well as from discipline to discipline. Nevertheless, at last, if not at first, a stylized version of Durkheim’s scheme provided many functionalists a covering rationale for their enterprise. And once social scientists had devised their revisionist origin myth, Durkheim’s ideas became in fact central to self-conscious articulation and dissemination of functionalist argument. Indeed, the importance of these ideas as a legitimating creed for the functionalist school in British anthropology, at least, may be judged from the self-conscious identification of its creators and followers as ‘sociologists’.
Durkheim provided nothing less than a justification for the existence of sociology as a separate discipline, arguing that it addresses a ‘new variety of phenomena’, qualitatively different from those of other disciplines. These are ‘social facts’, the only data to which ‘the term social ought to be applied’ — ‘ways of acting, thinking, and feeling, external to the individual, and endowed with a power of coercion, by reason of which they control him’ (1938 [1895]: 3). Thus, Durkheim dismissed the so-called reductionist research programme, which was widely endorsed by scientists in his day (and many find compelling in ours), premised on the assumption that scientific inquiries become more rigorous in direct proportion to the degree to which the phenomena they treat are understood in terms of their presumed underlying causes — sociology translated into psychology, psychology into biology, biology into chemistry, chemistry into physics, and so on. Indeed, he questioned the value of reductionism for any sort of scientific inquiry: because the whole is invariably greater than the sum of its parts, even the chemist considering the inorganic molecular constituents of organic matter should perceive that ‘these molecules are in contact with one another, and this association is the new phenomena which characterize life, the very germ of which cannot possibly be found in any of the separate elements’ (1938 [1895]: 102). The apparently counter-intuitive, sociological interpretation of such data as suicide rates exemplified the explanatory power of the Durkheimian paradigm: a reductionist, psychological account of suicide seems mandated because it seems a quintessentially individual act, the product of an individual’s ‘temperament, character, antecedents, and private history’(1951 [1897]: 46); yet, in both their frequency and motivational type, suicides are consequences of broad social trends. Certainly, Durkheim was not entirely consistent. He allowed that fundamental ‘sociological laws can only be a corollary of the more general laws of psychology’, because ‘the ultimate explanation of collective life will consist in showing how it emanates from human nature in general’ (1938 [1895]: 98, emphasis mine). And Durkheim also made assumptions about the relationship between the sociological phenomena he observed and human biological nature, tacitly resting his generalizations on Lamarckian biology.
Durkheim’s anti-reductionist pronouncements served as rallying cries for anthropologists and sociologists, and had perhaps particularly strong force for the British social anthropologists who developed functionalism. We should note that their two founding figures, Bronislaw Mal-inowski and A.R. Radcliffe-Brown, were joined in accepting Durkheim’s fundamental premises, notwithstanding their critiques of each other’s work (and the very real differences between them). Malinowski taught his students to repudiate Radcliffe-Brown’s ‘dangerous and one-sided … sociological determinism of culture’ (quoted in Kuklick 1992: 120), and Radcliffe-Brown charged that Malinowski was not a genuine functionalist because he invoked the biological needs of individual human beings (1949). Yet, in his Argonauts of the Western Pacific (1922), which together with Radcliffe-Brown’s The Andaman Islanders (1922) heralded the emergence of the functionalist school, Malinowski proclaimed that ‘as sociologists, we are not interested in what A or B may feel qua individuals . only in what they feel qua members of a given community.
Installing Durkheim at the head of function-alism’s lineage required a somewhat revisionist construction of his research programme, however, for he was an evolutionist, concerned to identify the developmental trajectory that led to the emergence of modern society. He explicitly decried sociologists’ assumption that ‘they have accounted for phenomena once they have shown . what social needs they satisfy . To show how a fact is useful is not to explain how it originated or why it is what it is’ (1938 [1895]: 89, 90). Observing that a given practice could assume variable significance for social actors over time, he concluded that attention to the processes of historical transformation is essential (1938 [1895]: 91). But his functionalist followers judged that anthropology could not become ‘the science it should be’ if it persisted in the search for origins — and that Durkheim himself was ‘misled’, rendering his theory in a form ‘which has caused it to be misunderstood by many of his readers’, and clinging to ‘some of the ideas and some of the terminology of the older social anthropology’ (Radcliffe-Brown 1932: 153).
Functionalism embattled
By the late 1940s, a standard litany of objections to functionalism was developing. The functional unity of society should not be assumed (and, indeed, could not be assumed when the society in question was a Western one); in the study of any given society, integration was a variable to be measured by empirical test. Routinized practices were likely to be functional for some members of a society and dysfunctional for others — and some practices might have no meaningful consequences for a society whatsoever. That a given practice could have variable significance for social actors indicated that patterns of conflict might be inherent in a social order, and that significant social change might derive from endogenous as well as exogenous factors. When such criticisms were made in the period immediately following World War II, however, they usually represented efforts to render functional-ism more rigorous. The habit of explaining all observed practices as indispensable for the maintenance of the whole, for example, could be identified as merely the product of conceptual confusion, rather than a telling illustration of the defects of the functionalist mode of analysis per se (see Merton 1968 [1949]: 76-93).
Perhaps more important to the future debates over functionalism was the charge levelled in the 1940s that it conveyed an inherently conservative political message, that its analyses of thoroughly integrated communities suggested that any sort of change in them was inadvisable. There was an irony in this accusation, for as political argument functionalism arguably began as social criticism – as an elaboration of a vision of optimal social order shared by many European and American intellectuals in the era of World War I, described in such tracts as R.H. Tawney’s The Acquisitive •Society (1921), a meritocratic polity in which individuals and groups would be rewarded in direct proportion to their contributions to the collective good. Post-World War II critics could, however, point to presumptive evidence that functionalism was the ideology of the status quo: the rhetoric of its chief spokesmen. In no small part, functionalism prevailed in the disciplines because its entrepreneurs made calculated appeals to powerful patrons. By arguing that only the functionalist brand of anthropology could serve the practical needs of colonial administrators, Malinowski became the broker of much of the research support available to anthropologists during the interwar period, eliminating his rivals from the field. Nevertheless, his consummate grantsman-ship must be recognized as just that. Malinowski and his students were highly critical of all manner of colonial practices. Furthermore, colonial administrators made little use of functionalist anthropology — and frequently suspected that the functionalists who came to do fieldwork in their jurisdictions were fomenting discontent among colonial subjects (Kuklick 1992: 182-241).
A similar story can be told for sociology. In the United States, such figures as Talcott Parsons persuaded Congress that the national interest required inclusion of the social sciences in the National Science Foundation established after World War II, arguing that the sociologist’s role was analogous to that of a physician – that of an objective professional dedicated to curing social pathology. And in the postwar decades, those universities which were centres of functionalist argument consumed a disproportionate share of the funds available from both government sources and private philanthropies (see Buxton 1985: 117- 25; Geiger 1988). Privileged access to funding did not enable functionalist sociologists to police their field as ruthlessly as social anthropologists: because few aspirant anthropologists can finance their own field-work – which functionalists made a necessary prelude to an anthropological career – the discipline’s gatekeepers have been those who hold the purse strings; by contrast, sociological research has taken many forms (particularly in the large and diversified American university system), a number of which can be pursued without lavish funding. But sociologists’ position in the disciplinary prestige hierarchy has tended to be a straightforward function of their ability to attract research support. It should not be presumed, however, that 1950s functionalist sociology invariably assumed the Panglossian posture exemplified by American interpretations of social stratification; by contrast, at this time British sociologists undertook to expose the rou-tinized practices that perpetuated inequities in the British class system — and they counted among their most important mentors Edward Shils, an American Parsonian who maintained a recurrent presence in British academe (Halsey 1982).
During the 1960s and 1970s, the chorus of functionalism’s critics swelled, reiterating earlier arguments – in louder tones. The usual interpretation of their motivation has become a cliche, but is a reasonable explanation nonetheless: in every country in which functionalism had dominated social scientific inquiry, ideological differences grounded in cleavages of class, ethnicity, and generation belied the consensual model of society embodied in functionalist analysis. And the crisis of authority in Western democracies, coupled with the politics of decolonization in erstwhile subject territories, facilitated recognition of routinized conflicts in even the most apparently stable of societies – the non-literate, technologically underdeveloped polities on which the functionalist model had been based.
Functionalism’s critics in the 1990s
Functionalism could withstand the critiques of the post-World War II generation, with relatively minor modifications. The accusation that functionalism had constructed an ‘oversocialized conception of man’ (Wrong 1961) by confounding ‘what A or B may feel qua individuals’ with ‘what they feel qua members of a given community’ (Malinowski) could be countered with the observation that many functionalists – such as Malinowski – had never made this equation. The charge that functionalism exaggerated social integration could be met with attention to structured patterns of conflict within a social order. In the 1980s and 1990s, however, the very idea of a coherent social order — one which may be understood ‘as a whole composed of parts’ – became suspect; instead, social scientists posited ‘lasting incongruities between actors, others, and third parties in their construction of the meaning of events’ rather than ‘norms and shared ideas [that serve] as blueprints for acts’, so that what order obtains is an ‘emergent property’ of individuals’ interactions — and ‘differently positioned persons … live together in differently constructed worlds’ (Barth 1992: 19, 23, 24 and passim). Indeed, even ‘unitary accounts of the person have . become deeply problematic’ (Lave et al. 1992: 257). When social scientists emphasize historical contingency, situa-tional particularity, and conceptual disjunction, no sort of grand theory can seem plausible.
