Durkheim presented a definitive critique of reductionist explanations of social behavior. Social phenomena are "social facts" and these are the subject matter of sociology. They have, according to Durkheim, distinctive social characteristics and determinants, which are not amenable to explanations on the biological or psychological level. They are external to any particular individual considered as a biological entity. They endure over time while particular individuals die and are replaced by others. Moreover, they are not only external to the individual, but they are "endowed with coercive power, by . . . which they impose themselves upon him, independent of his individual will." Constraints, whether in the form of laws or customs, come into play whenever social demands are being violated. These sanctions are imposed on individuals and channel and direct their desires and propensities. A social fact can hence be defined as "every way of acting, fixed or not, capable of exercising on the individual an external constraint."

Although in his early work Durkheim defined social facts by their exteriority and constraint, focusing his main concern on the operation of the legal system, he was later moved to change his views significantly. The mature Durkheim stressed that social facts, and more particularly moral rules, become effective guides and controls of conduct only to the extent that they become internalized in the consciousness of individuals, while continuing to exist independently of individuals. According to this formulation, constraint is no longer a simple imposition of outside controls on individual will, but rather a moral obligation to obey a rule. In this sense society is "something beyond us and something in ourselves." Durkheim now endeavored to study social facts not only as phenomena "out there" in the world of objects, but as facts that the actor and the social scientist come to know.

Social phenomena arise, Durkheim argued, when interacting individuals constitute a reality that can no longer be accounted for in terms of the properties of individual actors. "The determining cause of a social fact should be sought among the social facts preceding it and not among the states of the individual consciousness." A political party, for example, though composed of individual members, cannot be explained in terms of its constitutive elements; rather, a party is a structural whole that must be accounted for by the social and historical forces that bring it into being and allow it to operate. Any social formation, though not necessarily superior to its individual parts, is different from them and demands an explanation on the level peculiar to it.

Durkheim was concerned with the characteristics of groups and structures rather than with individual attributes. He focused on such problems as the cohesion or lack of cohesion of specific religious groups, not on the individual traits of religious believers. He showed that such group properties are independent of individual traits and must therefore be studied in their own right. He examined different rates of behavior in specified populations and characteristics of particular groups or changes of such characteristics. For example, a significant increase of suicide rates in a particular group indicates that the social cohesion in that group has been weakened and its members are no longer sufficiently protected against existential crises.

In order to explain regular differential rates of suicide in various religious or occupational groupings, Durkheim studied the character of these groups, their characteristic ways of bringing about cohesion and solidarity among their members. He did not concern himself with the psychological traits or motives of the component individuals, for these vary. In contrast, the structures that have high suicide rates all have in common a relative lack of cohesion, or a condition of relative normlessness.

Concern with the rates of occurrence of specific phenomena rather than with incidence had an additional advantage in that it allowed Durkheim to engage in comparative analysis of various structures. By comparing the rates of suicides in various groups, he was able to avoid ad hoc explanations in the context of a particular group and instead arrive at an overall generalization. By this procedure he came to the conclusion that the general notion of cohesion or integration could account for a number of differing specific rates of suicide in a variety of group contexts. Groups differ in the degree of their integration. That is, certain groups may have a firm hold on their individual members and integrate them fully within their boundaries; others may leave component individuals a great deal of leeway of action. Durkheim demonstrated that suicide varies inversely with the degree of integration. "When society is strongly integrated, it holds individuals under its control." People who are well integrated into a group are cushioned to a significant extent from the impact of frustrations and tragedies that afflict the human lot; hence, they are less likely to resort to extreme behavior such as suicide.

For Durkheim, one of the major elements of integration is the extent to which various members interact with one another. Participation in rituals, for example, is likely to draw members of religious groups into common activities that bind them together. Or, on another level, work activities that depend on differentiated yet complementary tasks bind workers to the work group. Related to the frequency of patterned interaction is a measure of value integration, that is the sharing by the members of values and beliefs. In collectivities where a high degree of consensus exists, there is less behavioral deviance than in groups in which consensus is attenuated. The stronger the credo of a religious group, the more unified it is likely to be, and therefore better able to provide an environment that will effectively insulate its members from perturbing and frustrating experiences. Yet Durkheim was also careful to point out that there are special cases, of which Protestantism is the most salient, in which the credo of the group stresses a shared belief in individualism and free inquiry. Protestantism "concedes a greater freedom to individual thought than Catholicism . . . it has fewer common beliefs and practices." In this case, higher rates of such deviant behavior as suicide cannot be explained as a lack of consensus, but as a response to the group-enjoined autonomy of its members.

The difference between value consensus and structural integration can now be more formally approximated in terms of Durkheim's own terminology. He distinguished between mechanical and organic solidarity. The first prevails to the extent that "ideas and tendencies common to all members of the society are greater in number and intensity than those which pertain personally to each member. This solidarity can grow only in inverse ration to personality." In other words, mechanical solidarity prevails where individual differences are minimized and the members of society are much alike in their devotion to the common weal. "Solidarity which comes from likeness is at its maximum when the collective conscience completely envelops our whole conscience and coincides in all points with it." Organic solidarity, in contrast, develops out of differences, rather than likenesses, between individuals. It is a product of the division of labor. With increasing differentiation of functions in a society come increasing differences between its members.

Each element in a differentiated society is less strongly tied to common collective routines, even though it may be bound with equal rigor to the differentiated and specialized tasks and roles that characterize systems of organic solidarity. While the individual elements of such a system have less in common, they are nevertheless much more interdependent than under mechanical solidarity. Precisely because they now engage in differentiated ways of life and in specialized activities, the members are largely dependent upon one another and networks of solidarity can develop between them. In such systems, there can be some release from external controls, but such release is in tune with, not in conflict with, the high degree of dependence of individuals on their fellows.

In his earlier work, Durkheim stated that strong systems of common belief characterize mechanical solidarity in primitive types of society, and that organic solidarity, resulting from the progressive increase in the division of labor and hence increased mutual dependence, needed fewer common beliefs to tie members to this society. He later revised this view and stressed that even those systems with a highly developed organic solidarity still needed a common faith, a common conscience collective, if they were not to disintegrate into a heap of mutually antagonistic and self-seeking individuals.

The mature Durkheim realized that only if all members of a society were anchored to common sets of symbolic representations, to common assumptions about the world around them, could moral unity be assured. Without them, Durkheim argued, any society, whether primitive or modern, was bound to degenerate and decay.

From Coser, 1977:129-132.