Rules of Sociological Method

Émile Durkheim first published The Rules of Sociological Method in 1895. A second edition appeared in 1901 with a new preface responding to critics. This volume includes the full text of the second edition along with subsequent writings, debates, short notes, and letters in which Durkheim clarified, modified, and defended his positions.

The book functions as a manifesto for sociology, intended to establish its credentials as a natural science with its own distinct subject matter and rules of method. In the preface to the first edition, Durkheim warns that a science of societies must challenge accepted opinions, just as all genuine scientific discovery does. He anticipates controversy over his treatment of crime as a normal feature of collective life and preemptively counters: The normality of crime does not preclude the normality of punishment, since both are universal. He accepts the label of rationalist, defining his goal as extending scientific rationalism to human behavior by reducing it to relationships of cause and effect. In the preface to the second edition, he addresses the hostile reception the book provoked. He insists he never claimed that social facts, the external and coercive phenomena he identifies as sociology's subject matter, are material things, only that they are things in a different way: objects of knowledge that must be studied from the outside through observation rather than introspection. He proposes the word "institution" to express the special mode of existence of social facts, defining sociology as "the science of institutions, their genesis and their functioning" (15).

The book's six chapters lay out a systematic set of methodological rules. Chapter One defines social facts as ways of acting, thinking, and feeling that exist outside the consciousness of the individual and exert coercive power. Durkheim argues that the term "social" is used too loosely and that sociology needs a clearly delineated domain, separate from biology and psychology. Examples include legal and moral rules, religious beliefs, monetary systems, and professional practices. He also identifies "social currents," such as waves of enthusiasm or indignation in public gatherings, which sweep individuals along from outside. He extends the definition to include morphological facts such as population distribution and communication networks, arguing these are consolidated ways of acting. He formulates his definition: A social fact is any way of acting, fixed or not, capable of exerting external constraint over the individual, or which is general over the whole of a given society while having an existence of its own independent of its individual manifestations.

Chapter Two establishes the fundamental rule that social facts must be treated as things, not as ideas. Durkheim argues that every discipline in its early stages substitutes preconceptions for realities, producing ideological analysis rather than genuine science. He invokes Francis Bacon's doctrine of idola, or false notions that distort understanding, and demonstrates that sociology remains in this ideological phase. Auguste Comte took the idea of humanity's progressive evolution as a starting point rather than a hypothesis to be tested; Herbert Spencer defined society through the preconceived notion of cooperation and classified societies accordingly. From this basic rule Durkheim derives three corollaries: the sociologist must discard all preconceptions; the subject matter must be defined by externally observable characteristics rather than conformity to an ideal notion; and social facts must be studied through stable forms such as legal codes and statistics rather than through individual manifestations.

Chapter Three argues that science can objectively distinguish the normal from the pathological in social life. Durkheim rejects several common criteria, including pain and diminished survival probability, proposing instead that phenomena common to the majority of individuals within a species at a given phase of development are normal, while exceptional ones are pathological. He introduces the concept of the "average type," a hypothetical composite of the most frequently occurring characteristics within a species. He applies this framework to crime, arguing it is a normal social phenomenon because it exists in all known societies and appears "closely bound up with the conditions of all collective life" (61). Crime is not merely inevitable but "a factor in public health, an integrative element in any healthy society" (61), because eliminating it would require a uniformity of individual consciousnesses so rigid as to prevent the moral evolution societies require. Crime can also serve as a vehicle for necessary moral change: Socrates, whose independence of thought was legally criminal under Athenian law, prepared the way for a new morality his society needed.

Chapter Four addresses the classification of social types. Durkheim rejects both the nominalist position that each society is unique and the universalist position that all societies are expressions of a single human nature, proposing the concept of social species as an intermediate category. He advocates classifying societies according to their degree of organizational complexity, beginning with the simplest unit, the horde, a social aggregate containing no more elementary component. He sketches a progression from simple clusters of clans to increasingly complex polysegmentary societies, meaning societies formed by combining multiple clan-based segments into larger aggregates, and distinguishes varieties according to whether component segments retain their individuality or merge into the larger whole.

Chapter Five establishes rules for explanation. Durkheim insists that sociological explanation must be causal rather than teleological: demonstrating that a phenomenon serves a useful purpose does not explain why it exists, because needs cannot create things out of nothing. He separates the investigation of efficient causes from the investigation of functions, arguing the former should generally precede the latter. The determining cause of a social fact must be sought among antecedent social facts, not among states of individual consciousness. As the properties of a living cell cannot be found in its component molecules, social phenomena cannot be reduced to the psychology of isolated individuals. Durkheim identifies the inner social environment, particularly its volume (the number of social units) and dynamic density (the degree of moral concentration, measured by the coalescence of social segments), as the primary factors sociological explanation must address.

Chapter Six argues that the method of concomitant variations, which tracks continuous parallel changes in two phenomena to infer a causal linkage, is the sole experimental method suitable for sociology. Because social phenomena cannot be artificially manipulated, Durkheim dismisses other standard methods as impractical: the method of residues (inferring causes from what remains unexplained after known factors are accounted for), the method of agreement (identifying a common factor across cases with the same outcome), and the method of difference (comparing cases that differ in one factor and one outcome). He distinguishes three levels of comparison: within a single society, across societies of the same species, and across different species. He concludes that comparative sociology is not a special branch of sociology but sociology itself, insofar as it aspires to account for facts rather than merely describe them.

In the Conclusion, Durkheim summarizes his method's three defining characteristics: independence from philosophy and practical doctrines, objectivity grounded in treating social facts as things, and an exclusively sociological character that refuses to reduce social phenomena to psychological or biological explanations.

The volume's supplementary texts trace the development of these positions over the following two decades. In a review of Marxist theory (1897), Durkheim endorses the idea that social life must be explained by causes that escape participants' consciousness but rejects economic determinism as unproven, arguing that religion, not economics, is the most primitive social phenomenon. In a co-authored essay with Paul Fauconnet (1903), he argues that sociology must become the system of the social sciences, permeating each discipline with the principle that social facts are interdependent and subject to natural laws. In a debate with the historian Charles Seignobos (1908), he challenges the privileging of conscious motives reported by historical witnesses, insisting that every causal relationship constitutes a law. In shorter notes and letters, he proposes "social morphology" as a science of society's material substratum, clarifies that sociology is a distinct psychology with its own subject matter, and traces his intellectual debts to Comte, to his mentor Émile Boutroux's principle that every science must explain its own principles, and to William Robertson Smith's school of religious history.