Sociology's "One Law"

Abstract

Though accepted widely enough for nomination as sociology's one law, there has never been an adequate empirical test of Durkheim's proposition that Catholics have lower suicide rates than Protestants. As a result, the empirical foundation for one important application of Durkheim's assertion that egoism and suicide vary proportionately remains uncertain. We use crossnational, longitudinal data from twelve nations to test Durkheim's proposition. Our comparisions (1) use national, female, and male suicide rates; (2) are made both with and without control for nation's level of development; and (3) are for four different time periods between 1919 and 1972. Results largely refute Durkheim's proposition. However, application of Durkheim's entire theory of integration (egoism and altruism) rather than just his theory of egoism shows that the data may nonetheless be consistent with his theory of variation in suicide rates. After noting sociology's lack of scientific laws, Merton identifies Durkheim's generalization that Catholics have a lower suicide rate than Protestants as an approximation to such a law. More recently, LaCapra calls Durkheim's theory of suicide perhaps the only significant lawlike statement in sociology (147; see also Johnson). Whether in books on theory construction and logic of inquiry (e.g., Hage; Stinchcombe; Wallace; Winton), or methods (e.g., Cole; Riley; Selltiz et al.; Smith), or statistics (e.g., Loether and McTavish), or in introductory texts (e.g., Broom and Selznick; Demerath and Marwell; Goodman and Marx; Levin and Spates; Light and Keller; Wilson), sociologists continue to discuss Durkheim's generalization relating to differences in Catholic/Protestant suicide rates as if it were a scientific law. Presumably because of its widespread acceptance Merton