THE STATISTICAL TURN IN AMERICAN SOCIAL-SCIENCE - COLUMBIA-UNIVERSITY, 1890 TO 1915

Drawing on recent work in the sociology of science, we propose a sociological approach for understanding the process by which statistical methods were originally incorporated into the social sciences in America. In a departure from past accounts, which have viewed early statistical developments in the United Stares as part of the history of separate academic disciplines, we analyze interdisciplinary relations and local institutional conditions in turn-of-the-century America to elucidate the adoption and use of statistical methods by James McKeen Cattell in psychology, Franz Boas in anthropology, Franklin H. Giddings in sociology, and Henry L. Moore in economics. We argue that these four thinkers were doing boundary work to legitimize their disciplines in a competitive interdisciplinary field, where they confronted the ''newcomer's dilemma'' of conformity versus differentiation in relation to other discilplines. All four innovators turned to statistical methods to demonstrate compliance with acceptable scientific models and at the same time carve out a distinctive mode of statistical analysis to differentiate their own discipline from the others. Our analysis also shows that these developments occurred only under certain local institutional conditions. Cattell, Boas, Giddings, and Moore were faculty members at Columbia University at a time when the University had gained a competitive lead in the area of statistics over rival universities. Determined to preserve this institutional advantage, Columbia provided a conducive setting for the interdisciplinary process of the incorporation of statistical methods into the social sciences.