When Psychologists Were Naturalists: Questionnaires and Collecting Practices in Early American Psychology, 1880 - 1932

This dissertation reshapes our understanding of the earliest years of American psychology by documenting the discipline’s methodological plurality from its very inception. In tracing the use of questionnaires over the first half century of the discipline’s existence as a science, I argue that a natural historical orientation, wherein collection, analysis, and categorization are central to the scientific enterprise, has been a persistent facet of the field. Manifested in a recurrent interest in collecting information on mental life, this natural historical perspective facilitated a moral economy of data, wherein the discipline’s affect-laden norms and values sanctified the objects and practices of mass data collection. This in turn lent itself to the adoption of statistical analyses as a central component of psychological science. Although, at first glance, falling outside of the bounds of the mechanically objective practices that characterized the new psychology’s laboratory endeavours, with their use of standardized instrumentation, projects with this orientation adhered to this form of objectivity in their own way. Seeking precise accounts of mental life, including information on its physical correlates, these enterprises engaged the public in collection practices in the field. Taking up subjects with widespread interest outside of purely scientific spheres – including child study, psychical matters, and dreaming – questionnaire projects had broad appeal. Undertakings with less popular allure deliberately and necessarily confined themselves to more restricted university populations. Issues of social relevance remained mainstays of this kind of research, but by the 1920s the public’s relation to questionnaire research shifted so that they were no longer active participants in collecting activities. Instead, questionnaires were circulated in more restricted circumstances and their findings served as the basis for broad claims about the state of the public’s mind. To do so effectively, I argue, practices of collecting with questionnaires shifted from thick to thin description; no longer were rich descriptive accounts of mental life the aim of these endeavours. Rather, increasingly restricted ranges of information were accumulated, a process that culminated in the development of numerical Likert scales and the use of more sophisticated statistical analyses. Scales of this kind continue to dominate questionnaire research today.