This article provides a history of the treatment of observational errors where conditions cannot be controlled to reduce inaccuracies, more specific, a history of the discussion of errors in social statistics. The main focus is on Oskar Morgenstern's atypical position in this discussion. In...