ABSTRACT
In his written comments on this paper, John Tukey strongly objects to our interpretation of the role of subjective, nonempirical conventions and beliefs in deciding among alternative causal interpretations of social science data. He writes “ to say that causal interpretations are resolved by ‘context, beliefs, and a priori theory’ is to demote causal interpretations to the level of transient opinions or fads.” He also states that “ the authors fear—or are strongly repelled from—the idea of confirmation as the final standard of correctness.” He further writes that “some of my deepest discomfort stems from the feeling that the authors equate ‘assumption’ to ‘truth.’ If we assume it, it is so—if we don’t, it isn’t!”
