Lost in Translation: An Epistemological Exploration of the Relation between Historical Analysis and the NOMINATE Algorithm

The NOMINATE algorithm has become the most important analytical tool used in the study of the United States Congress. As such, congressional scholars have developed a great many social conventions, practices, and assumptions that enable interpretation of the statistical artifacts the algorithm produces. However, as many of these scholars recognize, serious problems emerge whenever we try to translate these statistical artifacts into language and thus attempt to assign them meaning in historical analysis. These problems are irresolvable because they reside in the very construction of the algorithm itself.