How/Should We Generalize?

Abstract

In Carl Kaestle's 1992 essay “Standards of Evidence,” generalization is how we know when we know. Kaestle sketches a model of increasing certainty in historical claims as they are developed and refined at increasing scales of research, from local to international. A historical claim might originate in the study of a particular place or case, but to know that the claims were true, the historian needed to move from the microlevel view to a more macro one, perhaps at the national rather than local level. Once tested and refined through comparison with other cases, possibly smoothing some of the rougher edges in the process, the claim could then be transferred beyond national borders. When a historical claim is polished enough to fit other contexts, we know it is true. Kaestle illustrates this increasing certainty through increasing scale with reference to the history of literacy and, more specifically, to scholarship on how Western European and US industrialization shaped literacy rates. Bringing studies from various locales into connection, and then comparing these cases with the national context, Kaestle summarizes that it was the commercial processes of urbanization, rather than industrialization itself, that helped produce rising literacy in the seventeenth, eighteenth, and nineteenth centuries. Generalization at greater scale becomes not only the means through which to claim the value of historical work, but the basis for constructing historical knowledge in the first place.