The strange career of Millian methods in comparative social science.

For half a century, comparative social science has been closely associated with John Stuart Mill's methods of comparison. However, few social scientists had heard of Mill's methods in 1970. Within a decade, the methods of agreement and difference had become part of the methodological canon—despite Mill's objections that these methods should under no circumstances be used in the social sciences. Comparativists continued to overlook the methods that Mill actually proposed for the social sciences, which relied on an analogy with astronomical observations rather than chemistry experiments. Yet Mill's own empirical research offered substantive findings without dwelling much on methods. Over the past half-century, successful works of comparative social science have pursued all three versions of Millian methods: the comparative methods that he widely associated with; the alternative methods that he proposed for social science; and the actual methods that he pursued, whose success lay in their creativity, not in methodological recipes. [ABSTRACT FROM AUTHOR]