Introduction

Historians cannot always work with problems for which all the relevant data are based on quantitative measurements. Very often the only information available for analysis relates to the number of cases falling into different categories; the category itself cannot be quantified. Thus household heads might be classified according to their sex, political affiliation, or ethnicity. Wars might be grouped into epic, major, and minor conflicts. Women might be subdivided by their religion, the forms of birth control they practised, or the socio-economic status of their fathers.

Alternatively the historian might have data that can be ranked in order, but the precise distance between the ranks either cannot be measured or is unhelpful for the problem under consideration. One example might be a ranking of all the universities in the country by a newspaper combining on some arbitrary basis a medley of criteria such as the quality of students admitted, library expenditure per student, and total grants received for research. Another might be a ranking of the power of politicians in an assembly on the basis of some measure of their influence on voting in the assembly. Similarly, an historian of religion might construct a ranking of the intensity of religious belief of the members of a community according to the frequency of their church attendance in a given period.