I Do not much hold with ardent debate about method. A historian method is as individual—as private—a matter as a novelist's style. There is good reason for reserve—even reticence—on this subject, except insofar as we seek to share each other's unique professional adventures and to listen occasionally, in a mood of interest tempered with skepticism, to such general reflections as we each would draw from those adventures.

This paper is an attempt to examine critically the function of theory in historical research and particularly in economic history. the function of We shall take as our starting point the assertion that the historian is not interested simply in collecting facts or true statements about some segment of previous experience. He wants to find causes and to explain what happened. The purpose of this paper is to introduce some of the problems attached to the concepts of historical causality and explanation in a stochastic universe and to suggest how the analytic tools of scientific inference can be applied in economic historiography.

Our conference today on comparative economic history is in some clanger of rushing into the wide-open spaces of ambiguity, for the term is new, and to agree too quickly on its meaning and implications may not even be desirable. In order to avoid engaging in a mere game of definitions, this paper will deal first with three general types of comparison in relation to their bearing on problems of evidence. It will then review some of the chief uses to which these types of comparison have been put in building up our body of knowledge about Western economic history. The survey will close with particular reference to our own preindustrial stages of economic growth, when western Europe was, in our uncomplimentary phrase, an underdeveloped or backward area.

In this paper I shall limit myself to various comments on integration narrowly conceived, that is, within the discipline of economic history itself, and to a few historical cases used to illustrate the possibilities and limitations of comparative method as a means to this end. Comparative study, as Miss Thrupp has stressed, scarcely represents a new departure in economic thought. The novelty seems to be rather in the almost universal appeal it now has for those who would explore the mysteries of change. This interest appears to be a consequence of the impact of recent developments in backward-area studies and the closely related pursuit of a general theory of growth; these tend to focus attention on different kinds and rates of growth and comparison becomes inevitable. In turn, accumulations of knowledge about economic change in very different contexts lead to the need for some means of correlating diese, partly for aesthetic reasons but also because a synoptic view of the landscape of change has its very practical use. Unless the whole field can be surveyed from a single point of view, and by this I do not mean the same point of view, it is very difficult to know where the need for more intensive exploration is greatest.