It is easy to see that the extension of economic research into the realm of dynamics will necessitate much closer co-operation between economic theorists and historians than has existed so far. Not only have historians to provide the empirical material with which to test our dynamic models. But, as it is impossible to conceive of dynamic processes as taking place against the quasistationary background of neoclassical economics, we have to provide, first of all, a proper setting for our models, a world in motion of which each dynamic process forms an element and from which it derives its impulse, its direction, and, hence, its significance. This world in motion provides the substance of, if it is not identical with, the historical process itself.
There are, however, obstacles to this closer co-operation which cannot be removed merely by repeated expressions of goodwill. Theorists, for instance, might profitably abstain from flaunting 'general' theories which are readily seen to rest on assumptions which are-and often enough are meant to be-mere transcriptions of the conditions of economic activity during some particular period of time, usually the theorist's own lifetime. Economic historians, on the other hand, are often confronted with a peculiar dilemma. In order to interpret the facts they have collected they need a theory, but the kind of theory best suited to their needs, viz. a theory with a time-dimension, theorists have thus far been unable to provide for them. Are we to blame them if then they evolve theories of their own, schemes of economic development 'by stages', models at the crudity and inadequacy of which the sophisticated may smile, but for which they have no substitute to