comparisons, or contrasts, with other civilisations are essential for an assessment of Europe's progress. Otherwise conjectures based on a winnowing of the European historical literature are uncontrolled. They may be tested for internal consistency and fit with the evidence, but since there is no generally accepted overarching theory of very long-term economic change there would be no proper check on the chosen explanatory variables. Economic theory is too patently derived from nineteenth- and twentieth-century industrial circumstances to explain the long-run emergence of the developed world. It helps with the sub-processes, but with the very long term precisely those elements that theory takes as given, blur and shift and become variables. Where ceteris paribus ceases to apply, the comparative method offers the remaining hope for a test of significance.

If we cast around for continents and cultures to set alongside European experience and turn first to Africa, we find that the general level of development and the size and density of population lagged well behind in the historic period. Africa's fascination is sui generis. It had no major direct influence on the other continents, except maybe as a source of slaves. Certainly all was not barbarism.