Why study international relations? Is it simply because, like Everest, they are there? Or is it in order more effectively to empathize with international statesmen? Or is it to seek such a measure of understanding of structures and processes that interventions to change them in desirable ways may have increasing probabilities of success? Garnett most neatly attaches himself to the first of these positions. The contributors to the scholarly Bull and Watson volume (or more exactly, most of the contributors) seem to hold the second view. Kubalkova and Cruickshank, Strange, and Banks have the third objective in mind, although in very different ways, and with varying degrees of explicitness. The choice of objective made by a scholar is determined not merely by his interests and the intellectual and analytical skills of which he has made himself master, but also by his epistemological assumptions. Thus Garnett and Bull would assert that, in contrast with the ‘exact’ sciences, it is erroneous to suppose that scientific methods can be employed in the study of social questions, and that any attempt so to do and thereby to arrive at advice about action must be mistaken—and dangerous to the extent that it creates expectations about results that are as likely to be falsified as to be confirmed.