It has been one of the dominant themes of this book that educational assessment, perhaps more than any other aspect of education, has suffered the thraldom of'methodological empiricism' in which questions of technique have effectively predominated over the more fundamental issue of its effects. We have sought to account for this state of affairs in the equally unquestioned domination of the liberal reformist ideology of education and the associated consensus view of society which likewise dominated sociology until recently. Hence the need for a rational and just means of'role allocation* can be justified both pragmatically, in educational terms, as crucial to the establishment of a meritocracy and theoretically, in such functionalist sociological analysis, as necessary in a society characterized by a division of labour and social mobility. The acceptance of the assumptions contained in this ideology logically precludes the asking of certain questions.