It will, I suppose, be readily agreed that questions concerning the scope of moral principles are of vital importance for the moral philosopher. If we are to accept a view, such as Professor Hare's, that there are formal criteria for any morality if it is to count as a morality, then it is necessary to know what is left outside the fence. Are Egoism and Caste Morality to count as moralities, restricted to the interests of one and of a limited group of persons respectively; or is there a logical absurdity in calling them moralities at all?

Hare's thesis of ‘prescriptive universalisability’, as definitive of the formal structure of any moral system, rules out such restricted ‘moralities’. Using ‘ought’ as a key-concept in moral discourse, Hare suggests that the word has two basic functions:

(i) to prescribe: meaning to commend courses of action sincerely to oneself and others.

(ii) to prescribe consistently, or universalisably; that is, in the same way, on relevantly similar occasions, to anyone, including oneself, unless valid reasons for discrimination of persons can be offered.