Page:Philosophical Review Volume 24.djvu/441

No. 4.] that all men should agree that 2+2=4. But there is no similar agreement with regard to such a proposition as 'Polygamy is wrong.' Now why could not God have secured that all men should agree on moral matters? Locke, indeed, had suggested that God had laid down in the Gospels "so perfect a body of Ethics that reason may be excused from the enquiry." But Berkeley saw that the ethical ideas of the Gospels were accepted by a portion only, and as he feared, by a diminishing portion, of mankind. If God had intended ethics to be as demonstrable a science as mathematics, he would have arranged that the definitions of ethics should be recognized by all men to be as eternal and immutable as those of mathematics. But God has not done this, therefore it cannot be his will that there should be a demonstrable science of ethics.

In Berkeley's works subsequent to the Principles no mention is made of a possible mathematical science of ethics. The writings of his middle and later periods, in so far as they are concerned with ethics, are largely controversial. Perhaps the most systematic account of his views is to be found in the Discourse on Passive Obedience, where he makes "some enquiry into the origin, nature and obligation of moral duties in general, and the criterions by which they may be known." He takes it for granted that there are moral rules or laws of nature, which carry with them an eternal obligation. He holds that these natural principles of morality have three characteristics:— (i) Natural principles of morality are also rational. In saying that moral rules are natural laws we interpret nature in the highest sense. The best moral principles are those which may be rationally deduced by the maturest reason. These natural-rational principles "grow from the most excellent and peculiar part of human nature." They are laws of nature, but they are also eternal rules of reason, because they necessarily result from the nature of things and may be demonstrated by the infallible deductions of reason.