His influential work Principia Ethica is one of the main inspirations of the movement against ethical naturalism (see ethical non-naturalism) and is partly responsible for the twentieth-century concern with meta-ethics.
The naturalistic fallacy
Moore asserted that philosophical arguments can suffer from a confusion between the use of a term in a particular argument and the definition of that term (in all arguments). He named this confusion the naturalistic fallacy. For example, an ethical argument may claim that if a thing has certain properties, then that thing is 'good.' A hedonist may argue that 'pleasant' things are 'good' things. Other theorists may argue that 'complex' things are 'good' things. Moore contends that even if such arguments are correct, they do not provide definitions for the term 'good.' The property of 'goodness' cannot be defined. It can only be shown and grasped. Any attempt to define it (X is good if it has property Y) will simply shift the problem (Why is Y-ness good in the first place?).
Moore's argument for the indefinability of "good" (and thus for the fallaciousness of the "naturalistic fallacy") is often called the open-question argument; it is presented in §13 of Principia Ethica. The argument hinges on the nature of statements such as "Anything that is pleasant is also good" and the possibility of asking questions such as "Is it good that x is pleasant?" According to Moore, these questions are open and these statements are significant; and they will remain so no matter what is substituted for "pleasure". Moore concludes from this that any analysis of value is bound to fail. In other words, if value could be analysed, then such questions and statements would be trivial and obvious. Since they are anything but trivial and obvious, value must be indefinable.
Critics of Moore's arguments sometimes claim that he is appealing to general puzzles concerning analysis (cf. the paradox of analysis), rather than revealing anything special about value. The argument clearly depends on the assumption that if "good" were definable, it would be an analytic truth about "good," an assumption many contemporary moral realists like Richard Boyd and Peter Railton reject. Other responses appeal to the Fregean distinction between sense and reference, allowing that value concepts are special and sui generis, but insisting that value properties are nothing but natural properties (this strategy is similar to that taken by non-reductive materialists in philosophy of mind).
Good as indefinable
Moore contended that goodness cannot be analysed in terms of any other property. In Principia Ethica, he writes:
- It may be true that all things which are good are also something else, just as it is true that all things which are yellow produce a certain kind of vibration in the light. And it is a fact, that Ethics aims at discovering what are those other properties belonging to all things which are good. But far too many philosophers have thought that when they named those other properties they were actually defining good; that these properties, in fact, were simply not "other," but absolutely and entirely the same with goodness. (§ 10 ¶ 3)
Therefore, we cannot define "good" by explaining it in other words. We can only point to an action or a thing and say "That is good." Similarly, we cannot describe to a blind person exactly what yellow is. We can only show a sighted person a piece of yellow paper or a yellow scrap of cloth and say "That is yellow."
Good as a non-natural property
In addition to categorising "good" as indefinable, Moore also emphasized that it is a non-natural property. This means that it cannot be empirically or scientifically tested or verified - it is not within the bounds of "natural science".
Moore argued that once arguments based on the naturalistic fallacy had been discarded, questions of intrinsic goodness could only be settled by appeal to what he (following Sidgwick) called "moral intuitions:" self-evident propositions which recommend themselves to moral reflection, but which are not susceptible to either direct proof or disproof (PE § 45). As a result of his view, he has often been described by later writers as an advocate of ethical intuitionism. Moore, however, wished to distinguish his view from the views usually described as "Intuitionist" when Principia Ethica was written:
In order to express the fact that ethical propositions of my first class [propositions about what is good as an end in itself] are incapable of proof or disproof, I have sometimes followed Sidgwick's usage in calling them 'Intuitions.' But I beg that it may be noticed that I am not an 'Intuitionist,’ in the ordinary sense of the term. Sidgwick himself seems never to have been clearly aware of the immense importance of the difference which distinguishes his Intuitionism from the common doctrine, which has generally been called by that name. The Intuitionist proper is distinguished by maintaining that propositions of my second class—propositions which assert that a certain action is right or a duty—are incapable of proof or disproof by any enquiry into the results of such actions. I, on the contrary, am no less anxious to maintain that propositions of this kind are not 'Intuitions,’ than to maintain that propositions of my first class are Intuitions.
— G.E. Moore, Principia Ethica, Preface ¶ 5
Moore distinguished his view from the view of deontological intuitionists, who held that "intuitions" could determine questions about what actions are right or required by duty. Moore, as a consequentialist, argued that "duties" and moral rules could be determined by investigating the effects of particular actions or kinds of actions (PE § 89), and so were matters for empirical investigation rather than direct objects of intuition (PE § 90). On Moore's view, "intuitions" revealed not the rightness or wrongness of specific actions, but only what things were good in themselves, as ends to be pursued.