Pathways to PhilosophyKindle eBooks by G Klempner

on this page

Or send us an email

Application form

Pathways programs

Letters to my students

How-to-do-it guide

Essay archive

Ask a philosopher

Pathways e-journal

Features page

Downloads page

Pathways portal

Pathways to Philosophy

Geoffrey Klempner CV
G Klempner

International Society for Philosophers
ISFP site

Home   Hughes 1   Hughes 2   Hughes 3   Hughes 4

pathways (essays)

Shirley Hughes

Can Ethical Terms Be Defined?

The answer might seem obvious. Ethical terms can be defined because they have been. "Good" means pleasure; "good" means utility; "good" means self-realisation, or self interest and so on.

Classical moral philosophy philosophers have apparently had no difficulty at all in defning terms like "good". It was just this multitude of different and incompatible definitions however, which led Moore to have some doubts about whether philosophers knew what they were doing when they attempted to define "good". Is it really possible to define "good" as one might define "triangle" or "horse"?

Are there not some important differences?

Moore is convinced that there are. In the first place, when we define "triangle" or "horse" we know what we are defining in the sense that we can see or at least formulate an empirical representation of what we are talking about. We aren't able to see goodness, or point to it, at least in the same way.

Furthermore when we define "triangle" as "an enclosed three sided plane figure", it makes no sense to ask, "but is an enclosed three sided plane figure a triangle"? — not at least if we know what we are talking about, i.e. a triangle. But if we define "good" as pleasure for instance, it does seem to make sense to ask "but is pleasure (really or always) good? Moore is convinced that it makes sense to ask this question, not merely because we may happen to be ignorant of what goodness is, and have thus made a mistake such as would be the case if we defined a triangle as a four sided figure; rather the error occurs because we have confused two quite different kinds of things with one another. We have confused a natural property (pleasure) with a non natural property (good). He calls this kind of error a "naturalistic fallacy". Since it is bound to occur whenever we attempt to identify good with something that isn't, all purported definitions of "good" commit this fallacy. "Good" he concludes is indefinable This does not mean however that the term "good" is meaningless. On the contrary it is no more meaningless than the term "yellow" which is also indefinable in the requisite sense.

Still the question remains. "What does "good" then refer to ?" Certainly not to any sensed property like yellow. It refers, according to Moore, to an intuited and unanalysiable property of goodness which some things have and others do not have.

In Principia Ethica Moore does not adopt the same approach to the term "right" as he does to the term "good". "Right" is there defined as productive of the greatest possible good.

While agreeing with Moore's position with respect to "good", Ross argues in effect that Moore himself commits the naturalistic fallacy with respect to "right". The same argument by which Moore shows that "good" is indefinable is used by Ross to show that "right" is also indefinable, and that it refers to an unanalysable property which some things are known to have, not through the senses to be sure, but by intuition.

Thus both Moore and Ross, insofar as their positions are concerned on the meanings of ethical terms are "intuitionists" or as they are sometimes called "non naturalists".

Ethical naturalism in contrast to ethical intuitionism may be defined as the view that moral terms are definable in naturalistic terms and are capable of being understood by empirical means. In contrast to both intuitionism and naturalism is another view known as emotivism. A.J. Ayer is a proponent of this position regarding the meaning of ethical terms. He basically agrees with both Ross and Moore that such terms are unanalysable notions but for quite different reasons. They are not unanalysable notions because they refer to simple properties of things but because they are totally non-descriptive of anything whatsoever. They are in other words, cognitively meaningless.

Instead he proposes the theory that words such as "good" and "bad" and "right" and "wrong" are used to express one's feelings, attitudes and emotions. it follows from Ayer's view that judgments which include ethical expressions are also non-descriptive utterances.

If I may say "theft is wrong" I an not describing anything in the sense that what I have said is either true or false. I am rather using words to express a certain feeling or attitude of disapproval, and the expression of attitudes is neither true or false. In Language Truth and Logic Ayer says we should think of moral attitudes as certain patterns of behaviour and that the expression of a moral judgment is an element in the pattern. In no way, however, does this modification of his view affect his fundamental belief that moral judgments are non-cognitive i.e. not such as to be claims to knowledge.

In contrast to this view of moral Judgments, Moore, Ross and the ethical naturalists are "cognitivists". That is to say, both intuitionists and naturalists believe that moral judgments make assertions which are either true or false and therefore are, if true, contributions to knowledge. They differ insofar as the cognitive status of moral judgments is concerned only in that the intuitionists believe that moral judgments are descriptive of non-natural properties whereas naturalists think that such judgments are descriptive of natural properties and are thus empirically verifiable.

R.M. Hare holds the view that two features distinguish moral judgments from other kinds of judgments. The features are prescriptivity and universalisability. By saying that moral judgments are prescriptive he means they primarily serve to "command" or "recommend" rather than to describe certain kinds of behaviour. By saying that moral judgments are universalisable he means that they can be applied to all relevantly similar persons in relevantly similar circumstances. (A relevant similarity is a feature of the original situation which the person making the judgment thought entitled him to make it in the first place).

Thus a moral judgment such as "don't lie" or "it is wrong. to lie" is different from an ordinary imperative statement such as "close the door" because the latter is not universalisable (i.e. we don't mean to say "let everyone close the door").

A moral judgment is different from an ordinary descriptive statement such as "the door is closed" because he latter serves no commendatory or prescriptive function. His general view differs from cognitivism (or as he calls it "descriptivism") in both its natural and non-natural forms. He believes that the cognitivists claim that the descriptive component in moral terms exhausts their meaning. is misleading for it suggests that moral Judgments are merely statements of fact. He does not believe that that is an accurate account of them. On the other hand his view differs from emotivism in that he believes that moral judgments do more than merely express our feelings and attitudes or evoke responses.

Since he holds that there are logical relations between moral prescriptions (even imperatives can contradict one another) and that they can serve as reasons for drawing moral conclusions, he believes that they are in that sense "rational".

Hare calls his own view "universal" prescriptivism. It is clear that the issue whether moral judgments are true or false cannot be settled short of an examination of many other disputes. Since a judgment cannot be said to be true or false unless it makes an assertion, it is important to consider the conflicting claims of cognitivists versus the non-cognitivists.

Even if we should find the cognitivists' arguments convincing, it will still be necessary to evaluate the differences among them. Are the intuitionists right in their contention that moral terms denote objective but non-naturalistic properties of things or are the naturalistic philosophers correct?

To establish the latter it would be necessary to show that Moore's naturalistic fallacy is either not committed by the naturalists as he claims or that it is not in fact a fallacy. If we should agree with the contention of Hare and others including Kant that moral judgments are essentially prescriptive in character then we should have to explain in what sense it can be said that a prescription such as "don't kill" or "thou shalt not kill" is cognitively meaningful.

Normally a statement in imperative form such as "close the door" is not regarded as having "truth value". Why should "don't kill" or "don't lie" be regarded any differently?

To judge that an action is generally a means to good is to judge not only that it generally does some good but that it generally does the greatest good of which the circumstances admit.

In this respect ethical judgments about the effects of action involve a difficulty and a complication far greater than that involved In the establishment of scientific laws.

The question of what things are related as causes to that which is good in itself can only be answered by the method of empirical investigation.


G.E. Moore Principia Ethica Ch. III (1903)

R.M. Hare The Language Of Morals

A.J. Ayer Language Truth And Logic 2nd Edition

A.J. Ayer Philosophical Essays (London Macmillan 1959)

W.D. Ross The Right and The Good (1930)

W.D. Ross Aristotle (1923)

W.D. Ross Foundations of Ethics (1939)

Mel Thompson Mastering Philosophy: Ethics