Standards of Evaluation
The purpose of these notes is to clarify and exemplify the standards we are using to judge moral normative theories. To begin, however, it may prove useful to answer the question, why do we need to consider such standards in the first place?
Dr. Jack Kevorkian is famous for assisting in the suicides of nearly thirty people who had terminal and painful illnesses. His actions raise questions about the morality of euthanasia. There is a long tradition rooted in the Christian religion which holds that acts of suicide are morally wrong regardless of circumstance. The desire to commit suicide is usually viewed as a symptom of poor mental health. In fact, in some states it is illegal to commit suicide. Obviously, Kevorkian disagrees.
Let us reconstruct plausible arguments on either side. Note that these arguments are not exactly the arguments given by Kevorkian or his opponents. Still, they are passably close to the mark. Suppose that Alfred has stomach cancer--an extremely painful form of cancer--and Alfred is wondering whether or not it is morally right (morally permissible) to commit suicide. Alfred first consults a Christian church leader and is given the following argument:
1. Divine Command Theory (DCT) is true.
Therefore, by 1,
2. An action x is morally wrong (morally impermissible) if God forbids x.
3. The Ten Commandments consist of what God commands and forbids.
4. "Thou shalt not kill" is one of the Ten Commandments.
5. Suicide is an act of killing.
Therefore, by 3, 4, and 5,
6. God forbids that Alfred should commit suicide.
Therefore, by 2 and 6,
7. It is morally wrong for Alfred to commit suicide.
Alfred then consults Kevorkian. Kevorkian is not a Divine Command Theorist. Rather, Kevorkian believes that pain is bad and to be avoided while pleasure is good and should be pursued. Thus Kevorkian evaluates the morality of an action not according to God's commands but according to the action's consequences: that action is morally right which promotes a greater balance of pleasure over pain than any alternative action, for the greatest number of people. We will study this theory, and theories like it, throughout the semester. For now, call the theory "Utilitarianism". Kevorkian gives Alfred the following argument:
1. Utilitarianism is true.
Therefore, by 1,
2. An action x is morally right if, and only if, x's consequences promote at least as much of a balance of pleasure over pain as any alternative action, for the greatest number of people.
3. Alfred's options are
i. To not commit suicide.
ii. To commit suicide.
4. If Alfred does not commit suicide, Alfred will live in extreme chronic pain for a year before dying. Alfred's loved ones will have to endure watching Alfred in pain and Alfred's illness will absorb considerable medical resources.
5. If Alfred commits suicide, Alfred will no longer live in extreme chronic pain, Alfred's loved ones will not have to endure watching Alfred in pain and Alfred's illness will not absorb considerable medical resources.
Therefore, by 3, 4, and 5,
6. The consequences of Alfred's suicide promote a greater balance of pleasure over pain than any alternative action, for the greatest number of people.
Therefore, by 2 and 6,
7. Alfred's suicide is morally right.
Both arguments begin by assuming the truth of an ethical theory and conclude with the moral wrongness (in the case of DCT) or the moral rightness (in the case of Utilitarianism) of Alfred's suicide. Both can easily be rewritten into valid form and both make plausible assumptions . How is Alfred to decide?
Clearly ethical theories can differ, as the two previous arguments demonstrate, in their implications. Ethicists have a special term for this. When two moral normative theories agree in all their implications we say that they are extensionally equivalent theories. DCT and Utilitarianism are thus not extensionally equivalent. So before Alfred can determine whether or not committing suicide is morally right, he must determine which of the two theories is true, ignoring for this discussion the very real possibility that both are false.
In sum, the question of whether Alfred's suicide is morally right becomes a question of which theory, DCT or Utilitarianism, is, in fact, true. Alfred's problem is not just the problem of determining the morality of committing suicide. It is the problem of determining which is the true theory. This is why we need to consider standards for the evaluation of moral normative theories. Alfred requires standards by which to evaluate DCT and Utilitarianism since DCT and Utilitarianism are not extensionally equivalent. In simpler terms, Alfred is being told that committing suicide is morally wrong by DCT and Alfred is being told that committing suicide is morally right by Utilitarianism. Which theory is true? They can't both be true; at least one must be false. So how, at least, do we decide which theory is false? This is where the Standards of Evaluation enter.
Ethicists have been concerned with the question of how we evaluate the truth of a theory for some time. To quote Richard Brandt (Morality, Utilitarianism and Rights, Cambridge 1992, pp. 112-3)*
The standard that I suggest an acceptable normative moral theory has to meet is this: The theory must contain no unintelligible concepts or internal inconsistencies; it must not be inconsistent with known facts; it must be capable of precise formulation s o that its implications for action can be determined; and -- most important -- its implications must be acceptable to thoughtful persons who have had reasonably wide experience, when taken in the light of supporting remarks that can be made, and when compared with the implications of other clearly statable normative theories. It is not required that the implications of a satisfactory theory be consonant with the uncriticized moral intuitions of intelligent and experienced people, but only with those intuitions which stand in the light of supporting remarks, etc.
Out of this we can distill three types of standard:
A. Standard of Clarity
i. An acceptable ethical theory must not contain any unintelligible concepts.
ii. An acceptable ethical theory must be capable of precise formulation so that implications for action can be determined.
B. Standard of Coherence
i. An acceptable ethical theory must not contain any contradictory principles or statements.
ii. An acceptable ethical theory must be consistent with known facts.
C. Standard of Reflective Equilibrium
An acceptable ethical theory must cohere with the moral intuitions together with the arguments of experienced and intelligent moral agents.
The Standard of Clarity merely supposes that if we either do not know what a theory says or we cannot formulate a theory with precision and in such a way that the implications for action can be determined, then we have no reason to think that the theory is acceptable. Clearly, if we do not know what a theory says then we wont be able to formulate the theory in such a way that its implications for action are clear. But it may be the case that we know what a theory says and yet, for whatever reason, we cannot formulate the theory precisely.
Some have claimed that Natural Law Theory (NLT) fails to meet the Standard of Clarity. According to NLT,
An action x is morally right if, and only if, x is consistent with all relevant proper purposes in a world of proper purpose.
The argument presented by those who think that NLT fails to meet the Standard of Clarity has two steps. The first step is to note that the concept of a proper purpose is vague. It is vague because we don't always know what to count as a purpose, or even if there are such things as purposes. Thus, in the first step, it is argued that NLT cannot be stated precisely because of the vagueness of the concept of purpose. Given the first step, the second step notes that if we do not know what is to count as a purpose, then we have no way of determining the implications of the theory. Presumably, we determine which actions are right and wrong according to NLT by considering relevant proper purposes. But if 'proper purpose' is mysterious or vague, then discovering relevant proper purposes will be impossible. Thus, in the second step, we conclude that because the concept of proper purpose is vague it is impossible to determine the implications of NLT.
The Standard of Coherence contains two related standards. First, it requires that an acceptable moral normative theory must not contain any contradictory principles or statements. As an example, consider the popular ethical theory, Cultural Ethical Relativism (CER). According to CER,
A. An action x performed in a culture C is morally right in C if, and only if, x is consistent with the practices/codes/polices of C.
B. It is wrong to judge the moral practices of another culture.
It can be argued that A contradicts B. That is, if A is true, B must be false. Or if B is true, A must be false. A and B cannot be true together. The reason, of course, is that A is a statement about morality; namely that moral judgments are relative to a particular culture. B, however, is a moral judgment that is not relative to a particular culture. B is a universal moral judgment, which is exactly what A denies can be true.
The Standard of Coherence secondly requires that an acceptable moral normative theory must be consistent with known facts. It is a known fact that stomach cancer is extremely painful. A theory which begins by denying this fact would have to be discounted at the outset. Notice that the DCT argument above does not deny this fact, it merely ignores the pain as being morally irrelevant. Pain--and pleasure--on this view should not enter into moral considerations.
NLT, however, has a problem: it is a generally accepted fact that we do not live in a universe of purposes. Indeed, modern ears, accustomed to Classical Physics, cringe to hear about rocks and rivers and wind having purposes. Thus NLT fails the Standard of Coherence because it is inconsistent with known facts.
The final standard, the Standard of Reflective Equilibrium, is the most complicated of the three standards. The idea here is that we compare a moral normative theory's implications against common, reflective, intelligent, experienced, moral intuition backed by supporting arguments. Acceptable theories must cohere with what, to the best of our ability, we are able to determine is right or wrong pre-theoretically. This is tricky, in part because the moral normative theories are supposed to determine what is right or wrong. But surely a theory which implies that the random killing of innocent people is morally right cannot be included as an acceptable theory. Further examples may help.
CER fails to meet the Standard of Reflective Equilibrium. For consider:
1. The Nazis Argument
CER implies that it is wrong to say that other cultures are morally inferior.
But according to moral intuition, some cultures are plainly morally inferior, e.g. the Nazis.
2. The Arbitrariness Argument
CER implies that the rightness or wrongness of actions is decided merely by consulting the practices of a given culture, which entails that morality is arbtrary.
But according moral intuition, morality is not arbitrary.
3. The Reformer's Dilemma
CER implies that there is no such thing as moral progress.
But according to moral intuition, it is possible to improve morally. Witness pre-civil war U.S. versus post-civil war U.S.
Thus CER is at odds with moral intuition in a number of extremely important ways. We conclude that CER fails the Standard of Reflective Equilibrium.
*I want to thank Fred Feldman for bringing Brandt's work on standards of evaluation to my attention.