In this paper I assume that if morality and prudence do not conflict, the agent has most reason to do either moral act or prudential act or both. However, I want to argue that if they do conflict, there is no objectively commensurate basis for comparing the powers of moral and prudential reasons or the values of their ends to determine what the agent has most reason to do. If we still want to find some guidelines for our action, what we can do at most is to appeal to reflective public opinion or to vote. In that case, sometimes we can determine what the agent has most reason to do but sometimes we cannot, all depending on reflective public opinion.
I
What does the agent have most reason to do when moral and prudential reasons conflict? Bruce Russell provides the following example to show that sometimes it is difficult to answer this question.
Suppose your son has robbed a rich man of some of his jewels, the police are after him, and he asks you to help him escape to Brazil. You know you can arrange things so that neither of you will get caught. You also know that if he is caught he will be sent to prison and his life will be ruined. It would be wrong of you to help him escape, but why isn’t it true that what you have most reason to do is to help your son escape justice? [1]
A moral person would ask: “Why isn’t it true that what you have most reason to do is to send your son to the police, not to help him escape justice?” There are two fundamental principles by which we can determine what the agent has most reason to do: the prudential and the moral. The prudential principle can be expressed as follows:
P: If doing A most promotes the agent’s self-interest or intrinsic reflective preference, the agent has most reason to do A.
The moral principle can be expressed as follows:
M: If doing A most promotes the justice of society or the good for society or the performance of a duty for the good of a non-agent person, the agent has most reason to do A.
According to P, the father has most reason to help his son escape justice since helping his son escape would most promote his intrinsic reflective preference. According to M, the father has most reason not to help his son escape justice since not helping his son escape most promotes the justice or the good of society. Let us suppose that there is no disagreement on the truth of the other premises except P and M. Then, the evaluation of the strength of two arguments eventually concentrates on P and M. Generally, whenever a moral reason and a prudential reason conflict, we need to examine two arguments for each of reasons to determine the better reason. If there is no disagreement on the other premises and the arguments are valid, such examination will eventually lead to comparing P and M. We should notice that P and M express the ultimate ends of action. P asserts that something good or valuable for the agent is the ultimate end of action, while M asserts that something good for the society, perhaps including the good for the agent (though not necessarily), is the ultimate end of action. Thus, when the two principles conflict in a given situation, what the agent has most reason to do depends on which principle is more reasonable or which end is more valuable. Can we compare the two principles (or the two ultimate ends) to determine the superior reason or value?
II
Some philosophers argue that it is impossible to do so. Their argument can be stated as follows:
(1) The superior reason or value between P and M can be determined if and only if P and M are commensurable.
(2) P and M are not commensurable.
(3) Therefore, the superior reason or value between them cannot be determined.
To deny (3), one needs to deny either (1) or (2). Consider the denial of (2) first. To show (2) is false we need to understand what “commensurable” means. There are two versions, corresponding to two types of commensurate bases. One version of commensurability can be defined as follows:
C1: A and B are commensurable if and only if they can be precisely measured by a single cardinal `scale’ of units of value.
Suppose I am looking for a job with good salary. I get two offers from two colleges: one college pays a salary of $30,000 and the other a salary of $35,000. The two offers can be measured by a single cardinal scale of units of value, i.e., units of money, by which I can determine the better offer. Thus, the two offers are commensurable.
However, it is very dubious whether there is a single cardinal scale of value for P and M or the prudential and moral ends. For instance, it is very dubious whether there is a cardinal scale for determining the value of maximizing the agent’s happiness and the value of maximizing the good of society. For, as Robert Nozick argued, a person is separate and his is the only life he has. Thus, no one is entitled to force sacrifice upon him in the name of the happiness of society or other people. [2]
However, many philosophers and economists claim that cardinal scales of value are not required for value or reason comparison. For instance, it seems that there is no cardinal scale for measuring the value of keeping a promise and the value of saving an ordinary person’s life, but it seems obvious that the latter is more valuable than the former.
When lacking a cardinal scale of value, how can we determine the superior value or reason between two compared items? We can determine the superior value or reason according to qualities of values rather than the quantities of values, i.e., some ordinal scales or qualitative rankings. Thus, the second version of commensurability can be defined as follows:
C2: A and B are commensurable if and only if there is an ordinal scale or value ranking in terms of which the superiority of A and B can be determined.
Keeping a promise and saving an ordinary person’s life are commensurable because there seems to be an ordinal scale that suggests a person’s life is more valuable than keeping one’s word. Using this scale, one can proceed with the comparison, though there is no single cardinal scale of value for those choices. We should note that moral values and prudential values could be internalized in one person (i.e., this person values both morality and prudence). In that case, there could be a personal ordinal commensurate basis for those values, say, morality takes priority over prudence, but such a commensurate basis is only subjective and arbitrary. There always could be another subjective value ranking, say, prudence takes priority over morality. Someone may argue that what the agent has most reason to do depends on the agent’s subjective value ranking system, but this answer is not satisfactory. First, we want to know objectively what the agent has most reason to do when morality and prudence conflict, regardless of the agent’s desires, personal goals and beliefs. Second, if there are two competing rankings, we want to know which ranking is better or more rational. Then we come back to the original question: which principle provides the agent most reason to act, moral or prudential? It seems that there is no further ordinal scale for evaluating P and M since P and M are already ultimate. If we do not want to accept (3), i.e., the conclusion of the argument I presented above, we need to consider the denial of (1), the first premise. We need to show that even though there is no commensurate basis (in the sense I have defined) for P and M, we still can determine the superiority of P or M.
III
In “Two Forms of Ethical Skepticism,” Bruce Russell gives some interesting cases of the superiority of moral or prudential reasons. One of his cases is that a rational, callous person takes candy from a baby instead of spending a couple of Canadian quarters on a candy bar. Russell believes that this case clearly shows that moral reason overrides prudential reason, i.e., the force of moral reason in this case is greater than that of prudential reason. [3]
Russell also gives some other cases to show that prudential reason sometimes, if not overriding moral reason, is at least not overridden by moral reason. One of the cases is this. Suppose a young woman has a lifelong dream to become a physician. However, even after much hard study she has been unable to score high enough on the MCATs to be admitted to any medical school. Suppose she has a good opportunity to cheat and that will ensure her an MCAT score high enough to get admission to some medical school. Her patients will not be harmed since she will become a good physician after the necessary training at medical school. The only person who will be harmed is the one denied admission to medical school because of her occupation of one of the available positions, which leaves one less slot to be filled. However, that person does not really care about his denied admission. And suppose, for some reason, that young woman knows all those facts and consequences. Now the question is: what does she have most reason to do? Russell seems to believe that in this case prudential reason overrides moral reason, or, at least, it is difficult to show moral reason overrides prudential reason. [4]
If we accept Russell’s cases of the superiority of moral or prudential reasons and if we assume that there is no commensurate basis for moral and prudential reasons, then (1), the first premise of the argument, must be rejected. The question is: what is the basis for such acceptance?
Philippa Foot argues that we should not think of prudential and moral theories of practical rationality as rival theories but rather as different parts of practical rationality, none of which should be mistaken for the whole. She proposes that it is not always rational to act in one’s self-interest or to act for a moral reason. If it is said that moral considerations are always overriding, it must be the overall ‘moral judgment.’ [5] But how can we compare or use those aspects of practical rationality to find out what we have most reason to do? Foot claims,
It is, I think, possible to see, even if not as yet very clearly, the common thread linking these different parts of practical rationality. The root notion is that of the goodness of human beings in respect of their actions; which means, to repeat, goodness of the will rather than of such things as sight or dexterity, concentration or memory. [6]
Foot argues that the possession of moral character (keeping a promise, telling the truth, helping others, etc.) or doing moral things without the application of physical force or just by conventional arrangements (i.e., the goodness of the will) is necessary for our survival, just as “it is necessary for plants to have water, for birds to build nests, for wolves to hunt in packs, and for lionesses to teach their cubs to kill.” [7]
The goodness of the will seems to be in favor of moral reasons. Yet, if this is the common thread linking the different grounds of practical rationality, why can we not show that moral reasons always override prudential reasons? Also, how can the goodness of human beings or the goodness of the will help determine what the agent has most reason to do in such situations as the Prisoner’s Dilemmas and the young woman aspiring to be a physician, where moral reason and prudential reason sharply conflict?
Some philosophers are looking for the third value or principle by which we can compare the rational forces of moral and prudential reasons and determine what the agent has most reason to do. The third value or principle is neither a cardinal scale nor an ordinal scale. It is supposed to provide guidelines for value or reason comparison between moral and prudential principles. It could be one value or principle but also could be a set of values or principles.
Ruth Chang believes that any value comparison between two items must proceed in terms of a value that covers the values of the compared items. She calls this value “the covering value of the comparison.” Then, what is the covering value of the values of moral and prudential ends? Chang believes that phrases like “What there is most reason to do, all things considered” or “betterness, all things considered,” indicate that there is such a covering value. She seems to take moral reasons and prudential reasons as different aspects or components of the covering value. Just as scientists use different principles, such as simplicity, precision, and prediction, to give an overall evaluation of a particular hypothesis, we can also use different principles, such as moral and prudential (none of them is privileged), to give an overall evaluation of what we have most reason to do. However, there are some dissimilarities between science and practical reason. First, there seems to be one, dominant principle among different principles of evaluating a hypothesis: prediction power or the range of the true observational statements implied by the hypothesis. When other principles, such as simplicity and precision, conflict, we can appeal to this dominant principle to settle the case. But it does not seem that there is such a dominant principle we can appeal to in practical reason, especially when morality and prudence conflict.
Furthermore, the meaning of “rational” or “most reason to do, all things considered” is used in either the prudential sense or the moral sense. Either sense of “rational” or “most reason to do, all things considered” is irreducible. If the superiority of reason in a given conflicting situation can be determined, any reason or value comparison would yield a result about what the agent has most reason to do in either the prudential sense or the moral sense but not both. Thus, there is no such sense (or value) that is independent of the prudential sense or the moral sense and there is no such an overall sense that can “cover” both the prudential and the moral senses. For instance, in the case of the callous person, if we believe that he has most reason not to take the candy bar from the baby, that reason must be moral. It cannot be prudential or the third sense of reason.
Ruth Chang seems to recognize such a difficulty. She says that there is no content of such a value except what is given to it. It functions more like a place-holder for any value, say, x, y, or z. If different conflicting values (principles) enter the place-holder, we need to find a way to compare them and determine the superiority of value. Chang introduces a “nominal-notable test” as a way to determine the superiority of different values and principles.
Chang seems to believe that the covering value for comparing moral and prudential acts consists of moral and prudential values. We can rank the acts of each sort according to their positions in the hierarchy of the values from the most nominal to the most notable. She claims that, “In general, a notable moral act is better with respect to both morality and prudence than a nominal prudential one.” [8] Consider again the case of taking candy from the baby. According to one aspect of the covering value, i.e., prudential value, the callous person has reason to take the candy bar from the baby but according to another aspect of the covering value, i.e., moral value, he should not. Then, what does he have most reason to do? It seems that, according to Chang, the callous person has most reason not to take candy from the baby. The reason is that he only has a nominal prudential reason to take candy from the baby but he has a notable moral reason not to do so. In the case of the young woman who wants to enter the medical school by cheating, what does she have most reason to do according to Chang’s theory? It seems that she has a notable prudential reason to cheat but a nominal moral reason not to do so. Chang does not consider a case like this, but from what she says, we seem to be able to draw a parallel principle in favor of prudence: “In general, a notable prudential act is better with respect to both morality and prudence than a nominal moral one.” If neither morality nor prudence is privileged and if we assume the acceptance of that parallel principle, we can conclude that what the young woman has most reason to do is to cheat. However, this conclusion is controversial. Genuine moralists may argue that the moral reason not to cheat is notable, since the rule of prohibiting plagiarism would be shaken if we allows such an exception and this is not in the interests of the public. If we accept the two nominal-notable principles I just mentioned, the debate about what the agent has most reason to do will concentrate on how to draw a line between nominalness and notableness.
IV
How and on what basis can we draw such a line? In most clear cases of the superiority of reason, such as the case of taking candy bar from a baby, we actually appeal to our normal sense or intuition. What is our normal sense or intuition? It can be defined as an ability or power to recognize some self-evident truths or statements. A truth is self-evident just in case if we understand it, we will believe in it and will be justified in believing it. There seems no doubt that there are self-evident truths in logic and mathematics, but are there any self-evident truths in morality? It seems that there are some self-evident principles or standards in moral area. For instance, “Do not kill an innocent person,” “Equals should be treated equally” and “Do not do evil things to other innocent people” seem to be self-evident. There could be questions of whether those statements are really self-evident or whether our normal sense or intuition about those self-evident statements is universal. For simplicity, I only consider whether there is a universal sense or intuition about the overridingness of a reason in the cases of the betterness of reason we have considered above. It might be self-evident for some that the reason for not cheating does not override the reason for cheating, but it might not be self-evident for everyone. Similarly, it might be self-evident for some that the reason for not taking the candy bar from the baby overrides the reason for taking the candy bar, but it might not be self-evident for everyone. Your intuitions might be different from mine.
If there is no universal normal sense or intuition, then on what basis we accept the cases of the betterness of reason or draw a line between nominalness and notableness? I want to argue that the basis is reflective public opinion. It is justified by the intuitions of the majority of rational people in a community or society who really understand the relevant situation. What should we do? What is notable or significant in our life and what is not? We actually have to vote or to appeal to the opinion of the majority of rational people in the community or society (sometimes the only rational people who really understand the situation may be professional philosophers). In the case of taking the candy bar from the baby, the rational but callous person and some other people of the same kind may not believe that the callous person has overriding reason not to take the candy bar from the baby but we hold (or reflective public opinion holds) that he has overriding reason not to do so, regardless of his preference or desire. In the case of the father and the son, why is it difficult for us to show that the father has most reason not to help his son escape? This is either because even by appeal to reflective public opinion it is not clear whether we can draw a line between nominalness and notableness about moral and prudential reasons or because by appeal to reflective public opinion we may say that the father has a notable moral reason to refuse to help his son escape but also has a similarly notable prudential reason to help him. Thus we cannot determine what he has most reason to do. Therefore, it is not always possible to determine what the agent has most reason to do when morality and prudence conflict.
In short, as I have argued, there is no objective way to prove what the agent has most reason to do when morality and prudence conflict. If we still want to find some guidelines for our action, what we can do at most is to find out true statements about reflective public opinion, take them as given premises, and then work on what the agent has most reason to do in those conflicting cases. The task cannot always be successful because there could be the cases like that of the father and the son where we cannot settle the case by appeal to reflective public opinion. In that case, we can only determine what the agent has reason but not most reason to do.
Notes
[1] Bruce Russell, “Two Forms of Ethical Skepticism” in Louis P. Pojman ed. Ethical Theory (Belmont: Wadsworth, 1989), 469.
[2] See, his Anarchy, State, and Utopia (New York: Basic Book, 1974), 32-3.
[3] See, Bruce Russell, “Two Forms of Ethical Skepticism” in Louis P. Pojman ed. Ethical Theory (Belmont: Wadsworth, 1989), 467, 469.
[4] See, ibid., 469.
[5] See, Philippa Foot, “Does Moral Subjectivism Rest on a Mistake?” in Oxford Journal of Legal Studies, Vol. 15 (Spring 1995), 5.
[6] Ibid., 7.
[7] Ibid., 8.
[8] Ruth Chang, “Introduction” to Ruth Chang ed. Incommensurability, Incomparability, and Practical Reason (Harvard University Press, 1997), 32.
（Zhen Chen, Department of Philosophy, Wayne State University）
（原载 Southwest Philosophy Review, Volume 16, Number 1, 2000。录入编辑：红珊瑚）
