There is…one rule that lies at the heart of every religion – that we do unto others as we would have them do unto us. This truth transcends nations and peoples – a belief that isn’t new; that isn’t black or white or brown; that isn’t Christian, or Muslim or Jew. It’s a belief that pulsed in the cradle of civilization, and that still beats in the heart of billions. It’s a faith in other people, and it’s what brought me here today.
--President Barack Obama, speaking in Cairo Egypt, June 4, 2009
The Golden Rule (GR): Do unto others as you would have them do unto you, is widely accepted as a valid moral rule. However, it is fairly easy to show that it yields incorrect and inadequate moral guidance in many cases. But there are reasons why it has been so widely accepted and continues to be taught, and these reasons have to do with its very incompleteness as a principle of ethics, and its connection with the idea of reciprocity.
The GR can be stated in either a positive or a negative manner:
GR+: Do unto others as you would have them to do unto you.
GR-: Do not do unto others as you would not have done unto you.
In the first version it states a principle of positive duty while in the second it states one of negative duty. However, it is possible to state the GR in a way that is neutral between acts and omissions:
GR0: Treat others as you would have others treat you (where "treatment" can be understood as involving both acts and omissions).
By employing the subjunctive "would" the GR is quite different in its implications than the principle:
GR%: Treat others as they have treated you.
The GR% is a principle of pure reciprocity. Using it prescribes returning kindness for kindness and cruelty for cruelty, presumably in the same proportions as they are received from others. One might regard the GR% as "fair" in that it would rule out repaying a kindness with cruelty, and also, repaying cruelty with kindness. But unless you have already been acted upon by others in some way, there is no way to interpret the GR%. The change to talking about how one would want to be treated, obviates this problem, but as we shall see, creates other problems.
Also by way of preliminary clarification, one should understand the GR as having implicit universal quantifiers:
GRQ: Everyone should always treat all others as they themselves would wish to be treated.
Making the quantifiers explicit allows us to see more clearly why GR provides incorrect moral guidance in many cases.
Consider the rich man who does not wish to be given money as charity in order to feed his children (because he doesn't need it) who reasons that since he does not want to be treated as a recipient of charity, others, who might be much poorer than he, should not either. Or consider a case from my own experience. I would like my wife to attend my lectures because I would want to get her honest opinions on how I did. She, on the other hand, has repeatedly told me that she does not want me to attend her lectures since she thinks I would distract her and make her more nervous. If I followed the GR I would attend her lectures and if she followed the GR she would not attend my lectures. But this is clearly wrong: the right thing to do is for me not to attend her lectures and for her to attend my lectures.
In order to be successfully applied the GR assumes that there is a symmetry in the conceptions of the good between the agent and the patient, that others would want to be treated as oneself wants to be treated, but this is obviously untrue in many cases.
But, it will be objected, the cases I have chosen to illustrate a problem with the GR are framed at the wrong level of abstraction. Rather than framing the application in terms of specific types of actions (giving charity or attending lectures) the GR should be understood as recommending that one treat other persons in accordance with their own conception of their good. Since I would want others to respect my own conception of my good I should assume that others would likewise want that I respect their conceptions of their own good.
So then we might restate the GR as: "Everyone ought always treat all others in accordance with their own conceptions of their own good" or more succinctly:
GR*: Treat others as they want to be treated.
GR* is an improvement over GR because rather than assuming that the agent's own preferences are automatically a good guide to the patient's preferences, it asks us to consider the patient's preferences directly. But one immediately sees there is another problem with this formulation. What if the moral patient involved has a mistaken view of what their own good consists in?
Suppose Janine has a fifteen year old son, call him Oscar, who thinks it is in his best interest to drop out of high school and get a job as a gas station attendant. Janine, as Oscar's parent, would be acting irresponsibly to simply agree to this idea. Oscar has a mistaken view of what his good consists in and his own wishes in this case ought to be set aside in favor of Janine's more adequate conception of her son's good.
The problem with GR* is that in some cases one ought to treat people in ways that those people should ideally wish to be treated rather than in the ways they actually do wish to be treated. So we need to revise the GR once again:
GR*I: Treat others as they should ideally want to be treated.
But now we can clearly see that in order to apply GR*I one needs to have a theory of the good which tells one what it is persons ideally should want for themselves. Since the GR purports to be a moral rule or principle, let's assume that the term "should want to be treated" can be understood as meaning how he or she "morally ought to be treated." On this interpretation, the GR tells us something uncontroversial but also rather unhelpful, namely that we ought to treat people the way they ought to be treated.
To avoid this tautology, those who employ the GR as a guide to moral action must have a substantive ethical theory which describes the kinds of treatment which persons are morally entitled to receive. The GR*I will thus yield rather different prescriptions depending on which moral theory the agent has. If the agent is an ethical egoist he would want to be treated in ways that maximize his self interest. As Brian Medlin points out, it is doubtful that this principle can be consistently universalized. If the agent is an act utilitarian he wound want others to treat him so as to maximize the utility of all concerned on each occasion in which he is acted upon. On this view, the agent must be prepared to accept the fact that his or her own interests may be sacrificed at the altar of utility. If the agent is a contractarian he will want that others keep their agreements with him in cases in which he keeps his agreements with them. In this case, the relationship between the GR and the principle of reciprocity comes into focus. To the extent that the rules of morality can be viewed as conventions that are chosen and followed because they are in the long run mutually beneficial, the GR functions as a way of reminding us to be faithful to these agreements and the practices they support. If you would want others to keep the promises they make to you, then you are required to keep the promises you make to others. If the agent is a Kantian he will want others to treat him as an end in himself. If he is a virtue theorist he will want others to treat him so as to promote his eudaimonia, and so forth.
Depending on the moral theory of the agents who apply the GR*I, it will yield rather different moral prescriptions, and this feature of the GR may, as Paul Weiss has observed, be the basis of its nearly universal appeal, "because it allows and even tempts as many interpretations as there are modes of self-regard and systems of ethics" (Weiss 1941, 421).
But the GR can also support quite wicked actions if only the agent is willing to be treated in a similar fashion. If my ethical theory prescribes wanton murder and mayhem, and I am willing to be the recipient of such actions myself, then the GR does not rule it out. According to Weiss's analysis, for the GR to function as a moral rule three conditions must also be met:
(a) We know what we want.
(b) What we want is identical with what we ought to desire.
(c) What is good for us is also good for the rest. (Weiss 422).
The key conditions are the second and third – that we know how ideally we should want to be treated, and this idea of the good is valid for others. But how can we prove (a) and (b)? Kant provides at least the beginning of an answer. In the Groundwork of the Metaphysics of Morals (1785) he observes:
Let it not be thought that the common "quod tibit non vis fiery, etc." could serve here as the rule or principle. For it is only a deduction from the former, though with several limitations; it cannot be a universal law, for it does not contain the principle of duties to oneself, nor of duties of benevolence to others (for many a one would gladly consent that others should not benefit him, provided only that he might be excused from showing benevolence to them), not finally that of duties of strict obligation to one another, for on this principle the criminal might argue against the judge who punishes him, and so on. (note 7)
In the Second Formulation of the Categorical Imperative he proposes a different principle, that persons (both ourselves and others) be treated as ends in themselves rather than as means only. We can then reinterpret the GR as follows:
CI2: Everyone ought always treat all other persons as ends in themselves rather than as means only.
CI2 says that one should always regard other persons as rational agents in their own right who have a good in themselves and for themselves. In the case of competent adults, this is usually understood as entailing that one respect the other person's autonomous wishes concerning themselves. So, in the case of competent adult moral agents, CI2 is equivalent to GR* -- we should treat other adults in the ways in which they autonomously wish to be treated. This reading of the GR follows on the assumption that competent adult moral agents are the best judges of what is in their own good consists in. In cases involving minors or adults with impaired judgment or mental capacity, the alternate rule would be to do what is in their best interest, which is equivalent to GR*I in that it allows a substituted judgment as to what a person's good consists in. It does not, however, automatically assume that the agent's conception of his or her own good is a reasonable approximation to the patient's good. If the patient is, say, suffering from advanced dementia, then if I am their caregiver I ought to take care of various tasks for them, such as paying their bills, but I would not consider it in my best interest to have them take care of paying my bills for me. Kant's Categorical Imperative can thus be understood as an improved version of the GR. It is an improvement because rather than relying on a subjective and perhaps idiosyncratic conception of the good which an agent may have and which the agent may or may not share with the patient, it proposes a general conception of the good for rational moral agents that could in principle be embraced by all rational moral agents.
Kant claims that the second formulation of the CI is equivalent to the first formulation: Act only on that maxim that one can at the same time will to be a universal law. So then can the GR pass the universalization test? Not, it seems in the standard version, since I would not want other people to treat me in accordance with their conception of their good in cases in which their conception was mistaken or in cases in which it disagreed with my conception of my good. Willing the standard version of the GR would result in a contradiction of my will since in cases in which my conception of my good implied A and other agent's conception of his good implied ~A, I would be willing both A & ~A by willing that both of us follow the GR. I can, however, will the GR*I without contradiction since in this case if I am a competent and autonomous moral agent I would be willing that others treat me in accordance with what I conceive to be in my own best interests, and if I am not autonomous, then I would be willing that they treat me in accordance with what I should regard as in my best interests. In both of these cases, other agents would be directed to treat me as an end in myself whose best interests are morally considerable and should be taken into account in their decisions about how to treat me.
But Kant's theory provides only a "thin" conception of the good. A utilitarian theory attempts to put some additional flesh on the bare bones of respect for persons. Utilitarians often argue that in deciding how one ought to act one ought to take into consideration the interests and preferences of other persons as well as one's own interests and preferences. They also agree that one should be impartial and give equal consideration to the interests of other persons. In these particular respects, the GR captures the notion of equality of consideration that is found in utilitarianism. GR* expresses a moral rule that is similar to the Principle of Utility (PU) as understood by a preference utilitarian. If we assume that utility can be measured in terms of preference satisfaction, the preference utilitarian version of the GR would be to "Do unto others so as to maximize aggregate preference satisfaction of all concerned, where each person's preferences should be considered to have equal weight to one's own." This is similar to GR* which advises us to treat others as they want to be treated. However, the idea that people's preferences provide an adequate conception of the good is highly doubtful for reasons already discussed.
Perhaps only a fully developed virtue ethics can provide us with an adequate conception of the personal good that would apply with equal force to oneself as well as to others. According to the virtue ethics view what we ought ideally wish for ourselves is what will make us better people; but "better" in the virtue sense, that is, what will make us more just, more courageous, more charitable, more wise, and so forth. Ideally we should all wish that we become the most virtuous people we can be. This leads to a version of GR*I which tells us that we ought to treat others in ways that will make them more virtuous persons. Ideally, everyone ought to desire that they become more virtuous persons, since on this view, the attainment of virtue is considered to be the highest good for human beings. We should aim for eudaimonia (moral virtue, flourishing, or happiness) both for ourselves as well as for others.
This then leads to another version of the GR:
GR*E: Treat others so as to promote their eudaimonia.
GR*E is more adequate than the earlier versions, but is still lacking as a practical moral rule because it is not easy to interpret and apply. The specific actions that one ought to take in order to fulfill this maxim will vary depending on both the characteristics of the agent and also those of the patients to whom the actions are directed. Rather than a guide to practical action, the GR*E can function as a way of interpreting the Golden Rule so as to fulfill Weiss's conditions (b) and (c), but applying it depends on our knowing what virtue consists in and how best to promote it.
REFERENCES
Paul Weiss. "The Golden Rule." Journal of Philosophy, Vol. 38. No. 16 (1941): 421-430.
