Those who believe in moral intuitions never offer an exhaustive list of them. They are right. They could be quite numerous. We can only make an inventory of the most widely discussed:
1. There exists a certain form of moral wisdom that recommends that we not “play around too much with nature”
2. Everyone prefers real pleasures caused by real things or people to artificial pleasures induced by a machine or by pills.
3. Everyone knows how to distinguish between a life worth living and another life that is entirely without interest.
4. No one is capable of imagining a moral world completely different from our own, in which it would be good, just, or admirable to inflict gratuitous suffering on others.
5. Intentions count in morality, which is not always the case in other domains, for example, in etiquette, where respect for rules is strict and admits no exception for good intentions.
6. No one gives the impression of seriously believing that there is nothing objective or universal in morality.
7. We tend to believe that, even if we are not really free to act otherwise than we do, this does not suffice to absolve us of responsibility for our actions and to render illegitimate the indignation of others at actions of ours that cause them harm.
8. Certain things have a moral importance (pursuing a politics of the Left or of the Right) and others do not (driving on the right or on the left).
9. Certain things are morally unworthy (behaving in a servile manner toward the powerful), and others are not (putting ourselves wholly at the service of those who are in urgent need of us).
The spontaneous, unthinking, nonlearned, universal, or innate character of these intuitions remains highly controversial, as does their precise meaning. Certain experimental studies show that people do not exactly have the intuitions that philosophers attribute to them. Others teach us that the predictions of certain philosophers regarding these intuitions are accurate. But the list of intuitions to be discussed is obviously not closed.
We can, on the other hand, fairly easily identify some elementary rules of moral reasoning, even if, of course, we cannot rule out the possibility that others may be discovered or that analysis may lead us to eliminate one of them because it is redundant or incoherent.
And as no one wonders (for the time being) if such rules are innate or learned, we are spared one more controversy in regard to them.
1The three best-known elementary rules of moral reasoning are: “ought implies can” (or “no one has to do the impossible”); “one cannot derive an ought from an is” (or “you must not confuse judgments of fact and judgments of value”); and finally, “like cases must be treated alike” (or “it is unjust to weigh two different weights using two different measures”).
We can certainly add a fourth rule, one that is a little less known but that is hard to do without when analyzing norms of permission, obligation, or prohibition: it is pointless to oblige people to do what they will necessarily do of their own accord; it is pointless to prohibit people from doing what they will not willingly do in any case.
In all, there are therefore four elementary rules of moral reasoning, at least.
In order to introduce a little order into the debate concerning them, I will refer to each by means of a letter, and I will rank them. The ranking is not intended to imply any precedence, however.
R1: One cannot derive an ought from an is.
R2: Ought implies can.
R3:
Like cases must be treated alike.R4: It is pointless to oblige people to do what they will necessarily do of their own accord; it is pointless to prohibit people from doing what they will not willingly do in any case.
R1: ONE CANNOT DERIVE AN OUGHT FROM AN IS
There are reasons for believing that people often act in a selfish, greedy, xenophobic, or sexist fashion. Let us admit that we are concerned here with an indisputable fact. Would it be logical to draw from it the conclusion that it is good to be selfish, greedy, xenophobic, or sexist, or that we ought to be?
No, and no one moreover seems to reason thus. Even the selfish, the greedy, the xenophobic, and the sexist look for other reasons (which will probably not be any better) to justify their attitudes. They too seem to apply the most famous of the rules of moral reasoning: “One cannot derive an ought from an is.”
In actual fact, “one cannot derive an ought from an is” is a more general rule. It places certain limits upon all reasoning in which it is a question of permission, obligation, or prohibition, and which for this reason is called “normative” or “deontic” (from the Greek deon, “ought”). It therefore also concerns juridical or epistemological reasoning.
It is sometimes called “Hume’s law” or “Hume’s guillotine,” for it is Hume who proposed the first rigorous formulation of it. Yet certain liberties have thereby been taken with the text in which he evokes the passage from certain factual assertions to moral injunctions.
2Indeed, in this text, Hume does not exclude every passage from what is to what ought to be. He simply observes that this intellectual movement should be explained, whereas it generally is not.
3 Now this passage has to be explained, for otherwise it remains irrational. In order to be faithful to Hume, we should therefore write: “one cannot derive an ought from an is, or at any rate not without further argument.”
4There exist other formulations of the rule “one cannot derive an ought from an is”
For example: “We must carefully distinguish judgments of fact—it is true, it is false—and judgments of value—it is good, it is bad—and avoid deriving the second from the first.”
But Hume does not speak about the passage from judgments of fact to judgments of
value. What poses a problem for him is the passage from facts to
norms of obligation or prohibition.
Karl Popper has proposed an epistemological formulation of the rule “one cannot derive an ought from an is” According to him, we can perfectly well agree as to the facts and continue to disagree over the norms. Hence the idea that the norms do not necessarily flow from the facts:
That most people agree with the norm “Thou shalt not steal” is a sociological fact. But the norm “Thou shalt not steal” is not a fact, and can never be inferred from sentences describing facts. This will be seen most clearly when we remember that there are always various and even opposite decisions possible with respect to a certain relevant fact. For instance, in face of the sociological fact that most people adopt the norm “Thou shalt not steal,” it is still possible to decide either to adopt this norm, or to oppose its adoption; it is possible to encourage those who have adopted the norm, or to discourage them, and to persuade them to adopt another norm. To sum up,
it is impossible to derive a sentence stating a norm or a decision or, say, a proposal for a policy from a sentence stating a fact.5
Another formulation is due to Poincaré: “we cannot derive a normative conclusion from nonnormative premises”
6It is in this logical formulation that the rule “from what is, we cannot derive what ought to be” has proved of most interest to philosophers, and it is in this same formulation that it has been the most fiercely contested.
Thus, John Searle has maintained that, from a factual statement such as “Smith has said to Jones ‘I promise to give you twenty dollars,’” it is possible
logically to derive the normative statement “Smith ought to give twenty dollars to Jones.”
7But the whole question is that of knowing if the premise of the reasoning “Smith has said to Jones ‘I promise to give you twenty dollars’” is purely factual or descriptive.
We might suppose, among other considerations, that this promise is only a promise if it has not been extorted and if it does not conceal a threat.
8Compare it with “Smith has
promised Jones that he will steal his DVD collection from him at the first available opportunity.” Can we derive from it “Smith
ought to steal Jones’s DVD collection”?
It is not sufficient to promise something for an obligation to flow from it. It is necessary that the promise be valid, that it not conceal a threat, that we see clearly the good it contains, and so on.
If we grant that the concept of promise contains some normative aspects of this kind, inseparable from its descriptive aspects, we could no longer say that the statement “Smith has said to Jones ‘I promise to give you twenty euros’” is purely descriptive. And Searle would no longer be able to assert that he has derived this normative statement from a purely descriptive statement.
Supposing, however, that Searle is in the right, and that his argument allows us to think that from certain descriptive statements we can draw normative conclusions, could we generalize this result?
Could we maintain that from every descriptive statement we can derive a normative conclusion? This does not seem to be the case.
Let us return to the argument rejected by Popper: “most people believe that one should not steal,” therefore “one must not steal.”
To argue thus is to commit what is called in informal logic the “paralogism of popularity,” that is to say, a mistake we encounter each time an argument concludes “from the popularity of a view … its truth”
9In reality, from the fact of everyone believing that P, it does not follow that P is true, and from the fact that no one believes that P, it does not follow that P is false.
In the same way, in order to justify a statement asserting that such an institution is good or just, other arguments are needed besides “this is what everyone believes” If this argument were well founded, the philosophers of Antiquity would have been right to assert that slavery was not immoral, since it was what everyone believed.
Those who accept the rule “one cannot derive an ought from an is” certainly do not think that we should avoid invoking the slightest detail relating to life in society in order to justify normative statements such as “you must not steal” or “you must always keep your promises.” But they will all tend, I believe, to judge that, from the fact of everyone acting in a certain way or believing it is good to act in this way, it does not follow that it is good to act thus or that it is our moral duty to do it. They will reckon, probably, that if this were the case, we should say goodbye to every moral critique, and let the clichés and the prejudices prevail.
R2: OUGHT IMPLIES CAN
According to the second rule, it is absurd to demand of someone that they do impossible things, such as being in two different places at the very same time.
It is generally formulated as follows: “ought implies can,” or in more popular versions “no one is held to the impossible.”
10It is the awareness of the existence of such a rule that causes us to find scandalous certain demands addressed to foreigners wishing to regularize their situation. In order to have employment, you need a place of residence, but in order to have a place of residence, you need employment, and so on. What is appalling about this kind of norm is the fact of its seeming to oblige us to do the impossible.
However, every in-depth discussion of the rule “ought implies can” requires that its principal terms, “ought” and “can,” be specified, and it is there that the difficulties begin.
Thus “ought” has at least two totally different meanings: probability (“it should rain,” “it should already be there”) and obligation (“you must pay me back what you have borrowed”). In “ought implies can,” “ought” must obviously be taken in the sense of the deontic modality of obligation. But it can involve an absolute duty or a duty that admits exceptions, a categorical duty that imposes itself irrespective of the aims of the agent or a hypothetical duty, with respect to certain aims of the agent.
The term “can” is likewise difficult to define. We may have in mind logical or physical possibilities, but also psychological possibilities. None is easy to identify. Thus, we may wonder whether the physical or psychological possibilities to which reference is made are those of a species or of an individual. What is possible for a particular individual is not so for other members of the same species. It is not impossible for human beings to run one hundred meters in under ten seconds, above all if they are drugged up on amphetamines. But it would be personally impossible for me to do it even with the most sophisticated products. I do not even know if I am still capable of running more than 20 meters without stopping.
These difficulties do not prevent the rule “ought implies can” and its popular equivalent, “no one is held to the impossible,” being of some use in the selection of obligations that have a meaning. It enables us to assert that it would be absurd to oblige people to run more quickly than the capacities of the species to which they belong allow. It could also justify the idea that it is absurd to oblige people to run more quickly than is personally possible for them.
R3: LIKE CASES MUST BE TREATED ALIKE
Among the thought experiments presented here, certain of them would not be so striking if “like cases must be treated alike” did not have the status of an elementary rule of moral reasoning.
Consider The Child Who Is Drowning in a Pond and A Violinist Has Been Plugged Into Your Back.
1. If you judge it to be monstrous to let a child die in order to preserve your new shoes and to avoid being under some pressure at work, you should also judge it to be monstrous to let children from poor countries die of hunger, when you would merely have to devote a tiny part of your income to saving them, for we are concerned here with like cases requiring a like response.
112. If you judge it to be morally acceptable to rid yourself of an intruder who wishes to immobilize you for nine months, you should also judge it to be morally acceptable to interrupt an unwanted pregnancy, for it is a matter of like cases having to be treated alike.
12
We may of course ask ourselves whether like cases—or merely analogous ones—are involved.
13We may ask ourselves whether similarity implies an absolute identity or a sufficient one, an identity in certain aspects (moral, for example) or in every aspect (such as the exact age of the child or the child’s color in the case of The Child Who Is Drowning in a Pond), which would be absurd.
But that takes nothing away from the value of rule R3, which is hypothetical and demands that we treat these cases alike if and only if there are reasons for thinking that they are alike in certain relevant respects.
R4: IT IS POINTLESS TO OBLIGE PEOPLE TO DO WHAT THEY WILL NECESSARILY DO OF THEIR OWN ACCORD; IT IS POINTLESS TO PROHIBIT PEOPLE FROM DOING WHAT THEY WILL NOT WILLINGLY DO IN ANY CASE
1. You are in the act of shutting the door when a person, who sees you doing this, nonetheless orders: “Shut the door!” At first sight, this order is redundant, absurd. What is the point of obliging you to do what you were in the process of doing?
2. You have decided to spend the whole day in bed because you feel that you have the flu. Let us suppose that a person forbids you from getting out of your bed when they know that you have neither the intention nor the means to leave it. At first sight, this ban is redundant, absurd. What is the point of forbidding you to do what you did not wish to do?
The examples could be multiplied. Our lives can be ruined by pointless obligations or prohibitions. Happily, philosophical reflection offers us some tools with which to rid ourselves of them. It is possible, indeed, to construct principles that serve to filter the norms of permission, obligation, and prohibition, to know which of them are coherent, intelligent, and valid and which are redundant, contradictory, and pointless.
Thus, a principle of normative parsimony, formulated by Kant, renders null and void the norms that oblige us to do what each of us would do naturally of our own accord (such as being in good health and happy).
This principle plays a very important role in his critique of the moralities of happiness. For Kant moral prescriptions that demand that we be happy would be ridiculous, for it is a goal that we inevitably aim at of our own accord.
14We could say that this same principle of normative parsimony renders null and void the norms that prohibit us from doing what we would not do willingly in any case (such as being ill or unhappy).
We would therefore have two rules of normative parsimony, which we could baptize
Kant’s razor, in homage to his argument against the moralities of happiness, and with reference to the famous Occam’s razor.
15We could formulate them as follows:
1. It is pointless to oblige people to do what they do necessarily of their own accord.
2. It is pointless to prohibit people from doing what they will not do willingly in any case.
These rules could play an extremely important role in the critique of the so-called naturalist theories of the moral sense, those that posit the existence of an innate moral faculty.
If people were equipped with an innate “moral sense,” how would we account for the unbelievable quantity of moral obligations and prohibitions in all known human societies?
