One of the most important methods for evaluating moral claims - both one's own and those of others - is the method, or technique, of counterexamples.  But this method is also applicable to factual claims that involve no moral content, and since the situation is simpler there, it may be best if we begin with its use in that area.

1.1  Non-Moral Generalizations and the Technique of Counterexamples

        Generalizations can vary greatly in complexity.  But it will suffice, for understanding what the technique of counterexamples is, if one considers generalizations of the following two simple types:

(1)  Anything that has property P ALWAYS has property Q as well;;

(2)  Something that has property P NEVER has property Q.

        Consider, for example, the following generalization:

"All swans are white."

        The technique of counterexamples points, incidentally, to one habit that it is very important to acquire - the habit of taking generalizations seriously.  Thus, if someone asserts that something is always the case, or that it is never the case, you should immediately ask whether there are any counterexamples.  If it has been claimed that all Ps are Qs, can one find a case of a P that is not a Q?  Or, if it has been claimed that Ps are never Qs, can one point to a case of a P that is a Q?

        Some people, by contrast, tend to put forward generalizations very casually, and when confronted with a counterexample, respond by saying, "That's the exception that proves the rule."  The appropriate reply is simply:  "No it isn't.  It's a case which shows that your generalization is false, and so has to be abandoned."

1.2  Moral Claims and the Technique of Counterexamples

        When the technique of counterexamples is applied to general moral claims, the same basic idea is involved as in the case of purely descriptive or factual generalizations that do not involve any moral content.  There are, however, some special considerations that arise when the technique of counterexamples is applied to moral claims, and that need to be kept clearly in mind.  These involve:
 

(1)  The distinction between absolute moral rules, and prima facie rules;

(2)  The idea of doomsday-style counterexamples;

(3)  The value of counterexamples that are not of the doomsday variety;

(4)  The use of purely imaginary cases;

(5)  Counterexamples and rights claims;

(6)  The scope of the method of counterexamples;

(7)  The reason why the method is so often fruitful;

(8)  Searching for the most effective counterexamples.

1.2.1  The Distinction between Absolute Moral Rules, and Prima Facie Moral Rules

        The first point, and perhaps the most important, is that many moral statements can be interpreted in either of two very different ways.  Consider, for example, the following claim:  "Adultery is always morally wrong."  A person who utters this statement might be saying either that

(1)  Adultery is always wrong, all things considered, regardless of the circumstances, and regardless of the consequences,

(2)  Adultery is always prima facie wrong; it is always wrong in itself; it is always wrong other things being equal.

        What is the difference between these two claims?  Suppose that Mary is in a rather odd situation, in which the only way in which she can save the life of some innocent person is by committing adultery.  Someone who maintains that adultery is prima facie wrong might say that, although adultery was always wrong in itself, nonetheless Mary should commit adultery in order to save the life of the innocent person.  In contrast, someone who held that adultery was always wrong, all things considered, regardless of the circumstances and regardless of the consequences, would say that Mary should not commit adultery, even though an innocent person will die if she doesn't.

        One needs to distinguish, therefore, between absolute moral rules and prima facie moral rules.  An absolute moral rule is one that asserts either that a certain sort of action is wrong without qualification, and so should never be performed, or, alternatively, that an action is obligatory, and so should always be performed in the appropriate circumstances.  A prima facie moral rule, in contrast, is one that specifies what have been called right-making and wrong-making properties of actions - a right-making property being one that makes it obligatory, other things being equal, to perform an action, and a wrong-making property being one that makes it wrong, other things being equal, to perform an action.

        The importance of this distinction in connection with the technique of counterexamples lies in the fact that it is often more difficult to find a plausible counterexample to a moral statement when it is interpreted as expressing a prima facie moral rule than when it is interpreted as expressing as absolute moral rule.  Someone who holds that adultery is absolutely wrong is saying that one ought not to commit adultery even if that is the only way to prevent a nuclear holocaust, and many people would view this case as a convincing counterexample to the claim that adultery is absolutely wrong.  But cases such as this leave untouched the claim that adultery is prima facie wrong.  A person who advances the latter claim can happily concede that one ought to commit adultery if that is the only way to prevent a nuclear holocaust, while maintaining that adultery is always wrong in itself.  For what he or she can say is, first, that while the fact that an action is a case of adultery is a wrong-making characteristic of that action, in the case we are considering the action also has a right-making characteristic - namely, that of being an action that saves the life of an innocent person - and, secondly, that the latter, right-making characteristic is more significant than the former, wrong-making characteristic, so that, thirdly, the action in question is one that is right, all things considered.

        How, then, does on go about constructing counterexamples to claims that some action is prima facie morally wrong?  This is an important question, and it will be addressed later, in section 1.3.

1.2.2  The Idea of Doomsday-Style Counterexamples

        Some counterexamples to absolute moral claims involve an instance of the type of action that the person is claiming is always morally wrong where not performing the action in those particular circumstances will have some extremely bad consequences - such as, in the above example, the death of an innocent person, or, more dramatically, the destruction of all life on earth.  Hence the label: "doomsday-style" counterexample.

1.2.3  The Advantage of Non-Doomsday-Style Counterexamples

        Doomsday-style counterexamples can be very effective against many absolute moral claims.  But they have the disadvantage that they cannot be used against prima facie moral claims, since there the person who is advancing the moral claim can simply reply that while the action in question is not wrong in the doomsday case, the reason is that the prima facie wrongness of performing the action is outweighed by the prima facie rightness of performing an action that will prevent a horrendous outcome.

        In addition, if someone advances an absolute moral claim, and you respond by offering a doomsday-style counterexample, the person may grant that your counterexample has force against the absolute moral rule, but then shift to the corresponding prima facie rule - at which point your counterexample will not have any force.

        Non-doomsday-style counterexamples have the advantage, then, of greater applicability, because they apply to prima facie moral rules, as well as absolute ones, and, as a consequence, someone who initially advances an absolute moral claim cannot escape your counterexample by shifting to the corresponding prima facie claim.

1.2.4  The Use of Imaginary Cases

        Confronted with a factual generalization, such as that all swans are white, it will be of no use to appeal to an imaginary case of a pink swan.  A counterexample to the generalization must be some actual swan that is not white; hypothetical non-white swans do not count.  The situation is different, in this respect, in the case of at least some moral generalizations.  For at least some moral generalizations are concerned, not only with all the cases of actions of a given sort that actually arise, but with all possible actions of that sort.  So if, for example, someone maintains that it is prima facie wrong to break a promise, he or she is not saying merely that every case of breaking a promise throughout all of history has some property which, other things being equal, will make the action wrong.  The claim is rather that the property of being a case of breaking one's promise is a wrong-making property so that there could not possibly be an action that had that property which did not have at least one wrong-making property.  To claim that it is prima facie wrong to break one's promises is therefore to make a claim that covers both actual and possible actions that involve the breaking of a promise.

        The upshot is that in applying the technique of counterexamples to moral claims, it will often be unnecessary to point to a case of some actual action that falls under the moral generalization, but that fails to have the moral property which, according to the generalization, it ought to have.  A purely imaginary, or hypothetical case, will often suffice.  This is illustrated by the discussion above of the claim that adultery is always wrong, all things considered, and regardless of either the circumstances or the consequences.  For it is unlikely that there has ever been an actual case where the only way in which someone could prevent a nuclear holocaust was by committing adultery.  The case is therefore a purely hypothetical one.  This in no way makes it, however, a less effective counterexample.

1.2.5  Counterexamples and Rights Claims

        Some moral generalizations involves claims about rights.  What does it means for something to have a right?  At the very least, for A to have a right not to be treated in a certain way - call it M - it must be at least prima facie wrong to treat A in way M.  But that is not sufficient, since the fact that it is wrong for Mary to drink Sue's beer does not mean that the beer has a right not to be drunk, or a right not to be drunk by Mary.  The right here belongs to Sue, not to the beer.  Thus, to say that something has a right is not merely to say that it would be wrong to treat it in a certain way: it is also to say that if one were to treat it in that way, one would be wronging the thing in question, rather than wronging something, or someone, else.

        In short, if one is considering a claim that something has some right, one must ask not only (1) whether treating it in a certain way would be wrong, but (2) if so, whether such treatment would also wrong the thing in question.

1.2.6  The Scope of the Method of Counterexamples

        How widely applicable is the technique of counterexamples?  The answer is that it is potentially applicable to every moral claim that anyone advances.  But how can this be so, given that while some moral claims take the form of general claims, others do not?  The answer is that in the case where someone puts forward a moral claim that, rather than being general in form, concerns only what someone did, or what someone is considering doing, in a particular situation, a defense of the specific claim will ultimately have to appeal to general moral principles.  As a consequence, every moral claim either is, or depends upon, some general moral claim, and this means that the technique of searching for counterexamples to generalizations is a technique that is always relevant in any moral discussion.

1.2.7  The Reason Why the Method is So Often Fruitful

        People sometimes wonder how the technique of searching for counterexamples can possibly be an effective technique to use in moral discussions.  For suppose that John claims that all actions of type M are morally wrong, and Mary tries to convince John that he is mistaken by describing a certain action which is of type M, but which she believes is not morally wrong.  Given that John believes that all actions of type M are wrong, won’t he simply reject her proposed counterexample, and hold that the action which she describes is morally wrong?  So that while the case that Mary proposes as a counterexample to the generalization may seem to Mary to be a plausible counterexample, it will not seem so to John, and so it cannot help their discussion to move ahead.

        The answer is that people, in advancing a general moral claim, do not always consider all of the different sorts of cases that fall under the general claim.  A person’s acceptance of a generalization may often rest upon quite a limited range of cases.  As a consequence, if Mary focuses upon a case ? either actual or hypothetical ? which is of a sort that John may not have considered, then she may very well come up with one where his moral intuitions do not square with those of the general moral claim which he has advanced.  What is a counterexample for Mary may therefore be a counterexample for John as well, and thus provide him with a reason for modifying or abandoning his original moral claim.

1.2.8  Searching for the Most Effective Counterexamples.

        Finally, it is important to keep in mind that one is searching for counterexamples that are as uncontroversial as possible, since the goal is to find something that will be effective in providing a person who accepts the generalization with a reason for entertaining doubts about its acceptability.

1.3  Finding Counterexamples to Prima Facie Moral Claims

        We have seen that it is much easier to find potentially convincing counterexamples to principles that are advanced as absolute moral rules than to principles that are advanced as prima facie moral rules, since doomsday-style counterexamples often work in the form case, but not in the latter.  How, then, does on go about locating potentially effective counterexamples when one is confronted with what is being claimed to be a prima facie moral rule?

        Suppose that someone claims that all actions of type A are prima facie wrong.  What properties will a potentially effective counterexample need to have?  The answer would seem to be as follows:

 
(1)  The action in question must be one of type A which most people would not view as morally wrong, all things considered.

(2)  Either there must be no right-making characteristic at all that is present in the action in question or else, if there is a  right-making characteristic that is present, it must not be sufficiently weighty that it could explain why the action in question is not wrong, all things considered.

        The idea underlying condition (2) is this.  First, if there is an action of type A that is not morally wrong all things considered, and that possesses no right-making characteristic, then that action cannot possess any wrong-making characteristic either, and so it cannot be the case that being an action of type A is itself a wrong-making characteristic.  Secondly, if the action in question does possess a right-making characteristic, but one that is not sufficiently important to outweigh the purported wrong-making characteristic of being an action of type A, then the right-making characteristic cannot suffice to explain why the action in question is not wrong all things considered, and so there is no alternative to the conclusion that being an action of type A is not a wrong-making characteristic.

        Let us now try finding effective counterexamples to some prima facie moral claims.  Consider, for example, the following five moral claims:

1.  Adultery is always prima facie wrong.

2.  It is always prima facie wrong to kill someone who wants to live.

3.  Breaking the law is always (at least) prima facie wrong.

4.  Divorce is always prima facie wrong

5.  Pleasurable experiences are always prima facie good.

        Let us now consider how one might arrive at promising counterexamples to the above five claims.  (Before reading on, try coming up with a plausible counterexample for each of the above five moral claims.  This will not always be easy.  If you have difficulty, see if the hints and suggestions that follow help.)

Hints and Suggestions

For claim 1: Try thinking about harm, deception, and breaking promises.

For claim 2:  Think about someone who is, or has been, very naughty.

For claim 3: Don't confine yourself just to laws that exist at the present time.  Consider laws that existed at earlier times, both in this country and in other countries.

For claim 4:  Try thinking about divorce and the idea of harm.

For claim 5: Can you think of any pleasures that do not make the world a better place?

Discussion

1.  Adultery is always prima facie wrong.

        What explanation might people offer of the wrongness of adultery?  There are many possibilities, but some are as follows:  (1)  Adultery involves breaking a promise to be sexual faithful to one's spouse.  (2)  Adultery involves deceiving one's spouse, or at least not being honest with him or her about a very important matter.  (3)  Adultery involves a significant chance that one's spouse will find out what one has done, and be harmed by it.  (4)  Adultery may destroy one's marriage, and, thereby, inflict suffering upon one's children.

        The idea, then, is to think of a case of adultery where the factors just listed are absent.  For example:  John and Mary are a swinging couple with no children who give each other permission to have sexual relations with others at a party they are both attending.

        Now, of course, not everyone would concede that this is a sound counterexample.  One might think, for example, that marriage is a contract that involves not just John and Mary, but God as well, and that in committing adultery, even with each other's permission and approval, John and Mary are breaking a promise that they made to God, and not just to one another.  Nevertheless, this is still a very promising counterexample, and one that many people would view as sound, and the key to arriving at it was to construct a case where various factors that bear upon the morality of adultery are not present.

2.  It is always prima facie wrong to kill someone who wants to live.

        Are there cases in which you would think it was justified to kill someone?  Suppose that you were being attacked by someone who was armed, and who it appeared wanted to kill you.  Or suppose you knew that the person in question was America's most wanted serial killer?

        In short, one good type of counterexample to claim 2 involves the case where you kill a guilty aggressor in self-defense, and the crucial thing that you needed to notice about claim 2 was that it did not specify that the person in question was innocent.

        Once that is noticed, another potentially effective counterexample may spring to mind as well - namely, the case of capital punishment involving a person who has been found guilty of some horrendous series of crimes.  (Since many more people reject capital punishment than reject killing in self-defense, this second counterexample will not be quite as effective as the first.  But it is still very relevant, and quite strong.)

3.  Breaking the law is always (at least) prima facie wrong.

        Here one should ask what grounds there might be for disobeying a law.  One answer is that the law may itself be immoral, in that it may itself involve a violation of the rights of people.  Consider, for example, the following laws from earlier times:

(1)  Laws in America concerning the ownership of slaves, or racial intermarriage.

(2)  Laws in America denying women the right to vote.

(3)  Laws in Nazi Germany concerning the treatment of Jews and certain other groups.

(4)  Laws in the middle ages concerning the appropriate punishment - namely, execution - for holding heretical religious beliefs.

Can you think of any present-day laws that you would hold are unjust?
 

        Here, as in the case of adultery, it is helpful to ask what reasons there might be for holding that divorce is harmful.  One reason might be that divorce may harm either the man or the woman.  Another is that it may harm any children that they have.  Once again, then, the strategy is to construct a case where these sorts of consideration are not present.  For example: the people involved do not yet have any children, they have found that they both prefer living on their own to living with someone else, and both of them want to abandon their marriage, and go back to living as single individuals.

        Here, too, as in the adultery case, not everyone will agree that this is a convincing counterexample - since, here too, some will hold that marriage is a contract with God, and not just with one another.  But, once again, one has a potentially very effective counterexample, since one has constructed a case where factors that might make divorce wrong have been removed.

5.  Pleasurable experiences are always prima facie good.

        Some pleasurable experiences may be bad all things considered, because they may lead to consequences that involve, say, suffering that outweighs the pleasure.  But how could a pleasurable experience not be prima facie, or intrinsically, good?  How could a pleasurable experience not be such as, in itself, makes the world a better place?

        The answer emerges if one considers people who enjoy doing things that harm others - the psychopath who enjoys torturing and murdering people.  So the suggestion is that the world is not a better place if such a person gets great pleasure from such activity than it would be if he or she got only a little pleasure from it, and thus that sadistic pleasures are a counterexample to the claim that pleasure is always prima facie good.

        Some people might argue that the judgment that sadistic pleasure is not prima facie good is based upon a confusion, and that one is mistakenly transferring one's disapproval of the sadistic action over to the pleasure that results from that action.  I do not think that there is a confusion here.  But one could avoid this objection by shifting to a slightly different case where there is no sadistic action - namely, the case of a person who simply takes pleasure in any suffering that he or she happens to see people undergoing.