Is using a person simply as a means always unacceptable?
THE DRIVER’S DILEMMA
The driver of a trolley realizes that his brakes have gone when he is hurtling down a valley hemmed in by steep banks.
On the track, ahead of him, but some way off, there are five trolley workers carrying out repairs.
If the runaway trolley continues in its course, the five trolley workers will inevitably be crushed, since there is not enough room alongside the track for them to take cover.
However, as luck would have it, the main line divides shortly before reaching the five people, leading onto a narrow secondary track. The driver can avoid killing them if he diverts the trolley.
But, unluckily, another trolley worker is repairing the secondary track. The situation is much the same as on the main line, since there is not enough room alongside the track for the trolley worker to take cover, and the driver’s maneuver will inevitably result in his being crushed.
The driver is therefore faced with the following dilemma: he could choose not to intervene and let the five trolley workers be crushed on the main line, or he could intervene and divert the trolley, thereby causing the death of the trolley worker on the secondary line.
Is it morally permissible for the driver to divert the trolley?
1This thought experiment was devised by Philippa Foot in 1967.
2Her idea was to contrast the driver’s dilemma with the numerous other cases in which philosophers wonder whether it is morally permissible to sacrifice one person in order to save several others.
One of the most celebrated cases she evokes is that of a spelunker so fat that he gets stuck when attempting to get out of a cave and will have to be blown up with dynamite in order to save the lives of the other spelunkers trapped within.
3 Would it be morally permissible to do this?
But Philippa Foot preferred to compare the trolley driver’s dilemma with the actions of a surgeon who kills a person in good health, dismembers them in order to remove the organs, and transplants them into five patients in order to save their lives.
According to her, we all have the intuition that it is not permissible for the surgeon to do what he does, but we all also have the intuition that it is permissible for the trolley driver to divert his engine toward one trolley worker in order to save five others. Yet in both cases it is a question of sacrificing one person in order to save five. Wherein does the difference lie? Are these intuitions not contradictory? How is one to justify them?
In a series of articles extending over more than thirty years, Judith Jarvis Thomson has proposed several variations on the original theme, in an attempt to refine the argument.
4 What follows are the two most important variations.
THE DILEMMA OF THE WITNESS WHO COULD THROW THE POINTS LEVER
You are strolling beside the tracks when you witness the scene described above. You quickly grasp that a trolley driver who is hurtling down a valley hemmed in by steep banks has lost consciousness. You see the five trolley workers trapped on the track and doomed to be crushed. What should you do? As luck would have it, there is a points lever right next to you. If you throw it, the trolley will be sent down a secondary track.
But, as bad luck would have it, another trolley worker is working on the secondary track. If you throw the lever, the trolley worker will inevitably be killed.
You are therefore faced with the following dilemma: either you fail to intervene and you let the five trolley workers be crushed on the main track, or else you intervene by throwing the points lever and cause the death of the trolley worker on the secondary track.
Is it morally permissible for you to throw the lever?
THE DILEMMA OF THE WITNESS WHO COULD PUSH THE FAT MAN
You are on a footbridge when you see on the track beneath you a trolley hurtling along and, on the other side of the bridge, five trolley workers working on the rails. You grasp immediately that the trolley will not be able to stop. But you know enough physics to understand that if a large object were to be thrown at that moment onto the track, the trolley would inevitably stop. Now, a fat man, who seems to have the necessary volume and weight, is in fact on the footbridge right next to you. He is leaning over the parapet. He is waiting for the trolley to pass, suspecting nothing. Just a flick of the finger would send the fat man toppling over onto the track.
Is it morally permissible for you to do this?
Judith Jarvis Thomson informs us that “most people” when told these two stories reckon that it is morally permissible to throw the points lever, but not to push the fat man onto the track in order to stop the runaway trolley. She herself shares these intuitions. But how is one to justify them? How is one to account for this moral asymmetry? After all, when we push the fat man, we are doing nothing else but causing the death of one person in order to save the lives of five others, that is to say, exactly the same thing as when we deliberately divert the runaway trolley onto the secondary line where there is just the one person.
Is it not somehow inconsistent to judge that there is a significant moral difference between the two examples?
Thomson does not think so. On the contrary, she reckons that our intuitions are consistent because they enable us to grasp the problem of rights that has been raised.
According to her, two features in fact characterize the dilemma of the witness at the points lever:
1. The agent saves the five persons by transferring to one isolated person the danger that threatens them.
2. The agent does not resort to any means liable to constitute in itself a violation of the rights of the isolated person.
Taking up again a comparison I have already used, I would say that, for Thomson, the situation resembles to some degree that of a pilot who would opt to crash his plane into the least densely populated zone of a big city.
It is a way of minimizing the number of deaths liable to be caused by a threat that anyway already exists, and that will bring about deaths whatever the agent does. We are simply “diverting the fatality” without infringing the fundamental rights of anyone whatsoever.
According to Thomson, in the dilemma of the fat man, we are also simply “diverting the fatality” We save five persons by transferring to one isolated person the danger that threatens them. But we do it by violating the fundamental rights of the fat man. It is because we are aware of this difference that we judge the two cases so differently.
Judith Jarvis Thomson constantly refers to “ordinary intuitions.” She contests sophisticated philosophical conceptions by invoking “what people think,” that is to say, what she thinks people think. But what do people really think? How do they justify their judgments?
FROM THOUGHT EXPERIMENT TO “SCIENTIFIC” STUDY
The most ambitious inquiry into these questions was conducted by a team of psychologists directed by Marc Hauser in the context of a massive investigation on the Internet between 2003 and 2004.
5The results were published in 2007, in other words, half a century after Philippa Foot’s thought experiment (one proof, among many, that a little story can have major consequences in the sphere of moral reflection).
Analysis of the answers to the whole set of experiments by and large confirms Judith Jarvis Thomson’s intuitions: for most people it is morally permissible to throw the points lever but not to push the fat man onto the track in order to stop the runaway trolley. However, these answers do not validate Thomson’s interpretation so far as rights are concerned.
Hauser’s inquiry involved around twenty-six hundred people, of both sexes and from several age sets, religions, levels of educational attainment, and ethnic or cultural groups, as well as in several different countries, namely, Australia, Brazil, Canada, India, the United States, and the United Kingdom.
Some respondents had been exposed to moral philosophy (a little over five hundred of them), while others had not. Several scenarios were outlined on separate sheets of paper and selected at random.
6 They differed in various respects from the original thought experiments devised by Philippa Foot and Judith Jarvis Thomson.
In the driver’s dilemma, it is a passenger who has to seize the controls if need be, for the driver fainted when he realized that his brakes had failed. The witness’s dilemma has been replaced by a pair of much more complicated dilemmas, which I will present separately below in order not to confuse the reader too much.
These scenarios are designed to enable us to determine whether reference to the doctrine of double effect can make it possible to understand the moral thinking of each and every one of us.
What is the “doctrine of double effect”?
When the psychologists conducting the inquiry use this term what they have in mind is a very simplified version of a complex notion that philosophers have elaborated over centuries of debate.
This moral doctrine, the original formulation of which is ascribed to Thomas Aquinas, designates two effects, one good, the other bad, of an action that, considered in itself, is good, or at any rate neither good nor bad.
One might think of the bombarding of a bunker in which the high command of a cruel army waging an unjust war is hiding, but in which civilians are also sheltering. One of these effects—the elimination of unjust aggressors—is good. This is what the action is designed to achieve, what its authors want. The other effect—the killing of innocent civilians—is bad. The authors of the action anticipate it, deeming it to be an inevitable “collateral effect.” But it is not this latter effect that is aimed at, or intended by its authors. It is not even conceived as a means of arriving at the result aimed at.
According to the doctrine of double effect, this kind of action having two effects is morally permissible under these conditions (the bad effect is not aimed at, it is neither an end nor a means). But the harm caused (in terms of innocent victims, for example) must not be
disproportionate.7The psychologists conducting the inquiry seem sometimes to be reducing this complex and controversial doctrine to Kant’s formula: “Act in such a way that you always treat humanity, whether in your own person or in the person of any other, never simply as a means, but always at the same time as an end”
8But they are not required to be the exegetes of medieval theories, and once we see with sufficient clarity what they have in mind when they speak of the “doctrine of double effect,” everything is fine. It is the notion that we must distinguish between those cases in which we treat a human person simply as a means (the fat man) and those in which we do not do so (the driver who diverts the trolley, the witness at the points lever).
What is important and enables us nonetheless to place the empirical inquiry on the same plane as the thought experiment is the fact that the three significant differences between the variants have been respected:
1. action aimed at diverting a threat (to the five trolley workers) or action creating a new threat (to the fat man);
2. action causing the death of one of the persons as a means of stopping the runaway trolley (pushing the fat man) or as a collateral effect of the fact of having diverted it (by throwing the points lever);
3. action causing a death in an impersonal fashion (by diverting the trolley or by throwing the lever) or in a personal one (by giving the fat man a shove).
We must also insist upon the fact that, in the original thought experiments, no mention was made of personal ties between the participants, of how visible the trolley was, or of the possibility of stopping it by some other means. The inquiry treats all these things as given.
What were the results?
A PAIR OF THOUGHT EXPERIMENTS: DIVERTING THE TROLLEY AND PUSHING THE FAT MAN
Of all the people taking part in the experiment, 89 percent judge it to be morally permissible to divert the trolley onto the secondary track on which there is a trolley worker, thus deliberately causing his death. There is no significant variation in terms of age, religion, gender, culture, educational attainment, or knowledge of moral philosophy.
Only 11 percent of the persons taking part in the experiment judge it to be morally permissible to push the fat man onto the track, thus deliberately causing his death.
9 There is no significant variation in terms of age, religion, gender, culture, educational attainment, or knowledge of moral philosophy.
These results are in complete harmony with Judith Jarvis Thomson’s predictions. Indeed, all of the respondents deem it morally permissible to divert the trolley, but only a tiny minority judge it to be morally permissible to push the fat man.
What strikes all those interested in these results is the unlikely convergence of the answers despite differences in age, religion, gender, culture, educational attainment, and exposure to moral philosophy.
The more optimistic proclaim that, thanks to the help given by psychologists, the philosophers have at last discovered a universal moral datum. The problem is that such data seem rather to show that what is universal in our moral reactions is their inconsistency!
THE PROBLEM
For 89 percent of respondents, it is morally permissible to divert the trolley and thus deliberately cause the death of one person in order to avoid letting five die. Upon reading this result, we could conclude that the respondents are overwhelmingly consequentialist, since it seems to them morally permissible to carry out an act that minimizes evil, independently of any consideration relating to the nature of the act itself. But only 11 percent judge it to be morally permissible to push the fat man in order to achieve the same outcome. What has become of the consequentialist intuition that seemed to inform the first judgment? If we think that it is morally permissible to deliberately cause the death of one person in order to avoid letting five other people die, why would it not be morally permissible to throw the fat man onto the track? Is it not a form of inconsistency to judge that there is a significant moral difference between the two?
Such is the problem we have to address here.
THE RESPONDENTS ARE ASKED TO JUSTIFY THEIR JUDGMENTS
When the respondents are asked to justify their judgments, they divide into three groups.
1. Those who offer satisfactory justifications.
They readily grasp what the significant differences between the cases are: physical contact or the lack of it; using another as a means or not doing so; averting an existing threat or introducing a new one.
They seek to justify their answers by taking such differences into account. They will say, for example, that it is permissible to divert the train but not to push the fat man onto the track, because it is acceptable to divert a threat but not to create a new threat.
2. Those who offer unsatisfactory justifications.
Some simply say that they are unable to justify their judgments. They find themselves declaring without any further justification that, in the one case, you cannot do other than let someone die, that it was inevitable, whereas in the second case you can abstain from killing. Others offer a consequentialist justification in the one case (it is better to save five persons than one, diverting the train is a lesser evil, and so on) and a deontological justification in the other (it is morally prohibited to kill, we have no right to take ourselves to be God and to decide who is to live or to die, not causing a wrong is more important than going to someone’s aid, and so on). But they are not able to see the contradiction and do not attempt to explain why we should be consequentialist in the one case and not in the other.
3. Those who offer justifications that are not pertinent (but that are at the same time more amusing). They have not grasped the significant differences and offer any old thing by way of justification: “A man’s body cannot stop a train,” “It’s an absurd story: the trolley workers would have heard the train coming and would have fled,” and so on.
According to Hauser, 70 percent of the respondents are incapable of justifying their spontaneous judgments.
The 30 percent who offer satisfactory justifications have no particular characteristics so far as religion, age, or gender are concerned.
The only significant factor, that is to say, the only factor that would have enabled us to predict the satisfactory justifications, is exposure to moral philosophy. Of the philosophers 41 percent (as against only 27 percent of the nonphilosophers) are capable of giving satisfactory justifications of their spontaneous judgments. But these figures also show that the simple fact of having studied moral philosophy is no guarantee that we will be able to justify our judgments in a consistent fashion, since 59 percent of the philosophers were incapable of doing as much (but perhaps they had not attended every course!).
Hauser, who devised the experiment, was not surprised by this outcome. One of his original hypotheses was that there exists a dissociation between moral intuitions—the rapid, and often not conscious, reactions that we all have—and their justifications, which we are often incapable of giving.
On the other hand, for Hauser moral intuitions, despite their rapidity and intensity, are not purely emotional reactions. We may without a doubt be concerned here with spontaneous judgments devoid of any affective content, with the kinds of principle applied with the utmost rapidity and unconsciously, and with complete conviction. This hypothesis is pivotal to Hauser’s construct. Indeed, it is this that he seeks to prove at all costs, since he is thereby able to envisage the possibility of our all being Kantians without being fully aware of it.
By means of his experiments, Hauser seeks to confirm that intuitions cannot be reduced to irrational emotional reactions lacking in moral relevance. This accounts for his bias in favor of a deontological interpretation of the results, one that is, however, not self-evident.
10There are in fact several ways of interpreting the spectacular difference between the results relating to the permission to divert the runaway trolley and those concerning the prohibition on pushing the fat man, even though the consequences of the two actions are identical.
THE CONSEQUENTIALIST INTERPRETATION
We formed a consequentialist judgment when we were faced with the dilemma of the witness who wonders whether he should throw the points lever. We judged that it was morally permissible to kill one person rather than five, when the alternative presents itself thus. If we were consistent, we ought not to change our conceptual framework when addressing the dilemma of the fat man. We should remain consequentialist and judge that it is morally permissible to push him even if,
personally, we would rather not do it. But that is manifestly not what happens. We react in the second case as if we had suddenly become fanatical deontologists who absolutely rule out certain actions even when they are for a higher good. Why is this?
One of the hypotheses that would account for this inconsistency has it that we are naturally “programmed” (through the evolution of our species) to be shocked by violent physical contact, and psychologically incapable of remaining coolly rational when faced with such a sight or thought.
11 The idea of pushing the fat man unleashes emotional reactions so intense that they block the processes of rational thought. This is why we would tend to distinguish between diverting the train and pushing the fat man, when the two cases are in fact morally equivalent.
12THE ROLE OF THE EMOTIONS
In order to sustain the above conclusions, which rank deontological reactions with irrational madnesses (with which perhaps not all deontologists will be best pleased), some researchers have invoked techniques borrowed from the neurosciences.
13When the possibility of pushing the fat man is evoked, it is the emotional areas of the brain that are presumed to be aroused. When the possibility of throwing the lever is evoked, it is the rational areas of the brain that are supposed to react.
14I obviously cannot pronounce upon the validity or otherwise of the data, having no particular competence in what is by common consent a highly problematic field of science, namely, that concerned with cerebral localization.
15The philosophical objection that I might nonetheless raise here is that, in a very general sense, from the fact of an action being motivated or accompanied by emotional reactions, it does not follow that it is itself irrational.
16THE DEONTOLOGICAL INTERPRETATION
It was in order to establish the inadequacy of this consequentialist and irrationalist explanation that Marc Hauser submitted to the respondents the following two cases, which are variants of the case of the witness throwing the points lever proposed by Thomson.
17We are concerned here with situations in which the fat man is used to block the runaway trolley, but without his having been pushed by the witness.
This time we must imagine that the secondary track onto which the witness can divert the trolley by throwing the points lever describes a loop and then returns to the main track.
THE CASE OF THE FAT MAN ON THE LOOP
A fat man is on the loop. In hitting the fat man and causing his death, the trolley will be greatly slowed down, and this will give the five trolley workers time to flee and to save their lives.
Is it morally permissible to throw the lever?
THE CASE OF THE MASSIVE OBJECT AND THE TROLLEY WORKER
There is a massive object on the loop behind the trolley worker. In hitting the object, the trolley will be greatly slowed down, thus giving the five trolley workers time to flee and to save their lives. But the trolley worker who is in front of the massive object will inevitably be killed.
Is it morally permissible to throw the lever?
In the pair of scenarios “with a loop,” the action taken is impersonal, without there being any violent physical contact with the victim.
However, in the first scenario, the witness who throws the points lever treats the fat man simply as a means to block the runaway trolley, which leads us to assimilate this case to the one in which the fat man is directly pushed.
In the second scenario, the trolley worker’s death is inevitable, but as a collateral effect of the fact that the trolley is heading for the massive object. The trolley worker is not treated simply as a means.
In this second scenario 72 percent of the respondents judge it to be morally permissible to throw the lever (the trolley worker is not treated simply as a means).
But in the first scenario only 56 percent judge it to be morally permissible to throw the lever (the fat man is treated simply as a means).
According to Hauser, the difference is statistically significant. It allows us to conclude that in the four scenarios, with and without a loop, the respondents judge that we can cause the death of one person in order to save five, if we do not treat him simply as a means. Furthermore, when the trolley is stopped on a loop, there is no violent personal contact, which enables us to eliminate the irrationalist interpretations.
For Hauser, taken as a whole, these results display a kind of consistency. In his view, everything happens as if the respondents were intuitively applying the doctrine of double effect, as it is understood by the authors of the experiment.
It is deontological, in the sense that, in the authors’ opinion, it absolutely excludes treating intentionally a human person simply as a means.
In the two scenarios “without a loop,” the doctrine might be phrased: it is permissible to divert the train but not to push the fat man. By diverting the train, we are certainly causing the death of the trolley worker, but we are not treating the trolley worker simply as a means to save five lives. On the other hand, that is precisely what would happen if we were to push the fat man.
But this pair of scenarios can admit an irrationalist interpretation, since there had been violent physical contact. In the two scenarios “with a loop,” the “violent personal physical contact” factor is eliminated. Yet the respondents continue to judge it to be far more inadmissible to treat the fat man simply as a means than to cause the death of one person as a collateral effect.
For Hauser, these results bear out his hypothesis: it is not the idea of violent personal physical contact that explains why we overwhelmingly reject the idea of pushing the fat man, but
the thought that we are treating him simply as a means.
18The results thus lend credence to the idea that the sample in question tends to employ the doctrine of double effect (as interpreted by Hauser).
19In other words, the best interpretation of the
full set of the results would not be consequentialist and irrationalist but deontological and rationalist. It shows that what counts for most people is their not treating a person simply as a means.
This interpretation is still contested by the irrationalists, on the basis of the following hypothesis. And if it were possible to cause the fat man to fall onto the track, with the aim of blocking the runaway trolley, without any violent physical contact, simply
by opening up a trapdoor beneath his feet?20 Would we be so shocked? Probably not.
However, we would thereby have treated the fat man simply as a means, and in exactly the same fashion as if we had pushed him! Before asserting that what shocks, when we push the fat man, is not the violent physical personal contact, but the fact that we are treating him simply as a means, we should perhaps devise some tests based upon thought experiments as simple as that of the trapdoor that opens up beneath the fat man’s feet.
21A DRAW BETWEEN DEONTOLOGISTS AND CONSEQUENTIALISTS
Marc Hauser is confident of having shown that people rapidly and unconsciously apply the doctrine of double effect (in his sense of the term).
If people find it permissible to throw the points lever but not to push the fat man, it is because a person is turned into a simple means in the second case but not in the first. Their moral intuitions are consistent, rational, and … deontological! But the consequentialists continue to defend the notion that the refusal to push the fat man is inconsistent, irrational, and due to the involvement of emotional factors. They reckon that the deontologists have not conclusively demonstrated that a feeling of revulsion at the idea of brutal, personal physical contact was not an important explanatory factor, but was perhaps in fact the decisive factor accounting for the different judgments regarding the fact of, on the one hand, throwing the points lever and, on the other, pushing the fat man.
22The debate surrounding the best interpretation of the spontaneous judgments of people exposed to the different scenarios of The Killer Trolley therefore ends up, at any rate for the time being, in “a draw”
THE POVERTY OF MORAL INTUITIONS
Numerous other research projects have been devised with a view to evaluating the above results. The latter should in fact lead us to doubt the solidity of our moral intuitions. For all we have to do is to add certain data to our presentation of the dilemmas, or simply to modify the way in which they are presented, and our moral intuitions alter.
23Various hypotheses lending credence to this supposition are plausible. Some of them have received empirical backing.
Our intuitions relating to the moral right to throw the points lever or to push the fat man might vary:
1.
According to the presumed moral qualities of the people involved. We should be less doubtful about the idea of pushing the fat man if we were told that it was he who had put the lives of the five trolley workers at risk by sabotaging the trolley, or if we learned that he was a sadistic torturer who went on footbridges to watch accidents.
242. According to whether the people at risk are relatives or are unknown to us. We would tend to be more moved by the fate of the trolley worker who was to be sacrificed if he were a friend or a member of the family.
3. According to the age of the people at risk of being sacrificed. We would tend to be more moved by the fate of the person who was to be sacrificed if they were young or very young.
4.
According to whether the people at risk of being sacrificed resemble us or not. We would be less inclined to sacrifice people who resemble us. Women would thus be less inclined to sacrifice their sisters, and men their brothers!
255.
According to the degree of responsibility of the people at risk. We would be less inclined to sacrifice people who were in no way to blame for being at the precise spot they happened to be at or whose duty it was to be where they were.
266.
According to the energy needed to obtain the result. We would be sensitive to the fact that more effort is required to push a fat man from a footbridge (especially if he resists) than to operate a points lever.
277.
According to whether the result is obtained by violent physical contact or not. We would be more shocked if the fat man were to be toppled by brute force than if his fall were to be brought about by opening a trapdoor beneath his feet.
8. According to whether the threat is diverted or created. We would tend to find less grave the action of diverting an existing threat without creating a new one (by throwing the lever) than that of creating a new threat (which is aimed at the fat man), even if it were with a view to diverting an existing threat.
9. According to the position occupied by the characters in the narrative. It would be a graver matter to threaten a person who was in a safe place (the fat man on the footbridge) than a person who was in a tight spot (the trolley workers).
10.
According to the order in which the stories are told. When we have been exposed first of all to a scenario of sacrifice as hard to bear as a transplant gone mad (one person is killed so that their organs can be removed and the lives of five other sick people saved), we will be less inclined subsequently to deem it permissible to throw the points lever, thereby causing the death of one person in order to save five.
2811.
According to the choices on offer. If the decision is taken to present only the case of the witness who pushes the fat man and at the same time to insist that there is no other way of saving the five trolley workers, the rejection of it will be less overwhelming.
2912.
According to the position that the respondent is supposed to assume. We all tend to be more moved by the fate of the trolley worker to be sacrificed when we ourselves are presumed to be responsible for throwing the points lever (in answer to the question “what would you do?”) than when we are presumed to be in the position of a moral judge who does not intervene (in answer to the question “what do you think about it? is it morally permissible?”).
30
If all of these hypotheses were to be confirmed, the upshot would be that the intuition whereby it is permissible to sacrifice one person in order to save five is operative only in cases in which:
1. We have to do with unknown persons, with whom we have no familial ties and no bond of proximity or friendship, and who have no gender or age, no rights, and no responsibilities;
2. Our point of view is that of a moral judge who does not intervene.
One cannot say that this would be a result decisively in favor of the idea that our moral intuitions are on balance consequentialist. It might simply allow us to think that our consequentialist intuitions, if we have any, are, most of the time, neutralized, because the cases in which they could be pertinent are very rare.
From another perspective, if all the above hypotheses were confirmed, the intuition “it is not permissible to treat a person simply as a means” would be completely neutralized or anaesthetized:
1. when treating a person simply as a means is the only choice possible;
2. when treating a person simply as a means does not require any violent physical contact;
3. when the person whom we treat simply as a means is morally repugnant;
4. when the person whom we treat simply as a means is at the origin of the threat;
5. when the person whom we treat simply as a means is responsible for the risks he has taken.
Results of this kind could establish that our minds do not have a tendency to orient themselves, in every case, consciously or unconsciously, around the principle: it is not permissible to treat a person simply as a means.
One cannot say that it would be a conclusive result in favor of the notion that our moral intuitions are on balance deontological. It might merely allow us to think that our deontological intuitions, if we have any, are often blocked because, in numerous cases, they do not take priority.
THE FRAGILITY OF MORAL INTUITIONS
One author has wondered whether, by increasing the quantity of people who could be saved by pushing the fat man (a hundred persons saved, let us say, instead of five), the intuition that forbids us to do it could be blocked. This does not seem to always be the case.
31But that does not rule out the possibility of there being certain “threshold effects” that weaken this type of intuition. The difficulty is in knowing where these thresholds are situated, and for what kind of people.
32Is it morally permissible to push the fat man in order to save a hundred, a thousand, two thousand, a million innocent adults?
And what if it were to save children, let alone our own children?
Once cases become too complicated, once the numbers exceed a certain threshold, once the qualities of people obtrude, our intuitions lose their solidity.
To summarize:
For a great many researchers, the problem posed by the experiment of The Killer Trolley is that of the consistency of the answers.
Why is there such an asymmetry between our ways of judging the act of throwing a lever to divert a runaway trolley toward one person, thereby bringing about his death, and those of causing the death of one person by pushing him off a footbridge in order to block a runaway trolley?
According to the so-called consequentialist interpretation, The Killer Trolley experiment shows that we are inconsistent, being under the sway of our irrational emotional reactions.
According to the so-called deontological interpretation, The Killer Trolley experiment shows that we are consistent. We apply rapidly, naturally, and effortlessly a principle that is well known in moral philosophy, whereby we are prohibited from treating a person simply as a means.
How are we to come to a decision?
The most difficult questions arise when we change the way in which we present dilemmas, adding particular details to the cases under discussion. We then realize that our intuitions change too. These intuitions appear to possess neither the robustness nor the degree of independence from their theoretical contexts that we generally expect to find in an intuition. The philosophical questions raised are then the following:
If we really do not have robust intuitions that could come to the rescue of deontological theories, how could we justify such theories?
If we really do not have robust intuitions that could come to the rescue of consequentialist theories, how could we justify such theories?
According to certain philosophers, appealing to moral intuitions is as vain and hopeless in ethics as it is in mathematics.
33 Is the comparison pertinent?
It seems to me that the contribution of experimental moral philosophy to armchair moral philosophy is a little clearer at present.
The empirical results show that the philosophers who defend theories, be they deontological or consequentialist, cannot rely upon ordinary intuitions to justify their point of view.
But that does not mean that they are wrong or that no other means of justifying a moral theory exists save that of appealing to ordinary intuitions.
Besides, other interpretations of these intuitions, of a kind that neither Hauser nor his critics have envisaged, are possible.
I have in mind here the interpretation in terms of rights that Judith Jarvis Thomson defended and that has never really been tested.
34We can also try to advance interpretations that would be founded neither on the doctrine of double effect, nor on the emotions, nor indeed on rights.
