Saturday, September 06, 2014

Churchill's Dilemma and the Neuroscience of Moral Judgment

In 1944, one week after the D-Day landing in Normandy, Winston Churchill faced a moral dilemma that philosophers today recognize as a "trolley problem."

During the evening of June 12/13, the Germans launched the first ten flying bombs or V1s towards London.  For the next few months, thousands would be launched, and over 2,000 reached London.  By the end of August, the British had become adept at shooting them down.  But then the Germans replaced the VI with the V2, a long-distance rocket with a warhead.

Scientists working for British Military Intelligence (MI5) had gathered information about the testing of the V1, and they concluded that the VI bombs often fell short of their targets.  This was confirmed when most of the V1s first launched fell on the neighborhoods of south London rather than on their intended targets in the center of London, which was densely populated and where the main offices of government were located.

Some people in MI5 realized that casualties and damage to the government could be reduced if the Germans could be deceived, so that they would not correct the targeting of the V1s.  The Germans had secret agents in England, but the Germans did not know that all of their agents had become double-agents for the British, and they had been supplying the Germans with misleading information.  The Germans asked their agents to report on the locations where the V1s were landing.  So if their agents reported that many of the V1s were overshooting their targets and landing north of London, the Germans would adjust their targeting so that the flying bombs would actually hit the south of London rather than the center.  People in MI5 recommended this plan to the government.  If it worked, thousands of people in south London would be killed, but many more people in central London would survive.

One of the scientists recommending this--R. V. Jones--knew that this would threaten his own family.  Years later, he explained: "I realized well that what I was doing was trying to keep the mean point of impact in the Dulwich area, where my own parents lived and where, of course, my old school was.  But I knew that neither my parents nor the school would have it otherwise" (Most Secret War, 1978, p. 421).

One member of Churchill's cabinet--Herbert Morrison--vehemently opposed this proposal as an "interference with Providence" that was immoral, because they would be responsible for killing some people who might otherwise have lived.  Churchill, however, accepted the argument that it was better to allow the foreseeable but not directly intended deaths of some people in the south of London if this prevented the deaths of many more people in the center of London.

As David Edmonds has pointed out in his new book--Would You Kill the Fat Man?--this dilemma resembles the trolley problem that I have written about previously.  In the Switch Case, we can prevent a runaway trolley from killing five people on the tracks by switching it onto a side track where it will kill one person.  When presented with this dilemma, most people around the world say they would pull the switch and thus sacrifice one life to save five lives.  That seems close to what Churchill chose in allowing thousands of innocent people to die so that many more could live.

In the Footbridge Case, by comparison, we can save the five people on the tracks by pushing a very fat man off a footbridge onto the tracks, and he's big enough to stop the train.  From a utilitarian point of view, we should decide this case the same way we decide the other case, because in both cases, the killing of one person is morally justified by the greater good of saving five people.  But most people around the world say they would not push the fat man.  And when they are asked to explain their reasoning, they often admit confusion about their motivation, and sometimes they'll say that pushing the fat man just feels wrong, even though the utilitarian principle of taking one life to save five apparently applies here.  (We might imagine Churchill having to decide whether he should personally push some people into the path of a flying bomb to save the lives of others.)

Edmonds and most philosophers today see this as a refutation of utilitarianism, or at least of pure utilitarianism.  Although our moral intuitions sometimes approve of utilitarian reasoning, as in the Switch Case, sometimes we reject utilitarianism, as in the Footbridge Case.

Fervent utilitarians like Peter Singer and Joshua Greene disagree with this.  They insist that maximizing happiness impartially is the fundamental principle of morality.  We all want to be happy, and we don't want to be miserable.  This universal pursuit of happiness becomes moral when we consider it from an impartial perspective.  I think I'm special in valuing my happiness and the happiness of those close to me.  But I can understand that you think you're special in the same way.  I can see, then, that if we are to achieve the benefits of social cooperation, we must recognize that each person's happiness is equal to the happiness of any other, and thus that the happiness of many might count for more than the happiness of few.  So if we are to maximize happiness impartially, we must push the fat man. 

The differences between the Switch Case and the Footbridge Case are not morally relevant, Singer and Greene claim, because all that counts morally are the consequences, which are the same: five lives are rightly saved at the expense of one.  The two cases might be different emotionally, but not rationally.  And morality requires that our rational comprehension of moral principles should prevail over any conflicting emotional responses.

How do we explain the fact that most people think the differences between these two cases really are morally relevant?  One explanation is that most people intuitively apply the doctrine of double effect, which was stated by Thomas Aquinas, and which has been associated with the moral teaching of the Catholic Church.  To explain why we prohibit intentional killing but permit killing in self-defense, Aquinas explained: "One act may have two effects, only one of which is intended and the other outside of our intention."  When we kill in self-defense, the death of the person threatening us is foreseen but not directly intended.  Our direct intention is to defend our life.  Killing the assailant is an unintended side-effect of our action.  So if the assailant were to stop his attack, we would not kill him, and this would be our preference.  If we could remove his threat to our life without killing him that would be the morally preferable choice.

This principle has become a fundamental standard in both domestic law and international law.  Homicide can be justified if the death is a foreseen but not intended side-effect of defending our lives or the lives of others.  In war, it is unjust to directly target innocent people--civilians or soldiers who have surrendered.  But it is just for some innocent people to die as a side-effect of an attack on a military target, as long as there have been reasonable efforts to minimize the danger for innocent people.

A full statement of the principle of double effect has been suggested by John Mikhail: "an otherwise prohibited action, such as battery or homicide, which has both good and bad effects may be permissible if the prohibited act itself is not directly intended, the good but not the bad effects are directly intended, the good effects outweigh the bad effects, and no morally preferable alternative is available" (Elements of Moral Cognition, 149).

Mikhail explains that most people feel that they should not push the fat man off the footbridge, although they would switch the trolley onto the side track and thus kill the man on the track, because they have an instinctive knowledge of the norm of double effect, and they sense that pushing the fat man is intentional homicide, while pulling the switch results in foreseeable but unintentional homicide.

Utilitarians like Greene and Singer argue that this distinction between foreseeing and intending a death is irrelevant to moral judgment: if it is right to kill one person to save five, then it doesn't matter morally whether this killing was directly intended or only a foreseeable side-effect of directly intended action to save the five, because the consequences are the same.

Greene concedes, however, that the human brain has evolved to recognize the principle of double effect.  In Greene's own research, in scanning the brains of people while they pondered trolley problems, he has shown that when people are thinking about the Switch Case, there is increased neural activity in the dorsolateral prefrontal cortex (DLPFC); and when they are thinking about the Footbridge Case, there is increased neural activity in the ventromedial prefrontal cortex (VMPFC) and the amygdala.  The neural circuits in the DLPFC enable cognitive control in which the brain organizes thought and action in accordance with decision rules that might require overriding competing impulses.  The neural circuits in the VMPFC and the amygdala enable strong emotional responses.

So when we are making utilitarian judgments (such as sacrificing one life to save five), Greene infers, our DLPFC is active, and the emotional response in the VMPFC and the amygdala is weak, which is what happens to most people in deciding the Switch Case.  But in the Footbridge Case, the thought of pushing the fat man arouses such a strongly negative emotional response in the VMPFC and the amygdala in most people that the utilitarian judgment is overridden.  Those people who do accept the utilitarian judgment in the Footbridge Case are able to exercise cognitive control in applying the utilitarian decision rule ("Do whatever will produce the most good") while suppressing the emotional inhibition against directly killing someone (Moral Tribes, 116-131). 

Here, then, we can actually see the conflict in the brain between reason and emotion in deciding whether the rights of the individual should take precedence over the greater good.  Greene sees this as supporting his "dual process theory" of moral judgment as based on the interaction between automatic emotions and deliberate reasoning.

It is a mistake, however, to think of this "dual process" in the brain as a sharp dichotomy or conflict between reason and emotion, with utilitarianian judgment requiring reason to conquer emotion.  Green thinks that neuroscience confirms David Hume's argument that reason is the "slave of the passions" (134-37, 184-85, 190-92, 201, 291, 328, 352, 385).  Pure reason by itself cannot move us to action without some motivating emotion or desire.  After all, utilitarianism is rooted in the natural fact that our wanting to be happy, and wanting not be miserable is an evolved gut feeling. 

Greene notes that the DLPFC and the MVPFC are interconnected, and he also observes: "the DLPFC, the seat of abstract reasoning, is deeply interconnected with the dopamine system, which is responsible for placing values on objects and actions.  From a neural and evolutionary perspective, our reasoning systems are not independent logic machines.  They are outgrowths of more primitive mammalian systems for selecting rewarding behaviors--cognitive prostheses for enterprising mammals.  In other words, Hume seems to have gotten it right" (368).

The Footbridge Case sets off emotional alarm bells in our brains in a way that the Switch Case does not.  Why?  Greene explains this through a "modular myopia hypothesis": "First, our brains have a cognitive subsystem, a 'module,' that monitors our behavioral plans and sounds an emotional alarm bell when we contemplate harming other people.  Second, this alarm system is 'myopic,' because it is blind to side effects. . . . These limitations make us emotionally blind--but not cognitively blind--to certain kinds of harm" (224).

Our evolutionary ancestors had an enormous capacity for premeditated violence.  Hobbes was right that in the state of nature human beings were roughly equal in that any healthy adult could kill anyone else, even the strongest.  And Hobbes was also right that anyone who was violent would provoke retaliation from his victims or their relatives.  Therefore, there was a natural need for all to seek peace so that they could enjoy the benefits of living together in cooperative groups.  We can assume, Greene argues, that this would have created evolutionary pressures for the evolution of an emotional alarm system in the brain to inhibit premeditated acts of aggressive violence, which would be adaptive for both individuals and groups.

This emotional alarm system would require that we have the ability to mentally inspect our premeditated plans of action, Greene reasons, so that if we see our actions directly causing the killing of someone, this would trigger our emotional alarm bells.  But perhaps this system is so simple that it cannot keep track of multiple causal chains, so that it sees the primary causal chain but not any branching causal chain.  In the case of the Footbridge Case, we think about the personal force in pushing the fat man onto the tracks in front of the trolley, which sets off our emotional alarm bells.  But in the Switch Case, we think about diverting the trolley away from the five people on the tracks as our primary end, and we are rather blind emotionally to the secondary effect of killing the man on the side track, although we are cognitively aware of this.

If this is right, then the utilitarian is someone whose neural circuitry does not attach more emotional weight to the killing of someone as the primary effect of an action than to the killing of someone as the secondary effect of the action.  That's why the utilitarian, with the rule that five lives are worth more than one life, will both pull the switch and push the fat man to save five lives.

Greene's general conclusion about all this is that "our moral intuitions are generally sensible, but not infallible," and whenever those moral intuitions deviate from utilitarian reasoning, the intuitions are mistaken, and they must be overridden by deliberate reasoning that suppresses the moral emotions.

So would Greene push the fat man?  No!  Here's what he says:
"Now, you may be wondering--people often do--whether I'm really saying that it's right to push the man off the footbridge.  Here's what I'm saying: If you don't feel that it's wrong to push the man off the footbridge, there's something wrong with you.  I, too, feel that it's wrong, and I doubt that I could actually bring myself to push, and I'm glad that I'm like this.  What's more, in the real world, not pushing would almost certainly be the right decision.  But if someone with the best of intentions were to muster the will to push the man off the footbridge, knowing for sure that it would save five lives, and knowing for sure that there was no better alternative, I would approve of this action, although I might be suspicious of the person who chose to perform it" (251).
This is complicated.  First, Greene admits that it's not clear that pushing people off footbridges to save others is really going to promote the greater good in the long run, because such behavior is likely to have bad consequences.  Second, Greene warns that the sort of people who could push people off footbridges are suspicious characters.  And, finally, Greene assures us that he is not one of those suspicious characters, because he feels that it is wrong to push people off footbridges, and he is proud that he has such virtuous emotions.

In fact, Greene admits that he is a hypocrite, in that he professes to be a pure utilitarian, but he cannot live completely by utilitarian rules because he's a human being with human feelings (254-68).  So, for example, he knows that in buying expensive birthday presents for his children rather than spending more of his money to help starving children around the world, he is failing to maximize happiness impartially, because he is partial to his children.  But, after all, this is only human nature.

One might be reminded of Peter Singer's confession some years ago that he was spending too much money on the care of his mother with Alzheimer's disease.  By his utilitarian principles, this was immoral, because his mother was going to die soon, and if he had sent this money to Oxfam, it would have saved many lives and thus maximized global happiness.  Like Greene, he is a hypocrite.

Greene says that if he were God, he would replace the human species--or Homo justlikeus--with Homo utilitus: "The members of this species value the happiness of all members equally.  This species is as happy as it could possibly be, because its members care about one another as much as they care about themselves.  This species is infused with a spirit of universal love.  That is, the members of this species love one another with the same passionate intensity that members of Homo justlikeus love their family members and close friends.  Consequently, they're a very happy lot" (267).

So pure utilitarianism is literally inhuman, because it contradicts our evolved human nature.  Purely utilitarian animals would have to belong to a different kind of species.

But then one might wonder whether even Homo utilitus would be truly utilitarian.  After all, the "spirit of universal love" in Homo utilitus does not seem to be truly universal, because it's restricted to members of that species.  Isn't this what Singer calls "speciesism"?  Doesn't utilitarian impartiality require universal love for all sentient beings--all beings capable of the experience of happiness--and not just members of our own species?

Doesn't this show how pure utilitarianism becomes incoherent?  In Greene's case, this incoherence comes from his religious longing for transcendence that conflicts with his scientific naturalism.  He repeatedly makes a "case for transcendence" (15-16, 25-27, 55, 63, 102, 200, 204, 264-67, 344-46, 349, 353), which culminates in his vision of himself as the Divine Creator who would replace the human species with Homo utilitus so that human beings could show universal love.  This is Gnosticism--the idea that we were created by an evil God who designed us so that we would be imprisoned in bodies with evil desires, and that our salvation requires a transcendence of our human nature through an enlightenment that transports us to a new realm of perfect universal love.

1 comment:

Kent Guida said...

We should all be grateful to Singer and Greene for making it clear why utilitarianism is a dead end.

How long will it be before utilitarianism ceases to be the default moral theory of the social sciences? Another generation or two?