Please note that this is BBC copyright and may not be reproduced or copied for any other purpose. RADIO 4 CURRENT AFFAIRS ANALYSIS Thought Experiments TRANSCRIPT OF A RECORDED DOCUMENTARY Presenter: Janet Radcliffe Richards Producer: David Edmonds Editor: Innes Bowen BBC White City 201 Wood Lane London W12 7TS 020 8752 7279 Broadcast Date: 28.06.09 2130-2100 Repeat Date: 29.06.09 2030-2200 CD Number: Duration: Taking part in order of appearance: Professor Anthony Appiah, Princeton University Joshua Greene, Harvard University Professor Jonathan Haidt University of Virginia Joshua Knobe University of North California Dr Helena Cronin London School of Economics Dr Nick Shea University of Oxford RADCLIFF RICHARDS: If you hear stories of some people who have stopped to help others in trouble, and other people who have passed by on the other side, what do you think makes the difference between them? Is it that some of them have been taught good moral principles and learned to apply them, while the others haven’t? Is it just that some of them are nicer by nature than the others? Well you might think so; but some recent experiments raise interesting questions about the differences between the ones who help and the ones who don’t. APPIAH: The classic experiment is one that was done with theology students, divinity students, who had been thinking about the problem of the Good Samaritan, and they were actually preparing a sermon on the topic. RADCLIFF RICHARDS: That’s Professor Anthony Appiah of Princeton University. APPIAH: And some of them were told that they were late to give the sermon and others weren’t. Then they were sent off across the yard to the place where they were supposed to give a sermon. And on the way they had staged a good-Samaritan like possibility. There was a person sitting in the doorway looking a bit dishevelled and unhappy. And the thing that predicted whether these people would stop or not, was simply whether they had been told that they were late or not. And it’s rather surprising because you would have thought that somebody who had been thinking about the good Samaritan problem would naturally realize that it’s probably more important to act like a good Samaritan than to be on time for a sermon about a good Samaritan. RADCLIFF RICHARDS: Well I suppose we might be charitable and recognize a moral conflict here. It would be good to stop and help, there is presumably a church full of people waiting to hear the sermon, and they are entitled to consideration as well. But other experiments suggest that a most extraordinary range of circumstances can affect our responses to people in need. Here is another of them carried out in a phone booth in an American shopping mall. APPIAH: They put a dime, a 10 cent coin, in the return slot of the phone booth for some people and not for others. And each time somebody came out of the phone booth somebody walked by and dropped a pile of papers. And the question was; what was the likelihood that the person coming out of the phone-booth would help the person who dropped the papers? And the answer was that finding that dime made it many, many times more likely that you would be helpful. RADCLIFF RICHARDS: You wouldn’t think finding a dime would affect your moral character. But even more surprising things seem to make a difference. APPIAH: They’ve done a similar experiment where whether somebody is going to give you change for a dollar is affected by whether they’re standing outside a bakery where the air smells good, or outside what they call a neutral smelling dry goods store, where the air doesn’t smell so good. And it turns out again that people are many, many times more likely to be helpful in terms of giving you change for a dollar if they’re somewhere where the air smells good to them. RADCLIFF RICHARDS: Professor Anthony Appiah is one of the advocates of a new movement of academics who claim that ethical questions philosophers have traditionally tried to answer from their armchairs can’t be dealt with that way. They call their subject experimental philosophy. I personally think it is more useful to think of what they are doing as an interdisciplinary collaboration between philosophy and the social sciences, rather than an extension of philosophy. But the relevant question is whether and how these investigators can contribute to ethics. Appiah suggests one important conclusion: APPIAH: I think the experiments like the one with the phone booth suggest that one traditional preoccupation of ethics which has been with character, perhaps needs to be somewhat modified because we have to bear in mind that actually what determines what people do is at least as much their circumstances as their character, perhaps surprisingly more their circumstances than their character. And what that suggests as a matter for moral life, I think, is that it’s terrifically important to shape the world in a way that gives people the context in which they’ll behave well, rather than shape a world around them in which they’ll do the things that we don’t want them to do. So I think that’s one big conclusion. RADCLIFF RICHARDS: Now of course no one is suggesting that we can stop knife crime or child cruelty by putting bakeries on every street, but the question of principle is important. Is this how we should try to improve society? Finding out how to manipulate people’s feelings and behaviour without their realizing, and adjusting the environment accordingly? It sounds like what T S Eliot described as ‘dreaming of systems so perfect that no one will need to be good’; and this appals a lot of people. Surely a good society must result from the deliberate choices of good individuals, not the manipulations of some benevolent puppet master? And this is where the fundamental problems of ethics arise. Before trying to create a better world we need criteria for deciding what counts as ‘better’. Traditionally such answers were given by priests; more recently by philosophers; and now the question arises whether the new ethical experimentalists can help. So, consider this puzzle about a runaway train – very familiar to philosophers – explained by another of the philosophically trained experimentalists, Joshua Greene, of Harvard: GREENE: So there’s a trolley that’s speeding out of control down a set of tracks and it’s headed towards five people, and if nothing is done those people will be run over and killed. RADCLIFF RICHARDS: This is by the way a trolley in the American sense – of a tram or a train – not our sense of a tea trolley or a supermarket trolley. GREENE: But you are standing at a switch where you can turn that trolley onto a sidetrack where it will only run over one person instead of five. The question: is it morally acceptable, morally permissible to hit the switch, so that you turn the trolley away from the five and onto the one and minimize the number of deaths? And if you ask people, most people say that that’s morally acceptable. RADCLIFF RICHARDS: So should you switch the train and kill the one person on the branch, to save the five on the track? These people are all strangers, to whom you have no special connection – so you are considering their lives as mattering equally, impartially. And that seems to be why most people think it’s better to save the five even though the cost is killing the one. But the thought experiment goes on... GREENE: Now you change the case a little bit - or a lot, depending on your point of view. This time the trolley’s headed towards five people and you’re on a footbridge standing over the track in between the oncoming train and the five and you’re standing next to a large individual. And the only way that you can stop the trolley from running over the five people this time is by pushing this large individual off of the footbridge, so that he ends on the track and is viewed essentially as a trolley stopper. He’ll die, but the five people will be saved. And of course you can’t jump yourself and use yourself as a trolley stopper because you’re not big enough to stop the train but this other person is. So the question here is again is it morally acceptable to perform this action that would save five lives at the cost of one? And there most people say that that’s not morally acceptable. RADCLIFF RICHARDS: So it seems wrong to most people to stop the runaway train, or trolley, by pushing the fat man off the bridge. But a lot of philosophers have argued that this response just doesn’t make moral sense. If it’s all right to divert the train, at the cost of killing the single person on the branch line rather than the five on the main line, how can it really be wrong to achieve the same end, at just the same cost in innocent life, by pushing the fat man off the bridge? And although this thought experiment may seem a rather pointless philosopher’s game, in fact it has real implications for much more realistic moral problems. Here’s another thought experiment along similar lines. GREENE: Decades ago the philosopher Peter Singer posed a question, for people in the affluent world. He said, “If you were walking by a pond and there was a child who was drowning in it and you could wade in and save that child’s life, but in doing this you would ruin the fancy shoes - would it be okay to let the child drown because you’re worried about your shoes?” And of course everybody says no. And then Singer’s question was: Why is this different from allowing people, let’s say children, across the ocean in Africa who desperately need food or medicine - why is that any different, letting that child drown, from letting those children starve or go without the medicine they need if we have the financial resources to prevent that from happening? RADCLIFF RICHARDS: If you will willingly sacrifice the cost of a pair of shoes to save the child in front of you, how can you justify not doing do the same for another child on the other side of the world? GREENE: And the response on the part of a lot of ordinary people, on the part of a lot of philosophers, was to think well there’s got to be some morally relevant difference between these two cases, and a lot of plausible attempts were made to say what that difference is and on closer inspection it’s actually a pretty hard thing to do. Both these children are presumed to be strangers; rationally, we can see that they matter equally. But our response to the two is radically different. And such strong differences of feeling inform our attitudes and even our laws about life and death in other areas as well. RADCLIFF RICHARDS: In medical ethics an important issues is physician assisted suicide. The position of a lot of people and I believe the American medical association is that it’s acceptable for a doctor to withhold life-saving treatment from a patient who would like to die, a terminally ill patient, but it’s not acceptable for a doctor to do something actively to help this person bring about his or her own death. Most people are absolutely convinced that there must be a morally important distinction, which is, in most countries entrenched in law, between killing patients and failing to save their lives. Patients who want to die have an absolute right to refuse life saving treatment, but a doctor who kills them at their request is guilty of murder. People feel strongly that there is a difference, but it’s notoriously difficult to find one that looks morally relevant. Philosophers have been arguing about these matters for a long time, but now the new experimentalists are approaching the problems from a quite different direction. One of these is the moral psychologist, Jonathan Haidt, of the University of Virginia. His research suggests that in practice we very rarely start from moral principles and then reason to conclusions about what people ought to do. Rather, we have the moral feelings first – we feel that something is right or wrong, that someone is to be praised or blamed – and then bring in the arguments afterwards, to justify them. HAIIDT: We have these quick gut feelings and then we engage in reasoning afterwards. But the reasoning isn’t really to find the truth; the reasoning is to find a justification. It’s really easy to see this in everyone else’s reasoning, especially when you listen to politicians. But it’s very difficult to see this in our own. We each believe that our reasoning actually causes our judgement, causes us to decide a certain way. RADCLIFF RICHARDS: Jonathan Haidt has done experiments designed to demonstrate the problems that often arise – unnoticed – in peoples’ attempted justifications for their moral beliefs. HAIDT: What I did in my early research is create stories in which people would instantly feel that it was wrong, but yet there was no victim. So about a brother and sister who are adults and who have consensual incest - consensual sexual relationships with two forms of protection; or about a family whose dog was killed by a car in front of their house and so they cut up the body and ate it for dinner. In these cases, there’s an instant yuck response and people would often say, “Well it’s wrong because well you know if you have sex with your sibling, there’ll be birth defects; and if you eat your dog, you’ll get sick”. And then I have to say, “Oh well, sure, you know if that were the case, I’d see why that’s wrong, but you know in this case you know they use two forms of birth control and they cook the meat fully and so how can anyone be harmed?” And what I found is that people never say, “Oh yeah, you’re right, I forgot about birth control. Oh then it’s okay”. Rather they say, “Oh gosh, yeah. Hmn, okay wait, let me think about this. I know it’s wrong, but I’m just having trouble thinking of why”. RADCLIFF RICHARDS: Jonathan Haidt refers to this as their being dumbfounded. They have run into a contradiction between two sets of moral intuitions – a contradiction between the immediate Yuck response – incest is yuck, eating your pet is yuck - and the idea that you can’t be doing anything wrong if you aren’t harming anyone or anything. The two beliefs are incompatible, but people want to hold on to both of them. And until someone forces them to confront the conflict, they don’t even notice it. Yet another of these experimenters, Joshua Knobe - has come to even further reaching conclusions about our moral psychology. He thinks our moral judgments actually influence our supposedly straightforward descriptions of what is going on. One of his studies involved people’s responses to stories about a hypothetical company chairman. KNOBE: Suppose that a vice president of a company goes to the chairman of a board and says ok we’ve got this new policy, it’s going to make huge amounts of money for our company but it’s also going to harm the environment, and the chairman of the board says, look, I know that this policy’s going to harm the environment but I don’t care about that. All I care about is making as much money as we possibly can. So they go ahead and implement the policy and sure enough it harms the environment. And the question is did the chairman of the board harm the environment intentionally? RADCLIFF RICHARDS: Knobe found that most subjects said he did do the harm intentionally. But then he changed the example slightly. Everything was exactly the same except that the new policy had a good effect on the environment. The chairman still didn’t care about anything but the profits. And when people were asked whether he caused this good effect intentionally, they said he didn’t. KNOBE: Yet it seems as though the only difference between the two cases is that in the one case he is doing something morally wrong, and in the other case he is doing something morally right. So it somehow seems as though peoples’ moral judgements are impacting their intuitions as to whether something was done intentionally or unintentionally. So Joshua Knobe thinks that our moral judgments actually influence our supposedly non-moral descriptions of what is going on. This is now widely referred to as the Knobe Effect. I think these experiments point to a really different view of human nature – of the basic question of how human beings understand their world. A common view is that human beings were like scientists; they were trying to develop roughly a scientific picture of the world, a picture that allows them to predict and explain things in the world. But these studies seem to be pointing to a very different picture – a picture of the human being as moralist; someone whose whole understanding of the world is suffused with moral consideration, who views the whole world through a moral lens. Jonathan Haidt agrees. We make our moral judgments quickly, and then present our descriptions of what is going on in ways that seem to support that judgment. HAIDT: If you’ve just heard a story about a CEO who harms the environment, you dislike him, you want to convict him. Now somebody asks you, “Did he cause the outcome?” Well of course he did because I want to convict him; and given that causality’s ambiguous, I’m going to pick the reading of causality that allows me to convict him. We make the evidence fit the crime. RADCLIFF RICHARDS: It does indeed sound like the police planting evidence on criminals when they are sure they are guilty anyway. This appears to be the kind of thing we all do, without noticing, most of the time. In the meantime Joshua Greene – whom we earlier heard describing the philosophical controversies – has moved up a technological step. He used a magnetic resonance scanner to investigate what was going in people’s brains when they thought about the trolley problem. GREENE: The first experiment that I did was essentially an attempt to look inside people’s brains while they’re thinking about these cases, so we had people respond to cases like these while they’re inside a brain scanner. A scanner detects which parts of the brain are being used more than normal. What Greene did was analyse the brain as his subjects thought about the trolley problem: as they thought about whether they should save five lives first by switching the trolley to a spur where it would kill one person, and second, whether they should shove the fat man off the footbridge. RADCLIFF RICHARDS: Greene looked in particular at parts of the brain involved with emotion and parts involved with reasoning. GREENE: The hypothesis that we were testing is that there is something more emotional about the case where you’re pushing the guy off of a footbridge and that something about the action in that case sets off an emotional alarm bell in your brain that makes you say no, that’s wrong. And if you don’t have that emotional alarm bell, then you’re going to default to a more cost-benefit analysis of the situation - what in philosophy we call a utilitarian or a consequentialist approach. That is thinking about right and wrong in terms of the consequences: better to save five lives than one life. RADCLIFF RICHARDS: And they found just what they were expecting. GREENE: And the initial neuro-imaging images that we got were at least you know consistent with that idea. We saw brain regions associated with emotion, among other things, exhibiting increased activity when people thought about cases like the footbridge case and, likewise, other brain regions that are more associated with reasoning, what we call cognitive control, exhibiting increased activity when people think about the cases like the switch case. RADCLIFF RICHARDS: So when we are faced with the prospect of killing a person standing next to us, the emotional part of the brain appears to become very active. When we are thinking of more distant people, like the one on the branch line, we can be more detached in our calculations. Now it must be emphasized that all this is work in progress. The research is new and in many respects controversial. But suppose it’s on the right track - Does it help us to decide whether we should follow our strong moral intuitions? It’s often been taken for granted that we should. Many religions regard that the moral part of our nature is divinely implanted, and hold that our conscience is the best guide to what’s right. But our experimentalists are starting from a secular world view. They regard us, body and mind, as wholly natural creatures whose emotions can in principle be explained in the same way as any other natural phenomenon. For instance, Greene speculates about the evolutionary cause for the emotions that make us recoil from pushing the fat man off the bridge: GREENE: There had to be some kind of brake on basic interpersonal violence if our species was to survive as a social species; and so my guess is that we have some kind of an alarm system that says don’t do that when you are contemplating doing something that’s harmful in a very basic physical kind of way. And my guess is that the footbridge test where you’re thinking about pushing a person off of a bridge, that action sets off that alarm bell; and that the case where you’re hitting a switch to redirect the trolley, that doesn’t set off that alarm bell. Now thinking about this from a scientific and an evolutionary point of view, it makes sense that we would have this response; that we evolved in an environment in which you could save people who were right in front of you in trouble - drowning or hanging from a cliff or being attacked by an animal; we didn’t evolve in an environment in which you could save people on the other side of the world who are total strangers, who you’d never meet by making a modest material sacrifice. So perhaps it’s not surprising that … Perhaps it’s not surprising that the up close and personal drowning pushes our moral and emotional buttons in a way that the far away starving child doesn’t. RADCLIFF RICHARDS: Or take the deeply felt difference between acts and omissions, killing and letting die; leading to the opposition of euthanasia: GREENE: Why do a lot of people have this judgement or this intuition? It could be because there really is this important moral distinction between causing somebody who wants to die, helping them die, or allowing them to die. But it also could be that we just think about actions and omissions differently for cognitive information processing reasons that have nothing to do with morality. For example right now I’m talking into a microphone, but there are a lot of other things that I’m not doing. I’m not grilling sword fish steaks, I’m not painting a picture of a zebra, there are infinitely many things that I’m not doing right now. And it would be cognitively impossible for me to represent all of the things that I’m not doing. And that’s true of actions in general. It makes sense that our minds are designed to represent actions differently than they’re designed to represent non-actions. But it’s not because there’s a morally important difference between bringing out a result through action and inaction; it’s just the way we think about it. We think about actions differently from non actions. RADCLIFF RICHARDS: Evolution, in other words, provides us with all kinds of moral heuristics – rules of thumb – that have been useful for our survival as a social species, but which fall very far short of providing coherent moral principles. Here’s Helena Cronin, who runs the Darwin programme at the London School of Economics. CRONIN: There’s really no reason at all to expect our moral and ethical outlook to be coherent and entirely consistent. We are involved to have moral heuristics that will work to oil the wheels of our social life, the social life of small bands of people of a couple of hundred at most. In order to oil those wheels, natural selection had to deal with lots and lots of different situations and it set up different moral stances, different moral heuristics for each of those situations. There’s absolutely no reason to assume that the moral heuristics that go for example with the disapproval of incest will in any way fit with the moral heuristics that go with the morality of sharing your meat. And they might indeed under some circumstances come into conflict. RADCLIFF RICHARDS: It’s often pointed out that evolution is a botch: full of elements that would never have been planned by a competent designer. If our moral emotions are just rules of thumb that happen to have made us successful as a species, how can they be regarded as guides to what’s morally right? They’re not even guides to continued evolutionary success - Helena Cronin again: CRONIN: Natural selection designed morality for very specific circumstances. Take for example the incest taboo. We understand the reason why it designed that, but the way it did it was to give us a horror of people who were brought up together having sex together; whether it’s ourselves or the way we regard others…. now just think in the modern world, how that trigger sometimes gets applied utterly inappropriately. For example, the most notorious and well known example is on the kibbutz, where cohorts of children who are brought up together, not related at all, but they never ever marry one other. There are no intermarriages on kibbutzim, which from an evolutionary point of view is nonsense, because these might be very suitable mates. Indeed, even from a social point of view I’m sure the parents would have been awfully pleased had they wanted to marry. All in all, our moral emotions can’t be regarded as reliable guides to anything that matters. RADCLIFF RICHARDS: So how should we think about our resistance to pushing the stranger off the footbridge to save five people? Or our strong feeling that doctors be allowed to let people die, if they want to die, but cannot intentionally speed up the dying process even if that’s what the patients want? Should we try to get rid of these emotions that reason tells us cause unnecessary harm? Well, if we did it would eliminate from life many of the things we most care about: all our special connections with family, friends, all our personal interactions with people. And that doesn’t sound very attractive. We’d certainly be an unrecognizably different species – not a social species at all. But anyway, even if we wanted to change human nature in radical ways, we haven’t the faintest idea how to do it yet. Instead of trying to drag our emotions into line with our reason, what we should be doing is using our reason to understand, and making the best of, what we are. For instance, perhaps some of our emotions can be redirected to support our moral judgments. My colleague Nick Shea, from the Oxford Uehiro Centre for Practical Ethics, believes we can sometimes train our deeply ingrained capacity to feel disgust. SHEA: So an example is moral vegetarianism. There’s been a little bit of research done on this that suggests that people who for moral reasons decide to become vegetarians then acquire a disgust response in relation to meat and that makes it in a way much easier for them to stick with their vegetarianism. And it also has a moral effect via contagion of that emotion, so they can effectively persuade other people to be vegetarians - like their children - by people seeing the disgust response in them, sympathetically experiencing disgust themselves, and thereby starting to find meat disgusting themselves. RADCLIFF RICHARDS: We might decide that there are all kinds of things we wanted people to feel strong disgust for that they may not feel at the moment. Income tax evaders, perhaps? But even if we can’t change or redirect emotions, we can still change rules and policies that our reason tells us are unjustified. For instance, if our emotions don’t lead us to treat the third world as well as reason dictates we should, we can institute policies to achieve those ends on our behalf. If we can’t get rid of our feelings of disgust, we can still refuse to allow them to inform our law. Even people disgusted by homosexuality, can recognize that this does not justify any denial of rights. If we can’t overcome our intuitive horror of active killing we can still remove the laws that treat voluntary euthanasia as murder and if we find that we can make ourselves nicer with the smell of baking bread – putting bakeries everywhere would not turn us into puppets because we would have made the decision that this is the kind of people that we wanted to be. Evolution began the process of ethics by giving us the capacity to sympathise with our relatives and friends but it was our reason, going beyond our evolutionary advantage, that made us realize that people mattered even if we did not personally care about them. Devising laws and environments to achieve what reason tells us makes the world better is to fulfil our moral capacities, for this, the scientific project of trying to understand the human animal is essential.