On the Appeal to Intuitions in Ethics
Excerpted from Dale Jamieson (ed.), Singer and His Critics, Oxford, 1999, pp. 316-318

Even though it has always seemed to me so evidently erroneous, the view that we must test our normative theories against our intuitions has continued to have many adherents [...]. But now it faces its most serious challenge yet, in the form of Peter Unger's Living High and Letting Die. On one level this book is an attempt to tighten the argument I advanced in 'Famine, affluence and morality'. Unger argues that we do wrong when we fail to send money to overseas aid organizations that will use it to save many lives. But he does much more than that. He makes his argument by presenting a wide variety of examples and telling us about the intuitive responses that he had found most people - especially his students - have to them. The responses are very difficult to reconcile with each other. Unger then offers explanations for them. His explanations are devastating for the view that we should take our intuitive responses to particular cases as the test of a sound theory, because the explanations show that our intuitive judgments are based on things that are obviously of no moral significance at all. Here is an example. Unger uses some variants on the 'trolley problem', much discussed by philosophers during the past thirty years. The problem is posed by a runaway trolley rolling down the railway track, on course to kill several innocent people further down the line. In one version of the problem you can throw a switch that will divert the trolley down another track, where it will kill just one innocent person. In another version, there is no switch, but you could push a very heavy person off a bridge in front of the trolley. The heavy person will be killed, but the trolley will be stopped and the six people will be saved. Most people think that you should throw the switch, thus causing one to die, rather than six; but they think it would be wrong to push the heavy person off the bridge into the path of the trolley. To a consequentialist this difference is puzzling. In both achieve this outcome? A Kantian, however, can claim that the responses show that our intuitions are in line with the Kantian idea that it is wrong to use someone as a means, even if by doing so there is a net saving of innocent human life. According to the reflective equilibrium model of moral philosophy, this ability to offer an underlying theory that can account for our responses to such cases is the mark of a successful normative theory.

In most versions of the trolley problem, the agent has only two options: to pull the switch or not pull the switch, to push the heavy person off the bridge or not to push. In one the agent is active, changing what would happen if he or she were not there, while in the other option, the agent does nothing. Unger introduces intermediate options, and shows that this affects the way people judge the extreme options. In other words, when presented with a choice between A and E (where A, for example, is doing nothing, and E is pushing the heavy person into the path of the trolley) people will say that E is the worse option. When presented with a choice between A, B, C, D and E (where B, C, D and E progressively save more lives by increasingly active forms of intervention) people will say that E ii the best option. The reason for this surprising result is that people see that B is better than A, C is better than B, D is better than C and E is better than D.

Why should adding or deleting intermediate options affect our intuitive judgments of pre-existing options? A defender of our intuitions might argue that Unger's intermediate options are a means of corrupting sound moral intuitions, but we would need to know why that should be so When we look more closely at the options that people are inclined to reject, the picture looks quite different. The intuitive reactions are, Unger argues, based on factors much odder than not using a person as a means:

First, when serious loss will result, it's harder to justify moving a person to, or into, an object than it is to move the object to, or into, the person. Second, when serious loss will result, it's harder to justify changing the speed of a moving object, or changing its rate of motion, than changing the object's direction of motion. Third, when there'll be big loss, it's harder to justify speeding up an object than slowing down an object. Fourth, it's a lot harder to justify taking an object at rest and setting it in motion than to justify taking an object in motion and increasing its speed... [Fifth] it's harder to justify imposing a substantial force on an object than it is to justify allowing a force already present (just about) everywhere, like gravitation, to work on the object. (Peter Unger, Living High and Letting Die, p. 102)

It's easy to agree with Unger's characterization of these ideas as 'silly'; but after reading Unger's exposition, it's not so easy to deny that they play a role in many people's intuitive reactions to the trolley problem and it variants. Clearly, if Unger is right, the method of doing moral philosophy that relies on our intuitive judgments of particular cases is in tatters These factors just cannot be morally significant. If our intuitions really are based on them, then our intuitions are systematically unreliable Perhaps these intuitions have been developed in situations where the factors Unger mentions frequently are pointers to something else of genuine moral significance, but in other situations we intuitively follow the pointers when they are no longer appropriate.

Admittedly, informal surveys of one's students do not prove anything.Unger is now exploring with a psychologist ways of more systematically testing his hypotheses, and I hope this will lead to more reliable data. Nevertheless, Unger has put the onus of disproving his hypotheses on to those who take intuitive moral judgments as the touchstone of a normative moral view.

Utilitarian Philosophers :: Peter Singer :: 'On the Appeal to Intuitions in Ethics'