Doing Harm

Doing Harm (pg. 70-84)

(From Normative Ethics)

By Shelly Kagan

In the assigned section of “Doing Harm,” Shelly Kagan promotes the idea of ‘moderate deontology.’ This theory suggests that there can be constraints on actions, such as a constraint on doing harm, which can override the fact that an action may produce the greatest good. This would mean that it is forbidden to do harm to someone, even if it would result in more good overall. However, she notes that there may be a threshold at which this constraint could be infringed upon. This helps to put consequentialism more in line with commonsense morality – i.e. it is not considered ok to kill one innocent person in order to save two others, however, it may be considered right to kill one innocent person to save a million others. However, when one allows for a threshold on the constraint, there will be difficulties in defining the level of that threshold. How many much good must be at stake to make it ok to harm an innocent person? This will also depend on the type (or size) of harm – if punching an innocent person will save a life, this would likely be seen as moral. There is also the consideration of the risk of harm – what level of risk is acceptable.

3.1 Deontology

Consequentialists (such as utilitarians) believe that there is one and only one factor with intrinsic moral significance in determining the status of an act: the goodness of that act’s consequences. However, others (and often commonsense) may suggest that there is more than one factor, or that there may be constraints on action. This is illuminated by a scenario.

Imagine there are five patients, each of whom will soon die unless they receive an appropriate transplanted organ: one needs a heart, two need kidneys, one needs a liver, and the fifth needs new lungs. Unfortunately, due to tissue incompatibilities, none of the five can act as donor for the others. But here is Chuck, who is in the hospital for some fairly routine tests. The hospital computer reveals that his tissue is completely compatible with the five patients. You are a surgeon, and it now occurs to you that you could chop up Chuck and use his organs to save the five others. What should you do? [This is the equivalent to the scenario in which a Sheriff has to frame an innocent person to prevent riots in which multiple innocent people would be killed.]

In this case, the utilitarian would likely admit that it would be better to kill Chuck, since you’d be saving five lives instead of one. However, many people feel intuitively that it is morally forbidden to chop up an innocent person. They might suggest that the moral status of an act depends not only on the goodness of its results, but also on whether or not it involves doing harm to someone, and that the importance of not harming someone outweighs the good results. Thus, commonsense morality recognizes a constraint against doing harm, even if the act is necessary to bring about the best results overall. Kagan defines a deontological theory as any moral theory that does incorporate constraints which erect barriers to the promotion of the good. This type of theory can work in concert with consequentialism, and the constraints do not need to be absolute.

It is possible that consequentialists might try to argue that their theory would not result in choosing to kill Chuck. For example, they may argue that organ transplants are risky, and the result could be six dead people instead of five. Or, if people find out about Chuck being killed, they may become afraid to visit hospitals, and this could be a much worse result overall. However, if the story is amended to ensure success of the transplants and complete secrecy, the person would likely agree that killing Chuck would be the correct choice. A consequentialist might argue, however, that in these extreme cases (or in any moral decision), we can’t just trust our intuitions, which is why a rule like utilitarianism is useful.

3.2 Thresholds

It may be true that at some point, there is a threshold on the ‘do not harm’ constraint. It makes sense that the significance of doing harm gives this factor sufficient weight to override goodness of results in cases like the organ transplant. However, what if killing an innocent person would save a hundred, thousand, or a million lives? Imagine that killing the one is the only way to prevent a nuclear war; or that the innocent person’s body must be used to produce a serum that will provide a cure for a plague that will otherwise kill millions. Some absolutists might not violate the constraint, even in these cases. However, most people believe that the constraint against doing harm can be outweighed if enough is at stake.

This theory is called moderate deontology, and it argues that up to the threshold point, it is forbidden to harm an innocent person, even if greater good could be achieved by doing it; but if enough is at stake (the threshold has been reached or passed), then the constraint is no longer in force and it is permissible to harm the person. However, if this position is taken, then the threshold would need to be identified. For consequentialists, this threshold is zero (not doing harm never outweighs the overall good of the consequences), while for absolute deontologists, it is infinity (not doing harm always outweighs the good of the consequences), so moderate deontologists must be between these extremes. Though commonsense morality suggests the threshold is fairly high, any particular level seems arbitrary.

In defining a threshold, a moderate deontologist would also need to take into account the size of the harm, kind of harm, and risk of harm. The size of harm matters since lower stakes may be required if the harm is smaller – e.g. it may be permissible to punch an innocent person to save one life. The kind of harm – whether to a person’s property or body – could also make a difference in the threshold. Finally, one would like to allow the threshold for the constraint against harm to vary with risk, as it seems to do in reality. Driving a car poses some risk to others, but is considered acceptable. However, acts with very high risk of harming others may be seen as immoral. Again, these aspects of the threshold will be difficult to define.