Can Violence Be Irrelevant?

by Sherry F. Colb

In my Verdict column for this week, I discuss an issue that arises in a book, co-authored by me and Michael C. Dorf, entitled Beating Hearts:  Abortion and Animal Rights.  The issue is how the pro-life and pro-animal rights movements ought to handle the violence dilemma, that is, the question when--if ever--violence is morally permissible in defense of one's cause.  My column concludes that violence is impermissible both for pro-life and animal-rights activists, because there are peaceful alternative means of rescuing fetuses or animals and of educating people and effectively communicating the message that fetuses/animals are entitled to be free of human violence, without having to resort to violent means.  In this post, I want to discuss an objection to this line of argument that emerged when Professor Dorf and I were presenting our work at a colloquium held at Queen's University in Kingston, Ontario.

Just to make the issue as clean and uncomplicated as possible, we asked our audience to consider two groups of people (or animals or fetuses) in need of rescue.  Both groups contain the same number of potential victims (10), and we know nothing about any one victim that would distinguish him or her from the rest (in terms of worthiness or other qualities that might break a tie in a triage situation). There is, however, one difference between the two hypothetical rescues that we would seek to accomplish.  In one case, rescuing the ten people would require us to kill a malevolent guard who is blocking access to the people.  In the other case, by contrast, rescuing the ten people would require no killing at all, perhaps because their malevolent captor was sufficiently confident in having hidden his victims well that he did not see any need for guarding them.  Assume that you have time for only one rescue of ten people, so you cannot save all twenty.  What should you do?

We believed, going into the colloquium--and in writing the chapter about violence--that the answer is straightforward.  When faced with this choice between killing a guard and rescuing ten innocents versus killing no one and rescuing ten innocents, the right thing to do is to rescue the ten innocents whose rescue requires no violence at all.  Quite a few audience members at Queen's University disagreed, however, and accused us of using a consequentialist metric to determine that it would be better to save ten without killing anyone than to save a different ten while killing someone.  On the assumption that it would ordinarily be justified, in isolation, to kill a guard who is holding ten innocents captive, our interlocutors suggested that it cannot suddenly become wrong to perform that same killing just because there is now another group of ten needing rescue, and this other group's rescue requires no killing.  In other words, either it is right to kill the guard to save the ten people he guards or it is wrong to do so, but its rightness or wrongness cannot turn, these audience members claimed, on whether some other group of people needs to be rescued as well, with different contingencies in effect.

Our view, however, is that the universe of options that one has available will always affect the morality of choosing to act in one way rather than in another.  We further questioned some of our audience, asking whether they would say the same thing if we believed that a group of ten people could be rescued only by killing their guard but then we learned information that enabled us to rescue those ten people without killing anyone.  Would you still say, we asked, that killing the guard would remain justified, in the absence of necessity (given the nonviolent alternative means of rescue)?  No, they replied.

If the very same group of 10 people could be rescued peacefully or violently, then the violent rescue is unnecessary and accordingly an immoral act.  It is immoral, as we understand it, because even though we would be killing the person in order to rescue 10 people--and the one person guarding them is culpable and therefore arguably entitled to less concern than the innocent 10--the availability to us of a peaceful option means that killing the guard is an unnecessary act of violence, and committing an unnecessary act of violence is wrong (at least on these facts, where we are not killing the guard as a means of retributive capital punishment).

Our audience members continued to reject our actual hypothetical example, however, claiming that there is all the difference in the world between our original example (with two groups of 10, one of whom may be rescued only with violence, the other peaceably) and the  example with the one group of 10 (who may be rescued either violently or peaceably).  Why the difference?  Because in the original example, the 10 people in the first group (guarded) are not interchangeable with the 10 people in the second group (unguarded), each of the people is as entitled to rescue as the others, and the choice to rescue one group of 10 rather than the other is accordingly arbitrary and should probably be done through a coin flip.  To say it is better to save the 10 whom we can save without violence, then--according to some of our audience members--is to treat people, or animals, or fetuses, as though they are just containers of utility that may permissibly be exchanged for one another while other orthogonal factors (such as the need for violence as part of the rescue) are attended to.

After talking about this issue back and forth for several hours, with professors and students alike, we came to the conclusion that perhaps a purely deontological view has no place for taking into account the difference between killing a guard and not killing a guard, if the underlying act of rescue is the right thing to do.  Perhaps this is in part because ideally, two rescuers would come along, one with a weapon and one without, and the two would successfully rescue all 20 of the people in need of rescue in the least intrusive manner possible (one group requiring violence, the other not).  But of course, in our actual hypothetical scenario, triage is unavoidable, and only 10 people can be rescued. And as people who consider ourselves at least deontologist hybrids, we regard as troubling the view that morality cannot prefer the nonviolent rescue option.  After all, the death of the guard is not merely a "consequence" of the violent rescue; it is the means by which we effectuate a rescue, when we could be effectuating a different rescue without deploying violence at all.

Trolley problems are legion in moral philosophy, so I shall invent one of my own to try to capture the moral dilemma we have been considering.  You are hurtling down a train track and you find that there are 10 people tied to the tracks whom you will run over if you do not flip the switch.  You can flip the switch to the right or to the left.  If you flip it to the right, you will hit the person who tied up the 10 people and you will kill him; if you flip it to the left, you will hit no one.  If you flip nothing, you will crash and kill everyone on your train as well as all of the people tied to the tracks.  The correct thing to do seems clear here: you flip the switch to the left, save everyone, and kill no one.

Now you find yourself in charge of two trains hurtling down two similar train tracks.  On one set of tracks, you face the 10 innocents tied up, the one evildoer who tied up the 10, and the clear track.  On another set of tracks, you face another 10 innocents tied up, the one evildoer who tied up the 10, but no other track.  You get to choose which train to control in this scenario, though, so you can save 10 but must also abandon 10 either way.  We believe that the right thing to do is to choose to engage with the train that has the option of going down a clear path and killing no one, even though this means that you abandon the other 10 innocents.  Remember that you necessarily will abandon 10, and the only question is whether you will also commit violence.  Under a "first do no violence" ethos, I would propose that the right choice is the choice that requires no violence of you at all, the one that has no "moral regret" attached (which attaches whenever hurting someone is necessary to accomplishing even a laudable mission).  If that position is counter-deontological, then -- to quote my co-author -- so much the worse for deontology (which ought, in my view, to be assessing the cost of moral regret in choosing between different options).  We're both very curious to see people's moral intuitions in response to these scenarios (including the one with which all of this started:  the choice between advocating for innocent victims--whether animals or fetuses--violently or nonviolently, when one can choose either path).