For better or for worse (I think for worse, but that's another post), People for the Ethical Treatment of Animals (PETA) is the face of the animal-liberation movement. Another group that makes headlines is the Animal Liberation Front (ALF). Animal liberationists such as me are often dismayed—and sometimes outraged—by the tactics used (encouraged, condoned, tolerated) by these organizations. One such tactic is violence (understood as intentional or reckless damage to person or property). Is violence ever justified as a means to protecting animals?
As is so often the case, the answer is, "It depends." It depends, specifically, on one's theoretical outlook. Let me mark out the ground. Act-consequentialists evaluate actions on a case-by-case basis. That act is right (they say) that maximizes overall good, taking all interests into account and counting all individuals (including the agent, or person performing the action) equally. That an act is of a certain type, such as "lie," "broken promise," "killing the innocent," or "torture," is irrelevant. That is to say, there is no presumption either for or against the doing of any particular type of act. Only the consequences of concrete (token) actions matter.
Deontologists deny this thesis. Deontologists say that the type of act one performs has (or can have) moral significance. Certain acts are wrong in themselves, independently of their consequences. Immanuel Kant (1724-1804), for example, held that lying, qua lying, is wrong. But it's not always noticed that there are two types of deontologist. Some are absolutists. They say that no amount of good consequences brought about or bad consequences prevented can justify committing one of the prohibited acts. Kant was an absolutist deontologist. He held that one must not lie, period. Lying is always and everywhere wrong, whatever the circumstances and whatever the consequences. (Do you see why the adjective "absolutist" is appropriate?)
Other deontologists are nonabsolutists. Let's call them "moderates," since they do not lie at one of the extremes of the theoretical spectrum. Moderate deontologists are deontologists, because they think that certain act-types are wrong in themselves, but they believe that acts of those types may be performed if enough good will be produced (or bad prevented) as a result.
Think in terms of thresholds. Absolutist deontologists have an infinitely high threshold. Nothing can reach it. Act-consequentialists have no threshold. There is no presumption against any type of act. Moderate deontologists have a finite threshold. It gets complicated here, because moderate deontologists can have different thresholds. One moderate deontologist may have a low threshold that allows lying if and only if it produces X units of good (or prevents X units of bad). Another may have a higher threshold, one that allows lying if and only if it produces X+10 units of good (or prevents X+10 units of bad). If you want to think numerically, the absolutist deontologist is at 100 (on a scale of 0 to 100); the act-consequentialist is at 0; and moderate deontologists fill the gap between 1 and 99. If someone tells me that he or she is a deontologist, I don't know much. He or she could be a weak moderate deontologist, and hence very close to the act-consequentialist, or an absolutist deontologist—or anything in between.
In case you're wondering whether anyone these days is a deontologist, the answer is yes. Emphatically yes. All of the following contemporary moral philosophers (some of them, alas, recently deceased) are deontologists in my sense (either absolutist or moderate): John Rawls, Thomas Nagel, Robert Nozick, Ronald Dworkin, Charles Fried, G. E. M. Anscombe, Bernard Williams, Alan Donagan, and John Finnis. These are among the most highly regarded of contemporary philosophers. Nor should one think that deontology is incompatible with atheism. Nagel is an atheist. Anscombe and Finnis are theists. Deontology has nothing to do with belief in a supernatural being.
Let us return to the question whether violence is ever a morally permissible tactic. The act-consequentialist will not rule it out. If an act of violence is the act, of all those available to the agent, that maximizes overall good, then it is the right thing to do. The end justifies the means. Everyone has heard this slogan. Now you know who subscribes to it and lives by it: act-consequentialists. Think about what the slogan means. It means that no path (means) to the end of maximizing overall good is ruled out. If torture is necessary to maximize overall good, then torture is justified. If lying is necessary to maximize overall good, then lying is justified. No means are forbidden. That an act is of a particular type (a lie, a broken promise, a killing of the innocent, a case of torture) is of no moment, morally speaking.
The absolutist deontologist rules out violence categorically. No amount of good can justify it. Evil may not be done that good may come. Moderate deontologists endorse a presumption against violence, but they differ, as we saw, in how strong the presumption is. Those who are close to absolutist deontologists will require that the violent act produce a great deal of good, or prevent a great deal of bad, in order to be justified. If the only way to save 10,000 innocent people is by blowing up a building that contains one innocent person, it might be justified, but not, say, if only ten people will be saved by killing the one. A moderate deontologist with a weak presumption (i.e., a low threshold) might allow the killing of one to save ten. An act-consequentialist would allow the killing of one to save two, or even the killing of one to save one, if the one saved has a longer expected lifespan (or a greater capacity for happiness) than the one killed. See the difference?
What this shows is that what one thinks about the moral permissibility of violence depends on one's theoretical outlook. I would be remiss if I ended the post here, because I have not done justice to act-consequentialism. I said that the act-consequentialist believes that the end justifies the means. But even act-consequentialists can adopt rules of thumb. Experience shows that acts of violence do less good than they are expected or hoped to do by those who perpetrate them. They also do more bad. And most people abhor violence, so any individual or group who is interested in social change would do well to use it only as a last resort, to avoid alienating or antagonizing those who are sympathetic to the group's ends. There are almost always other, more effective means to the end of social change.
Peter Singer, the famed (but in some quarters reviled) author of Animal Liberation (New York: Avon Books, 1975), which has been called the bible of the animal-liberation movement, believes that violence should be a last resort. He won't rule it out categorically, because he's an act-consequentialist, but he thinks that it is rarely (if ever) justified. Here is his discussion of violence in the second edition of Animal Liberation (New York: The New York Review of Books, 1990):
It would be a tragic mistake if even a small section of the Animal Liberation movement were to attempt to achieve its objectives by hurting people. Some believe that people who make animals suffer deserve to have suffering inflicted upon them. I don't believe in vengeance; but even if I did, it would be a damaging distraction from our task of stopping the suffering. To do that, we must change the minds of reasonable people in our society. We may be convinced that a person who is abusing animals is entirely callous and insensitive; but we lower ourselves to that level if we physically harm or threaten physical harm to that person. Violence can only breed more violence—a cliché, but one that can be seen to be tragically true in half a dozen conflicts around the world. The strength of the case for Animal Liberation is its ethical commitment; we occupy the moral high ground and to abandon it is to play into the hands of those who oppose us. (pages xii-xiii)
Singer is a wise man.
Note that in a particular case, each of the moral theorists I have described can condemn violence. Suppose someone kills the proprietor of a chicken farm in order to get publicity for the cause of animal liberation. The act-consequentialist would condemn it for the reasons Singer gives: It is counterproductive; it is not the act, of all those available to the agent, that maximizes overall good. The absolutist deontologist would condemn it because it's a case of killing the innocent (a species of violence), and no amount of good consequences produced or bad consequences prevented can justify killing the innocent. (Don't protest that the chicken farmer isn't innocent. He may not be to you, but he is to the absolutist deontologist, and we're talking right now about how the absolutist deontologist reasons about violence.) The moderate deontologist would condemn it because the amount of good consequences produced or bad consequences prevented are insufficient to justify it. This shows that theories can converge, or overlap. That they sometimes converge, however, doesn't make them the same theory. All it takes is one case of divergence to show this, and there are, take my word for it, many cases of divergence.