My essay on the non-aggression principle has generated a fair bit of discussion on the interwebs. Some of that discussion suggests to me that the title of the piece threw folks off a bit, or that my central thesis wasn’t as clear as I might have hoped.
So, let me say right off the bat that I don’t advocate “rejecting” norms against aggression per se. What I object to is the NAP as the sole and absolute criterion of libertarian justice. As I tried to make clear in the original essay, I think that a very strong presumption against coercion makes perfect sense. But to hold that that presumption is indefeasible, and the sole relevant criterion of justice, as Rothbard seemed to hold, strikes me as deeply implausible. And it was that absolutism and monism to which I was objecting in my essay.
So, I actually agree with most of what Jason Kuznicki has to say in his piece at Libertarianism.org, though I might disagree with just how overly-simplified a moral model ought to be to usefully serve its function. The point of my Copernican metaphor wasn’t to write non-aggression out of the moral universe. It was to argue that the moral universe doesn’t revolve around non-aggression.
On the other hand, I found David Gordon’s comments over at Circle Bastiat a little disappointing. David is one of my favorite libertarian philosophers, and I usually find his critiques to be incisive and thoughtful. But his arguments here just seem mostly to miss the mark.
To begin, he disagrees that Rothbard’s NAP implies that it would be wrong to trespass on someone’s property to feed a three-year old child whom someone was starving to death.
That is nonsense. To starve someone who cannot leave is to murder him. You don’t have to touch somebody to kill him: there isn’t a special libertarian concept of murder, different from the ordinary one. Neither is it the case that you are free to violate people’s rights, so long as you do so on your property. Rothbardian libertarianism is not the doctrine that each person is an absolute despot over his own property.
But when I wrote about children in my original essay, I wasn’t really drawing any implications from NAP that Rothbard didn’t draw himself. Rothbard’s own words on the subject were that “the parent should have the legal right not to feed the child, i.e., to allow it to die. The law, therefore, may not properly compel the parent to feed the child or to keep it alive.”
Now, why can’t the law compel the parent to feed the child? Presumably because doing so would be a violation of the parent’s rights. And I assume that if it is a wrong for the law to violate the parent’s rights, it is also wrong for private individuals to do so. Hence it would be wrong for people to trespass on the parent’s property to feed the child.
The parent, of course, can’t prevent the child from leaving. That would be to violate the child’s negative rights. But suppose it is a 8 month old who can’t leave the house of his own accord? In this case, I do not see how Gordon can say that the parent is violating the child’s rights, consistent with the Rothbardian view of property rights and the NAP. There’s a difference between killing and allowing to die, and if the parent is doing only the latter, then Rothbard seems to be committed to saying (and does say!) that the child’s rights are not violated.
With respect to risk and pollution, Gordon appeals to the role of convention in determining harm:
[Rothbard] recognized that setting the limits of harm is matter of convention, settled by the understanding that prevails in a society. Zwolinski here falls into a mistake that many libertarians make. They deny a role to convention in delimiting the boundaries for the application of a concept: unless “nature” settles the matter, use of a concept is an all-or-nothing affair.
Now, I think that to a certain extent, reliance on convention to determine the boundaries and contours of rights makes good sense. But where’s the limit here? If, as Gordon says, “setting the limits of harm is matter of convention, settled by the understanding that prevails in a society,” then would slavery not count as a harm in a society that did not regard it as such? Presumably Gordon would not want to draw this conclusion. But then why should your emitting noxious fumes onto my property and into my lungs not be a violation of rights, just because society has adopted the convention that it isn’t really harmful?
Finally, Gordon says that a prohibition on fraud is perfectly consistent with libertarianism’s fundamental emphasis on property rights:
[Zwolinski] s also correct that “aggression” in the principle must be understood to cover violations of property rights, as well as direct physical assault. Certainly Rothbard understood the NAP this way. But why, immediately before pointing this out, does he claim that a prohibition of fraud isn’t compatible with the NAP, because fraud is not physical violence? Is it too much to expect Zwolinski to realize that his point about the meaning of aggression invalidates his own objection regarding fraud?
This is a fair point. But in my view, Rothbard is inconsistent on violence/property rights. Sometimes he defines aggression as violence, period. And if that’s what aggression means, then fraud isn’t aggression and isn’t prohibited by NAP. Sometimes, however, he defines aggression as a violation of property rights. That’s probably a better way to go. But notice how, understood this way, libertarianism loses some of its intuitive appeal. Now the cornerstone of libertarianism is no longer an opposition to aggression – it’s the enforcement of property rights! Now maybe that’s a philosophically defensible view. I think there’s a lot to be said for it. But I worry about the misleadingness of saying that we’re opposed to one thing, when in fact what we’re really opposed to is something else altogether.