The non‐aggression principle isn’t sufficient to help guide most of our political decisions, and so isn’t sufficient to be the core argument for libertarianism.
There has been much buzz lately in libertarian Internet circles concerning the Non‐Aggression Principle (NAP), a favorite ethical maxim adopted by many libertarians in defense of anarcho‐capitalism. This increased attention to the NAP is due in large part to Matt Zwolinski’s excellent article articulating many of its counterintuitive implications. Zwolinski recommends that instead of radically editing the principle as stated – that is, instead of adding epicycles to our original understanding of the orbital paths of celestial bodies – libertarians ought to undergo a “Copernican Revolution” by developing a new ethical system to justify our preferred socio‐political order. Like any good controversy, there has been much response, most notably by Jason Kuznicki in his equally‐excellent essay “Non‐Aggression and Billiards.” In this essay I examine Kuznicki’s defense of the NAP, not necessarily as an absolute and inviolable axiom, but at the very least as a strong though rebuttable presumption.
We are all taught that certain things are true when in actuality they are false. We are taught, for instance, that there are three inviolable laws of motion governing all moving bodies in our universe, when in reality things work a bit differently when we start dealing with very small particles. We are also fed exaltations concerning the immutable axioms of Euclidean geometry when such an axiomatic system remains contingent upon a series of assumptions that may or may not hold true. Even so, these kind‐of‐true‐but‐strictly‐speaking‐false principles of physics and mathematics are very useful to us: they can allow us to understand how billiard balls interact with one another, or perhaps they help us find out how much manure we need to help fertilize a 16 foot by 24 foot field. The NAP, so Kuznicki argues, is of similar value: it gives us a good starting point to help determine when actions are morally right and morally wrong, though some refinement, or “engineering,” might be required for more advanced situations.
But many times Newtonian mechanics and Euclidean geometry are insufficient. Surely the rocket scientist and quantum cryptologist need to go beyond high school mathematics and physics classes in order to do their jobs properly. This, of course, does not change the fact that most people need only an elementary understanding of physics and mathematics to get through their day‐to‐day lives. But in comparing the NAP to something like Newtonian mechanics and Euclidean geometry, we need to see if the analogy carries over to this corollary: do practical agents, trying to navigate through life, require only the NAP to help guide most of their decisions, or do they need a more robust version of the NAP – or perhaps a new moral system altogether – in order to navigate life’s imbroglios? As practical agents, are we high school students or rocket scientists?
I contend that we are rocket scientists. We are rockets scientists not because life is that hard a thing to navigate through (though it certainly has its harsh contentions), but rather because the high school math many libertarians want us to adopt – the NAP – is so insufficient as a moral theory that it cannot stand up to the basic experiences we all, as practical agents, face. For example, most of us, it seems reasonable to suggest, will end up having and raising kids. Will the NAP (Newtonian physics) tell us how to raise them right? As Zwolinski points out, it will not only not tell us how to raise our kids correctly, it will, in actuality, approbate actions that any sane person would think constitutes morally impermissible methods of child rearing. Thus, quantum mechanics is required. Many of us also frequently engage in behavior that yields some sort of environmental pollution. Will the NAP (Euclidean geometry) gives us the right answer as to what we ought to do? Again, no. It will tell us that we cannot, as Zwolinski once again points out, drive an automobile or light a campfire, conclusions that are surely absurd. Thus, non‐Euclidean geometry is required. Several people are confronted every day with the chance to commit some kind of fraud to make a quick buck or reap some benefit. Will the NAP tell these individuals how they ought to act? As Zwolinski shows, it will not. Our high school physics and mathematics has failed; we need a more sophisticated theory.
Since the adoption of the NAP results in rocket scientists trying to make their way with Newtonian mechanics, we need to beef up the NAP, rendering it capable of handling the sorts of problems we as rocket‐scientist‐like practical agents face. In editing the NAP, there seems to me to be two paths we can take. First, we can radically alter or add to our definition of “aggression,” in hopes of reproaching the sorts of actions our original NAP failed to delimit as off‐limits. Kuznicki offers an attempt at doing this in hopes of rendering fraud impermissible; namely, he advocates we classify as morally impermissible actions that are “impediments to reflective self‐rule” to go along with our ban on brute physical aggression. This may work, but it is certainly not enough. What do we add to physical aggression and impediments to self‐rule if we are to outlaw a parent starving their child? What do we add to allow for reasonable amounts but not too much pollution? Moreover, we also run the risk of our amendments yielding results we find unacceptable. Would my neighbor playing loud heavy metal music several hours throughout the day impede my ability to reflectively self‐rule? Maybe, if it sufficiently hindered my concentration. Yet most libertarians would find it acceptable to do something on one’s property as benign as playing loud music. So now we need to further edit what is meant by “reflective self‐rule” to account for this result. Epicycles upon epicycles.
The second way we can strengthen the NAP proceeds as follows: we can keep our original understanding of what constitutes aggression, but then add a list of exemptions we consider to be legitimate instances of when the NAP may be transgressed. We then get something like this: it is morally obligatory to follow the dictates of the NAP except in cases x1, x2, x3,…, xn, in which case it is morally permissible (or perhaps obligatory) to disobey the dictates of the NAP. The problem, though, remains that variable n can be a high – indeed very high – number, to the point at which listing all exemptions becomes difficult, if not impossible. In fact, there is probably a case to be made that, given the tremendously narrow scope of the NAP and the ingenious ability of philosophers to dream up counterfactuals, the list of exemptions might be un‐codifiable, presenting a problem for this approach to amending the NAP. Instead, it might be argued that rather than listing each and every exemption to the principle, we might simply develop a secondary rule determining when some situation constitutes a legitimate instance of when the NAP may be permissibly violated. But this project also seems rather hopeless, for any rule that concretely adjudicates exemptions from non‐exemptions will have to contain excruciating detail – perhaps an un‐enumerable amount of detail (we could not, for example, simply say that situation s constitutes an instance permitting exemption from the dictates of the NAP just in case situation s yields a severe reduction in an agent’s autonomy, because it is not clear what constitutes a “severe reduction in an agent’s autonomy”). Once again, epicycles upon epicycles.
It is not the case that every ethical system yielding implausible implications through counterfactual analysis must be abandoned for another ethical system. It might be that an ethical theory is polished to the point that it can be revised to such a miniscule extent that it remains worth holding on to – after all, one or two epicycles does not grossly detract from the elegance of an otherwise perfectly elliptical orbital path. But there comes a point when the amount of revision borders on the absurd – after we have drawn the fifteenth, sixteenth, seventeenth epicycle. At this point, it makes sense to start anew in hopes of developing a better theory that does not require such extensive revision, allowing for maintained elegance. I think the NAP is in this latter category. As such, it is high time we, as libertarians, do as Zwolinski bids by undergoing our own Copernican Revolution in search of a better moral and political philosophy.