By Samuel C. Wheeler III, Department of Philosophy, University of Connecticut
This article originally appeared in Public Affairs Quarterly, Volume 11, Number 4 (October 1997) pages 431-443. It is published on the Independence Institute website with the permission of the author and of Public Affairs Quarterly. The endnotes in this article are not linked. If you want to read an endnote, just scroll to the end of the article.
The notion of an absolute right to self-defense runs into some serious problems with the logical structure of absolute rights. For almost any right, one is usually able to construct a sequence of cases such that there is no ethically significant difference between successive elements of the sequence, but the first element is clearly a case where the right exists and the last case is one where ascription of the right is absurd.1 Since it seems to be part of the concept of rights that two ethically indistinguishable cases must be treated alike, such results are paradoxical.
This paper proposes another way to think about rights and applies it to the arguments for a right to self-defense with deadly force. In particular, it will show that a popular sorites argument against such rights is unsound.
The paper argues that there are some clear cases where a person has a right to use deadly force in self-defense. Such cases are independent of the guilt or innocence of the assailant, are not punishment, and establish a prima facie liberty to prepare oneself for such exigencies. The existence of this right, however, depends on its not being over-ridden by conflicting rights. Whether such over-riding takes place is a matter of particular factors in the situation, not a matter of principles in hierarchy.
Given that such rights only exist when they do not conflict with countervailing rights, however, the sorites arguments which lead to the alleged right to own bazookas, 20mm cannons, and heat- seeking missiles do not succeed, given current circumstances.
Let me begin by examining some well-known features of rights which seem to distinguish rights from other kinds of considerations which we use in determining what we ought to do:
Rights can be over-ridden by considerations of utility, in the sense that there are situations where one ought to take actions which violate rights. But rights persevere when over-ridden by utility. The over-riding by utility leaves a situation where a right has been violated and compensation is owed.
Consider the following modification of an example made famous by Posner:2 I, a rare-book dealer, find myself stranded in a snow-storm on a tree-less pass. There is a concrete block cabin right by the road, with the door ajar. Inside is a stove and a pile of wood, with a sign "Property of Fred. Do not use." In my car are $450,000 worth of books from the late 1400's, many of them with genuine wood boards, and all of them folio-sized thick tomes. None of them are really rare, existing in many other copies around the world. As the temperature drops to twenty and then thirty below, I have a decision to make: Do I violate Fred's property rights by using his wood to survive, or do I observe his property rights and suffer a substantial loss by burning some or all of my books? (Rescue is assured within a day, and either my books or the wood will suffice to save me.) In this situation, there is no conflict of rights, but clearly only a conflict of utility and rights. Equally clearly, it seems to me, this is a situation where one ought to violate Fred's property rights, steal his wood, and take the consequences. Depending on Fred's character, those consequences could range from compensation to a short jail term.
A very different situation arises when rights encounter over-riding rights. Fred's property rights persist even when we ought to violate them, as is indicated by the fact that we owe him compensation no matter how reasonable our violation of his rights was. On the other hand, consider the following common situation: As a licensed driver, I have a right to drive down North Eagleville Road at the speed limit, namely 30 mph. However, Jones, a pedestrian, has a right of way in the crosswalk. In this case, when Jones is in the cross-walk, his right to be there erases my right to drive down the road at 30 mph. I have to slow down or stop. Furthermore, Jones does not owe me compensation.
Property-rights are erased by other rights when my use of property imposes risks on others. I cannot, for instance, burn my wood lot if that imposes risks on others. I cannot operate a nuclear reactor on my property if the existence of that poses enough of a threat to my neighbors. In such cases, I don't have a right which persists in the face of the countervailing right.
My general hypothesis about rights is that rights are only erased by conflicting rights. This means that, for any kind of right where there can be conflicting rights, rights are hypothetical on there being no over-riding conflicting counter-rights. Whether the right survives the countervailing right depends on the relative strength of the rights.
In the special cases I want to examine, the rights are rights not to be harmed by another's actions. The relative strength of rights in such cases amounts to a weighing of the risks imposed by the respective rights. Such risks are a multiplicative function of the degree of harm threatened and the objective likelihood of the harm.
I first argue that there are some clear cases where a person has a right to use deadly force to thwart an assault. I describe such a clear cut case, then make several points about this case:
The case is an actual occurrence, one of innumerable parking-garage abductions which have the following pattern: A woman is approaching her car in a deserted parking garage when she is approached and over-powered by a larger and stronger male assailant. The assailant takes the car keys, locks the woman in the trunk of the car, turns the radio up full-blast, exits the garage, drives to a deserted spot, and, more often than not, eliminates the only witness to a rape by murdering the woman. In the case at hand, however, the woman happened to be carrying a small .38 handgun, and when the abductor opened the trunk, he took four slugs in the torso.
It is difficult not to feel that this is an outcome preferable to likely alternatives. If the reader is not quite persuaded, perhaps we can distinguish the actual case, where the woman was prepared for such eventualities, from the possible case where she discovered a loaded gun in the trunk. Surely in that case, the woman has the right to shoot. I discuss possible objection to her right in these circumstances below.
There are several points which I need to review about this scenario:
1) The primary effect of the handgun is to render insignificant differences in physical ability, strength, and agility. In a world where men are on the average significantly larger and more powerful than women, a woman with a gun is almost as dangerous as a man with a gun. If there is any conditional right to self-defense, there is equally a conditional right to use appropriate technology.3
2) Shooting someone in self-defense is not punishment, as Locke points out.4 The woman is not legislating that attempted rape and murder is a capital crime. Self-defense prevents an unjust deadly harm. If there is, by chance, a police officer in the trunk of the woman's car who arrests the potential abductor, the potential victim does not then have the right to shoot. That it is not punishment, and therefore not a violation of any social contract, can be seen by reflecting on the following scenarios.
It could turn out that the abductor was having a psychotic reaction to Prozac, and would normally have never done any such act. Such an assailant is, to my mind, innocent. But the right to self-defense, since it is not punishment, persists even in situations where the assailant is in fact innocent.
It is important that, given even the minimal concession that the woman has a right to defend herself with a found hand-gun, there can be no "social contract" to leave defense to the commonwealth. That is, unless one is willing to say that it is the woman's duty to hope that the police will show up at the deserted spot in a timely manner, one has to limit the social contract to a surrender of rights to determine what the law is and administer justice. So, the woman's action cannot sensibly be described as "taking the law into her own hands." Self-defense is not in general a case of enforcement of laws against rape and murder. The woman may in fact be against capital punishment for such crimes as are about to be committed upon her and still be resolved to shoot to kill.5
3) Self-defense is a right even when there is a more optimal solution available. Let me discuss two rather different cases:
a) In another innocent-assailant situation, the woman's evil stalker has in fact hypnotized the assailant. The code-word for breaking the hypnotic trance is "hocus pocus." So, there was a much more optimal solution to the woman's difficulty, namely saying "hocus pocus" when the trunk was raised in the deserted spot.
In fact, during the ride, the woman might speculate that there is some possible world in which precisely that scenario has occurred. If this is that possible world, she reasons, then she does not have to take the chance of killing this man. Now, this would strike most of us as a very risky strategy, although not one with a zero chance of success.
What action ought to be taken in a situation is generally epistemological, not ontological. The fact that there is a better action available and within her power does not make that action either reasonable or obligatory.
b) Less remote chances, although just barely so, would be the possibility of appealing to the person's better nature, or shooting to hit the knees. If we demand such actions, we impose risks on the woman. In my opinion all of these risks are super-rogatory. Perhaps a saint would take them, but they are not actions we can reasonably compel.
It is debatable whether we have a right even to impose smaller risks; for instance by requiring that civilian women in such situations warn their assailant that they are armed and demand their surrender. The very real possibility that the young spry man would just slam the trunk lid back down, or that he has a gun of his own, make even such actions far beyond the call of duty.
4) Self-defense need not weigh utilities in seeking its justification. If there are three assailants, and she is lucky and skillful, the woman can legitimately shoot them all. Since there is a substantial chance that three lives are lost for the sake of one, if we think that it is permissible for her to shoot all three, then we cannot think that self-defense rests just on utilities. The idea that there is an obligation to submit to superior numbers rather than destroy more lives than would be saved by submission seems bizarre. To my intuitions, even the possibility that she has been abducted by a gang of a hundred does not diminish her right to shoot her way out of the situation if she can.
In the actual case, the woman was prepared for the eventuality of being assaulted. Being prepared for such attacks imposed risks on others. What are the risks that this preparedness imposes, and thus rights that this preparedness infringes? There are several sources for such countervailing rights.
1) first, every handgun can be used for offense as well as defense. Everyone else is threatened to the degree to which the woman herself might decide to assault someone for her own entertainment. More probably, there is some chance that an armed woman will shoot non-assailants who happen to be lurking for entirely other reasons, or who are taking actions which could be interpreted as the preliminaries to an assault, such as running towards her.
The empirical data suggest that in such cases, the mere display of a gun, rather than firing a gun, will clarify the situation. Certainly accidents will happen if a substantial portion of women are carrying guns. One would expect, however, that the knowledge that women tend to be armed would reduce behavior which could be interpreted as part of an assault.6 It will become more dangerous to pretend to stalk, or to amuse oneself by frightening women.
2) The existence of more handguns among the general outdoor population imposes some risks on the whole population. If someone gets very angry at me, the chances that I will be injured increase dramatically if the person has a gun. Do such risks outweigh the risks imposed on the woman by preventing her from protecting herself effectively?
Much depends on details. Consider automobile operation, and in particular auto operation by unlicensed drivers who have only seen driving on television. If many people on the road were driving for the first time, with no training whatsoever, the roads would be a dangerous place indeed. In many of the unfortunate accidents with guns, the person is firing a gun for the first time. Furthermore, the elementary precautions one learns to take in a pistol class, such as not having a round in the chamber under the firing pin, which prevents accidental discharge, are not taken. It is certainly unwarranted speculation to predict that accidents and ill-judged uses of firearms would increase because of an increase in properly-licensed concealed handgun-carrying.
We can only conjecture what the effect of treating gun-licensing the way we treat automobile-licensing would be: That is, what would happen if a person had to take a course to get a pistol permit, but the issuance of such a permit was a prima facie entitlement, so that the government would have to have a special reason not to issue a permit rather than the citizen being required to have a special reason to have a permit.7
One countervailing benefit often overlooked is the wider effect on unarmed women. If a third of women are packing concealed S&W .38 Specials, those who choose not to arm themselves are beneficiaries, as long as the weapons are small enough to be concealed. Many of them are. Concealed weapons, far from being especially dangerous, have positive free-rider effects.8
There is of course the position that the best solution is to have no citizens with guns whatsoever; thereby imposing, admittedly special risks on the usually weaker and smaller.9 Such prohibition might be justified, however, by the lighter risks the majority would be under. So suppose we decide that the risks in having armed non-criminals outweigh the risks a prohibition would impose on women, the small, and the elderly. Then we have to consider the effects of prohibition:
Given the realities of the population of handguns already in existence, it is reasonable to conjecture that the most likely effect of prohibition of handguns is at best the same as the effect of prohibiting recreational drugs. Those who use recreational drugs are those who have relatively little to lose by the penalties. A $100,000 fine is a great punishment for me, but a trivial and amusing punishment for a street person. Likewise, a year jail term loses me a tenured full-Professorship; while the typical drug-user suffers no such loss.10
Anything less than a serious diminution of civil liberties, involving laws which allow house to house-searches without warrants and stop-and-frisk operations without reasonable cause, would lead to a situation where only criminals and those with little to lose by violating laws would be armed. It is worth noting that a necessary condition for famous twentieth century genocides has been the disarming of the targeted population. The very stringent gun laws of Stalin and Hitler clearly prevented the sort of resistance that would have made the destruction of so many citizens expensive. The house-to-house search, as a way of making sure that gun-laws were obeyed, was a tactic common to Hitler and Pol Pot.11
The mechanics of guns and drugs are also very different. A single gun and a box of shells will last an armed robber an entire career. An armed robber very seldom has to shoot, since the threat of a gun, especially against known-to-be-unarmed victims, is so clear.12 Shells last indefinitely, as do guns. Drugs, on the other hand, are consumed, and need to be replaced almost every day. In brief, if there were a War of Guns of an expense and intensity equal to the War on Drugs, we can expect even less good results: The criminals will have guns and the law abiding will have risks imposed on them or else have their freedom of movement restricted.
Now, the debate on what a society would be like in which a substantial portion of the population carried concealed firearms could go on for a while and has. People have opinions about the moral effects, the unreliability of the bourgeoisie, the general disposition to poorly-thought-out action of those who haven't received doctorates, and so forth. The empirical research, in fact, is at best divided;13 and to those who had hoped to find reasons to ban guns, disappointing.14
Let me make a few more general remarks before proceeding to consider a sorites argument against the right of ordinary citizens to arms themselves with S&W .38 Specials and their ilk:
First, it is by no means clear that the risks imposed on the general population by licensed concealed handgun carrying citizens outweighs the risks (not to mention costs) imposed by forcing people to be effectively defenseless against kinds of assaults that are not uncommon. The weighing of these risks is also very tricky. How do we compare the harm in being locked in the trunk of one's car, raped, and then killed to the risks of being shot in a misunderstanding, for instance?
Second, the important point about this section is that the alleged right to be prepared against certain kinds of assaults is a matter of objective risks of prohibition to the self-defender compared to the objective risks imposed on the rest of the society by having armed citizens. The calculation will depend on the actual likelihood of such assaults, how bad they are, the likely risks to unarmed citizens, and a conception of what kinds of skills those with licenses to carry concealed weapons would have.
Third, since we are talking about rights, it is not a matter of how we feel about living in a society which has armed citizens, unless the strong and nimble are willing to compensate the weak and frail for the risks imposed on them. I regard my colleagues dismay that one of them is so uncivilized as to own actual handguns as inadequate appreciation of the value of diversity, something akin to a distaste for having homosexuals in the department.
A typical argument against the private ownership and use of handguns is that, if it were legitimate to own handguns, it would be legitimate to own repeating sawed-off shotguns, howitzers, ground-to-air missiles, and eventually, tactical nuclear warheads. There are several things to say about this argument:
1) First, there are indeed situations in which private ownership of weapons of far greater power than S&W .38 Specials would be appropriate. Let us consider two such weapons, a semi-automatic 12-gauge shotgun with a clip holding 50 rounds, and the 20mm cannon, a military weapon shooting exploding shells.
There are some kinds of reasonable expected assaults where a high magazine capacity semi-automatic shotgun would be appropriate. If I am a Russian Jew in the 18th century, I can expect that, several times during my lifetime, I and my family will be attacked by angry mobs, with the tacit approval of the authorities. If I can reasonably expect to face pogroms, a high-volume shotgun might be an entirely appropriate self-defense weapon.
There are kinds of expected assault in some situations where an automatic 20mm cannon would be an appropriate self-defense weapon. One of the reasons the Dark Ages stayed so dark for so long was Scandinavian raiding, which destroyed population and agriculture through much of Europe. If I am a French farmer on the banks of the Seine, subject to recurrent attacks by Viking longboats, a 20mm cannon would be very appropriate for dealing with such assaults.
I'm not claiming that people in such situations clearly have rights to private ownership of such weapons. It could well happen that the possession of a 20mm cannon would turn a peaceful French peasant into an aggressor who dominated his neighbors. Likewise, Russian Jews armed with semi-auto shotguns might turn into an oppressive minority.
2) The second point to make about the sorites arguments is that the situations where assaults for which very powerful weapons are appropriate are not reasonably to be expected in contemporary America. While there is some non-zero probability that I will be assaulted by a force against which automatic 20mm cannon would be an optimal defense, that probability is very small. Similarly, my chance of needing ground-to-air missiles to protect against a helicopter assault is miniscule.
In any given cultural-social situation, there are some kinds of assaults that it is reasonable to prepare for. In those cases, forbidding a person from adequately preparing for such risks would be imposing a serious objective risk. It might turn out that those risks are counter-weighed by the risks being prepared imposes on others, but there is a prima facie right to acquire such resources. However, other kinds of assaults have a very low likelihood, and if being prepared for such assaults imposes any substantial risks, any prima facie right to be prepared for such assaults is outweighed by the rights of others not to be exposed to the risks your preparedness entails.
In our current situation, there are negligible chances of assaults where such weapons as a 20mm cannon or high magazine-capacity 12-gauge shotguns are appropriate. Thus, imposing vulnerability to such attacks on a person is objectively a small imposition, even though the harms from attacks by, for instance, large groups of men arriving at one's farm in vans are very severe. Furthermore, given the very high risk imposed on others by such weapons, both on occasions of use (i.e. from wild shots landing on one's neighbor's house), and as potential offensive weapons, there are very high risks imposed on everyone else. Analogous remarks apply to large-magazine semi-automatic 12-gauge shotguns.
Thus the sorites argument collapses almost immediately. If every right to prepare for risks must be evaluated by the severity of the feared assault and its likelihood, weighed against the risks imposed on others by such preparation, then rights to exotic armaments do not exist in contemporary situations.
Of course, we can imagine changes in our social structure which would make private ownership of such weapons reasonable. If government effectiveness collapses, and roving gangs of predators become the rule, the different risks generated by that situation might well yield a right to own more dangerous weapons. Another possible scenario, not unprecedented in our century, is a government which unjustly attacks its citizens. The kind of political discussion which would examine the possibility of laws, traditions and institutions which make it possible for citizens to overthrow an oppressive government is beyond the scope of this paper.15
3) How do we draw the line between assaults it is reasonable to prepare oneself for, even at the cost of imposing some risks on others, and ones which it is not reasonable to prepare for? The simple answer appeals to the complex but not-finally analyzable notion of the reasonable person. Much of ethical decision-making, and therefore social policy setting, depends on non-algorithmatizable judgment about what is or is not reasonable.
Such judgments are of course subject to political discussion and therefore change. Consider the changes that the concept of liability has gone through in the past fifty years. It is just mistaken to suppose that there was a complete concept of liability in use in 1948 which successive court cases have managed to unpack into the current set of precedents. Rather, things could have gone otherwise. The concept of liability, and therefore the extension of what one is actually liable for (and negligent about, and so on) is subject to social and political evolution.
That is, we could have treated what we now regard as a spectacular out, where a fielder leaps above the level of the fence and prevents a homer, the way we treat goal-tending in basketball. In the same way, we could have decided that a power-mover operator who uses his hands to clear clogged grass while the mower is running, in spite of clear instructions on the machine advising against this, has only his own foolishness to blame for his missing digits. The manufacturer is not liable for failing to make it practically impossible to keep the mower running while reaching under it. Such decisions are arbitrary, not in the sense of being random but in the sense of having been arbitrated, resolved by discussions whose outcomes depend on logic and power relations.16
Note finally that non-risk-producing preparations for unlikely assaults are, while perhaps irrational, not subject to interventions by others, at least for the kinds of considerations which would limit the right to bear arms. If I decide to equip my house with a self-contained air-purifying and regeneration system and make my dwelling airtight, which steps will keep me safe from nerve-gas attacks, there is no reason for prohibition or intervention by others. I'm nuts, but its my money.
The answer depends, as I have argued, on the details. If we imagine a situation such as Florida, where anyone with the proper training and a clean record can get a permit to carry a concealed firearm, we get one kind of answer. Based on Florida's experience, the answer seems to be "Yes." If we imagine a situation where, because concealed carry-permits are nearly unobtainable, most of those otherwise law-abiding citizens are carrying guns they do not know how and when to use, the answer might be "No," since such gun-carriers are dangerous to themselves and others. But a more plausible answer is that the laws imposing risks or movement restrictions on people by restricting their right to prepare for assaults are unjust. The high risks and accident-rates that unlicensed concealed carriers expose us to are the result of bad laws, not a natural incapacity of law-abiding normal citizens to handle firearms responsibly.
Endnotes to "Self-Defense: Rights and Coerced Risk-acceptance "
1 An example of such an argument for property rights is Samuel Wheeler's paper, "Natural Property Rights as Body Rights", Nous, Volume XIV, Number 2, pp. 171-194, May 1980.
2 Posner, Richard, The Economics of Justice, Cambridge MA: Harvard UP, 1981.
3 This might be an appropriate place to say something about handguns as a weapon of choice in such social situations, and their preferability to any other available self-defense device short of an armed bodyguard. Many people suggest such devices as pepper-spray, martial arts training, and extreme swiftness of foot for avoiding harms from felons.
If there were devices as universally feared by assailants as handguns, my argument would be much weaker. If the typical sixty-year old woman could, without enormous investment of time and energy, become sufficiently skilled in kung-fu to equalize her chances against an armed assailant, or if most of us post-50-year-olds could outrun the typical parking-garage assailant, things would indeed be different. But those factors do not obtain, and the small but effective handgun is the most equalizing equipment for self-defense that's practically concealable.
4 Locke, John, Second Treatise on Government, sections 19 and 20.
5 The situation where the victim is already dead, and one is deciding whether the State can appropriately execute the criminal is very different from the situation where the victim is alive. Reflect briefly on the fact that in the happy event that the police were on the scene, they would be entitled to shoot to kill, even though attempted murder is not a capital crime. It is amazing that so many otherwise clear-headed people are confused on this distinction.
6 One of the pitfalls in thinking about criminal behavior is to suppose that they are just nuts, and beyond the reach of rational decision-making. Much good criminological research, though, finds it productive to view criminals as rational agents choosing lines of action which reduce risks. An early and important example of such research is Anthony R. Harris' "Imprisonment and the Expected Value of Criminal Choice," American Sociological Review 40: 71-87, February 1975.
7 There is some pretty good data from Florida, where, of the hundreds of thousands of concealed-carry permits issued, none resulted in accidental or intentional homicides. But Florida may be a special case, since the warm weather makes people mild, and Floridians are not surly New Yorkers.
8 One of the unfortunate consequences of the Florida concealed-carry law was that criminals selected tourists, identifiable by their rental cars, as victims. This is no longer the case, because Florida rental car companies responded by removing any identifying marks, as anyone who has been to Florida recently and tried to find their unfamiliar car in a parking lot will attest.
9 The idea that the United States would be a safer place if there were no handguns because, say, Sweden is a safer place, ignores history and cultural differences. Life before guns included the Assyrians, Tamburlane, Viking raiders, etc. The reasonable expectation of deadly assaults does not seem to have increased across cultures generally.
The main difference that guns made was a democratization of the possibilities of defense and offense. A Hun on a horse has a great advantage over a farmer with a scythe. The Hun loses such advantage versus a gun. The longbow can be regarded as a similar democratizing device, though much it was harder to learn to use effectively than an AK-47.
10 I think it is always useful to think of criminals as rational agents, who have normal desires to survive and prosper, rather than as just nuts. On that hypothesis, criminal behavior can be best understood as rational responses to situations. Criminals by and large wish to operate without being themselves threatened. As the risks of their occupation change, so does their occupation. In fact, if the risks and rewards change enough, a career change might be in order. Mugging is not a long-term livelihood if I get shot at by every tenth intended victim. See Harris' work for example of the illumination that treating criminality as rational can provide.
11 For a detailed account of the connection between disarmament and genocide, see Simkin, Zelman, and Rice's Lethal Laws (Jews for the Preservation of Firearms Ownership: Milwaukee, 1994). Simkin, et al. go over virtually every recent genocide, from Uganda to Cambodia to Armenia. Their book was published too early to deal with Rwanda.
12 One result of surveys of convicted criminals show a marked and quite rational preference for robbing convenience stores owned by chains, where the employees are under strict instructions not to resist, as opposed to Mom and Pop stores where the possibility of a gun under the counter is real and threatening.
13 Don Kates' collection, Restricting Handguns, (Croton-on-Hudson, NY: North River Press, 1979.) is a good place to start.
14 Gary Kleck has a number of surveys and studies which deserve a wide audience. See his Point Blank: Guns and Violence in America, (Aldine Press: New York, 1991) and Kleck and Patterson, "The Impact of Gun Control and Gun Ownership Levels on City Violence Rates", Journal of Quantitative Criminology, 249-287 (1993), Kleck and Bordua, "The Factual Foundation for Certain Key Assumptions of Gun Control", in Law and Policy Quarterly, Vol 5, pp 271-281, and Kleck and Delone, "Victim Resistance and Offender Weapon Effects in Robbery", Journal of Quantitative Criminology, pp 55-81 (1993).
15 There is a whole complex of considerations about the possibility that a government might be oppressive, so that it would be reasonable to prepare for the full-range of attacks a government might launch. In that situation, we have moved from self-defense which might be permitted by a country's laws to preparations for civil war.
Some of the arguments that a democracy which wants to stay a democracy should allow its citizens to be prepared to overthrow it are presented in Simkin, et al., mentioned above.
It is perhaps worth noting that the success record of well-equipped armies against guerrillas equipped with only the kind of weapons which are now legal in the United States is not very good. One need only think of the French and US experience in Vietnam, the Russian army versus Afghanistan, and the Yugoslav resistance during World War II. It is precisely because a few armed citizens can make tyranny costly that disarmament (under various guises) has always been a prelude to oppression and genocide.
It is further worth noting, especially in response to arguments that "it couldn't happen here," that the United States itself has been a party to some of this century's destructions of populations, as for instance in Guatemala. In addition, of course, there is our own record of forced-relocation-as-starvation, most famously against the Cherokees, which is quite parallel to the Armenian genocide.
16 See Samuel Wheeler's "True Figures", in The Interpretive Turn, edited by Hiley, et al., Cornell UP, 1991, pp. 197-218, for a more thorough exposition of this way of thinking about how concepts are extended.