Recently, Walter Block began a short thread on the LRC blog concerning the libertarian position on setting booby traps for the purpose of defending private property from trespassers. The discussion by no means exhausted all of the considerations involved in this topic, but a longer treatment may help to clarify some of the principles concerning self-defence in a libertarian legal order.

Every person in a free society is permitted to use defensive force against invasions of their person or property. Booby trapping does not question the principle of self-defence per se; rather, the difficulty is with whether this particular mode of protection may be considered defensive at all or whether the trap constitutes, in and of itself, an aggressive act in the event that it is sprung.


The primary concern with booby trapping is that the amount of force dispensed by the trap is likely to be “disproportionate” to the aggressive act it seeks to deter. Such a concern judges proportionality by quantifying the opposing forces, suggesting that the maximum magnitude of the defensive force permitted must be judged by the magnitude of the invasive force. For example, it would be “disproportionate” for a booby trap to kill someone who is “merely” engaging in the aggressive but relatively minor act of trespassing in your garden shed, and it would be woefully over the top to electrocute a burglar intent on stealing something as insignificant as a garden gnome.

In the first place, it is important not to confuse a requirement for proportionality with a requirement for equivalence, i.e. the notion that you can only fight fire with fire or fists with fists. Rather, proportionality simply means that the response must be appropriate to the level of aggression encountered. It is perfectly acceptable, therefore, for the defensive force to be more severe than the aggressive force, whereas any requirement for equivalence would lead to absurdities. A contest between aggressor and defender is not like a boxing match between two, equally valid competitors where the entire purpose of the contest is discover who is the mightier of the two. Instead, the aggressor is behaving illicitly whereas the defender is not, and so the just and desired result is for the defender to win. Thus, there is no requirement of “fairness” which prevents us from stacking the odds in favour of the latter.

However, the main problem with trying to quantify the force deployed is that it confuses the imposition of defence with the pursuit of justice. The purposes of defence mechanisms are to a) deter acts of aggression prior to their occurrence, and b) bring an aggressive act which does occur to a close in order to minimise the harm to the victim. The purpose of serving justice, on the other hand, is to subsequently right the wrong that has been caused by the aggressive act, the dominant part of which is (or should be) restoring the victim as close as possible to the status quo ante. Therefore, proportionality in terms of the quantity of harm inflicted by the aggressive act is entirely appropriate; it would be heinous to execute someone for stealing a packet of crisps, for instance. But when it comes to acts of defence such proportionality may be impractical and/or impossible.

The first problem is that the appropriateness of mechanisms of defence has to be estimated prospectively, i.e. before and/or during the aggressive act. Judging the magnitudes of force involved in an incident may be straightforward enough after the aggressive act has occurred when the facts can be assessed thoroughly and rationally without the threat of imminent danger. In the latter situation, knowledge of the invader’s motive and purpose may have become more apparent, and the precise weaponry he was carrying in order to fulfil that purpose can be examined. So once we know, in retrospect, that someone is climbing your fence just to steal a worthless garden gnome with nothing more than his bare hands then yes, subjecting him to a seriously injurious electric shock (or killing him) would seem woefully out of kilter. But it is not necessarily easy to determine how much of a threat an aggressive individual may pose while the act of aggression is still occurring.

We can see this most clearly in cases of invasions to the person. In a simple robbery the perpetrator may be aiming solely at stealing your valuables before otherwise leaving you unharmed. But the terrified victim may have no idea what the masked individual intends to do, fearing that the latter has set his sights on raping, murdering, or both. Moreover, any simple crime can easily escalate into something worse should the criminal become “spooked” and decide, in the heat of the moment, to finish off his victim. Similarly, when it comes to invasions of property, we may have no idea, during the act of aggression, whether the trespasser of the shed was simply trying to steal a plant pot or whether he was trying to lay his own deadly booby trap that would kill the owner when the latter next visited the shed. We may be clueless as to whether an invader is climbing over a garden fence to steal a gnome or to commit the much more grievous crime of raping one of the property’s occupants. If the aggressive individual is intent on inflicting a level of violence as serious as death or rape then surely it would be appropriate to inflict the harshest defensive force in order to stop him?

It is true, of course, that certain levels of aggression will be unlikely or even impossible in particular contexts, and so most people will only wish to spend time and resources installing security systems that are designed to meet the most typical threats. No one is likely to arrange the same levels of defence for a child’s piggy bank as they would for a bullion vault, for instance. Context is likely to matter also for legality, and where the possible extent of an aggressive act is clearly bounded by the circumstances then so too should be any defensive force. Moreover, it isn’t always true that the extent of aggressive force is indeterminable during its occurrence, and there may well be occasions when the victim has a reasonable opportunity to select the most appropriate defensive response. If, for instance, a man’s wife delivers him a slap on the face in return for an ill-judged comment then it would be woefully excessive for him to shoot her dead, whereas the same response to an intruder in the pitch black of night might be perfectly warranted.

However, once it is accepted that the force dispensed by the victim is prima facie defensive, then the burden of proving that it was excessive is borne by the aggressor. The latter must demonstrate that he presented the victim with a reasonably clear opportunity to assess the specific threat during its occurrence, and, thus, to deploy an appropriate level of defensive response. Thus, if someone does choose to install a booby trap which, after the fact, is revealed to be “excessive” compared to the level of aggression intended by the trespasser, then the benefit of the doubt in this regard should not be given to the latter, the very person whom we know is committing an unlawful act. In fact, given that any opportunity for an aggressor to communicate the level of aggression he intends is unlikely to present itself in trespass situations (as the perpetrator is usually hoping to avoid any kind of detection or face to face encounter), such a threshold is likely to be very high, and the aggressor may have to prove that the defensive force was completely out of kilter to any kind of reasonably possible aggressive action, given the circumstances.

Even then, however, certain people, regardless of context, may be excessively fearful of trespass and invasion (particularly if they are frail or were victims of prior criminal acts), and can sleep soundly at night only if they know their property is installed with “excessive” trip wires and traps. Shouldn’t the ability for an innocent party to feel secure with his own property trump the need for a trespasser to bask in the reassurance that his illegal act will be met only by a “proportionate” degree of force? The same is true also with invasions to the person, and excessively frail or fearful persons may feel safe from “minor” robberies or assaults only if they are able to dispense lethal force without hesitation. A common legal doctrine is the “eggshell skull” rule – the principle that an aggressor’s liability for an injury to the victim is not mitigated by any unusual frailty that the latter may have. This rule is often summarised as the requirement that an aggressor must “take the victim as he finds him”, and so bears the risk of that victim’s particular vulnerabilities. Applying the spirit of this rule to self-defence, it is difficult to suggest that an innocent, yet excessively fearful victim should bear the burden of not injuring the aggressor “too much” during an act of defence, and so it is the aggressor who should properly bear the risk of being subjected to disproportionate, defensive force.

There is also the question of the deterrence function of defence. It is important to note that deterrence is not the primary rationale for either a defensive mechanism or a punishment procedure, and any deterrent effect must be bounded by the limits within which a property owner can legitimately enforce his rights. (For instance, threatening to hunt and kill a trespasser’s entire family would probably serve as an adequate deterrent but carrying out such a threat would hardly be justified). Nevertheless, deterrence is an important factor, if only because the preference of the property owner is for any aggression to be avoided in the first place. Moreover, deterrence is likely to be effective particularly when it comes to defence against crime rather than with after-the-fact punishment once we account for the (typically) high time preference of criminals. Such a time preference rate may cause them to ponder an immediate and obvious fallout from a defensive mechanism more keenly than the possibility of subjection to some distant, post-trial ordeal. Thus, the prospective trespasser will need to realise that the (immediate) cost of being subjected to the defensive force outweighs whatever value he can attain from perpetrating the aggressive act.1 One way to increase that cost is to make the defensive force more powerful than anything the perpetrator is prepared to inflict. In other words, the bigger the risk the aggressor has to take in order to commit an aggression the less likely he is to go through with it. Given this, is it not likely that our strong, psychological aversion to the imminent threat of personal injury would make booby traps an ideal method of deterrence? Indeed, just the possibility that an aggressor may be met by more than a commensurate degree of force may, itself, suffice as adequate deterrence, regardless of whatever actual defence mechanisms a property owner has prepared.

Instead of judging proportionately by reference to the strength of the defensive force that is dispensed, libertarians should give greater weight to two other factors:

  • The defensive act should target the aggressor only and should not inflict any violence upon innocent third parties;
  • The defensive act should be directed solely at stopping the aggressive act by bringing it to a close, and cannot serve any other purpose such as seeking vengeance or inflicting punishment.

The first of these is relatively easy to grasp in principle. If the purpose of defence is to protect innocent people from harm then it would be ridiculous to suggest that defence itself can harm the innocent in the process. So it would be woefully disproportionate, for instance, to fire a gun randomly into a crowded street, spraying passers by with bullets, in order to protect yourself from a thief. Similarly, an explosive booby trap set to protect your garden shed cannot end up destroying a neighbouring house. These are both fresh, aggressive acts against innocent bystanders and the owner of the neighbouring property respectively.2

The only exception to this rule is when an aggressor has commandeered the property of an innocent party in order to perpetrate the aggressive act. For instance, if A steals B’s car and uses it to run over C then it would be a lawful act of self defence for C to shoot at B’s car in order to stop A. Recompense to B for the damage sustained by the car should be made by A. This principle may be extended to include situations where the aggressor’s action make it likely that an act of defence by the victim will harm third party property even if the aggressor has not explicitly commandeered that property. It also includes the unfortunate situation where the aggressor uses an innocent person as, say, a human shield in order to perpetrate an act of aggression – for instance, if A takes B hostage and shoots at C but C cannot defend himself from A without shooting through the innocent B. In this instance it would be lawful for C to shoot B so as to shoot A, and (critically) A would be responsible for the injuries caused to B. While the lawfulness of such action by C against the innocent B may seem counterintuitive there are, in fact, sound theoretical and pragmatic reasons for this being the correct libertarian position.3

First, it should be borne in mind that B is not left high and dry to bear the costs of his injury himself. He is definitely the victim of an aggressive act, and he is entitled to a legal remedy. The only question is who should be considered the aggressor, A or C, not whether there is an aggressor at all.

Second, A has commandeered the body of B as a means of perpetrating the aggression against C, aggression against which C can be reasonably expected to defend himself (indeed, the expectation of C’s defence is the very reason why A has forced B to act as a human shield). The correct causal connection of the harm to B is, therefore, A’s seizure of B’s body, not the defensive force imposed by C (we will explore causation in more detail below). However, this means also that B (if he is able) is entitled to defend himself against C and, once again, A is liable for any injury to C that this causes.

Third, if A’s use of B’s body as a means of aggression against C prevents the latter from lawfully defending himself, then this suggests that B’s right to self-ownership against C trumps C’s right of self-ownership against A – in other words, that C is lawfully required to sacrifice himself for the good of B. This may be the most virtuous and heroic of options for C to take and he may, indeed, have serious misgivings about defending himself if an innocent person has to be harmed. However, the purpose of the law is to protect rights, not to promote saintliness. The anomaly is removed if we recognise A as legally responsible for the totality of the force dispensed and both B and C as his victims, regardless of who actually pulls which trigger for which bullet. Thus, C can lawfully defend himself against both A and B, and B can lawfully defend himself against both A and C, but only A is answerable for all of the injuries to any of the parties that are caused in this encounter.

Finally, if C was held responsible for the injuries sustained by B this would allow A to offload the costs of his aggression onto third parties and would, in fact, incentivise the use of human shields. For if A does not use a human shield in shooting at C then A would have to bear the medical costs of injuries he receives to his body if C defends himself. If, however, A uses B as a human shield, it would be absurd to suggest that he should have to pay for injuries sustained to his own body but that he can leave C to pick up the bill for injuries to B. The clear result of such a stance is that an aggressor such as A would always want to use a human shield such as B, protecting himself from injury safe in the knowledge that any harm done to the innocent human shield would be the responsibility of the intended victim (C). This is clearly undesirable and, instead, A should be responsible for all the consequences that can be reasonably expected to flow from his act of aggression, and this includes injury to innocent parties that he chooses to put in harm’s way.

The second aspect of proportionality means that an aggressor can only be subjected to force for as long as he poses a threat. Once the aggressive threat is, from an objective viewpoint4, neutralised then any further force applied to the aggressor by the victim would be governed by rules concerning punishment and justice, not self-defence. For example, if you shoot a burglar in the leg and the wound is sufficient for it to become objectively clear that he can do you no further harm, then the threat of aggression is neutralised and so further defensive force is unnecessary. If, therefore, you proceed to grab a meat cleaver and start hacking the wounded invader to pieces then these latter acts are no longer serving the purpose of preventing aggression and so cannot be considered defensive. Similarly, if you were to chase the burglar off your property so that it became objectively clear that his flight had brought the burglary to a conclusion then you could not pursue him for several miles before apprehending him and beating him to within an inch of his life. And finally, if you design a booby trap to incapacitate an aggressor, you cannot then set another device to drip corrosive acid onto him and torture him. Such responses may, of course, end up being proportionate to the requirements of punishment and justice, others will not. However, given that, as we have argued before, the pursuit of justice and the infliction of punishment will nearly always require some form of adjudication in a public forum, it is likely that the vast majority of extra-defensive force applied unilaterally by a victim will be classified as aggressive in its own right.

Provided, therefore, that a booby trap does not breach either of these two rules, it should be considered a proportionate act of defence. In any case, however, even though courts may well rule that “excessive” quantities of force dispensed by a booby trap are unlawful if the threat posed by the aggressor is clearly miniscule, this would not preclude the use of such force against more serious aggressors. Thus, the legality of deploying a booby trap may be judged on a case-by-case basis. This is something that the property owner would have to weigh up for himself in determining whether a booby trap is the most ideal method of defence. If the risk of unlawfully harming relatively minor aggressors is greater than the need to protect one’s property from more dangerous threats then a booby trap will not be used.

A property owner might, of course, set booby traps for the purpose of sadistic pleasure rather than deploying them as the most “appropriate” (given the context) method of defence. He may revel in dark delight at watching a man’s head get blown off as the latter trespasses into a garden shed. However, unless the sadist has deliberately enticed people onto his property for the purpose of injuring them (thus negating the characteristic of the booby trap as an instrument of defence) then, morally abhorrent as he may be, his pleasure should be regarded as incidental. Provided that the trap does actually serve a defensive purpose by doing no more than bring an aggressive act to a close and by targeting only the perpetrator, then it is difficult to suggest that it should be legally impermissible. After all, the trespasser wounded by the trap has one clear way of avoiding it – don’t trespass!

Life vs. Property

A philosophical objection to setting booby traps is that the preservation of human life is inherently more important than the protection of property, and so booby traps should not take the risk of inflicting either death or serious injury upon a person to stop him from “merely” trespassing upon your property. This, of course, assumes that the purpose of the trespass is not to inflict any violence upon an actual person which, as we suggested earlier, may not be apparent until after the aggressive act has occurred.

In order to refute this argument, we do not need to dig too deeply into the question of the relative “importance” of life or property. But we might as well mention in passing that there are many examples of property that are required to enhance or sustain life – food, medicines, and even a car if that is your only means of transporting those other items. If a trespasser wishes to either steal or destroy these kinds of property then why they should they not be protected by booby traps? We could also point out that not all people’s lives are likely to be cared for equally. If the trespasser is a serial burglar who subsists entirely off stealing from other people then dispatching this parasite to the afterlife may be nothing short of a benefit for the lives of everyone else.

The more pertinent argument, however, is that valuation is subjective. The trespasser will, in ordinary cases, value his life more than whatever he could gain by breaking and entering someone else’s property (which, as we suggested earlier, is precisely the reason why booby traps would make good deterrents). But there is no good reason for expecting the property owner to regard the security of his hearth and home as less valuable than the life (or bodily integrity) of a complete stranger who would break into it. Indeed, any attempt to delve deeper into a more “objective” method of comparing valuations risks judging culpability by reference to something like “social utility” in the style of Chicago School economists Ronald Coase and Gary Becker – which, in this case, would mean that a method of defence would be permissible only if it imposed a lower “net social cost” than permitting the act of aggression to proceed. Prominent Austro-libertarians have pointed out the absurdity of this theory.5

Indiscriminate Force

A more practical problem with booby traps is that they are automated devices which will trigger in a uniform manner once activated, as they have no ability to assess the specific threat posed by a particular trespasser. It is quite possible, therefore, that a) a relatively weak individual posing a minor threat, even if he intended to trespass, would be subjected to the same force as a stronger person posing a more alarming threat; and b) parties who are lawfully present on the property (or those who are otherwise innocent) may unwittingly activate the trap.6 This objection is tantamount to saying that if you wish to protect your property then you must either resort to security measures that keep people out without subjecting them to force (such as walls, locks, steel doors, CCTV etc.); or, if the application of force is necessary, then only a human being with the ability to exercise judgment (such as a security guard) should dispense it.

It may well be the case that installing non-injurious security features is more cost-effective to the property owner than installing a booby trap. This, however, is an empirical question – if a booby trap is, in fact, the most affordable and effective way of either preventing or dealing with aggression then why should a property owner be prohibited from using it? Indeed, the fact that a booby trap is likely to be makeshift and relatively unsophisticated (but nonetheless effective) may well lend it a degree of attractiveness, particularly in an impoverished area with high crime rates where more complex security systems would be unaffordable to a property owner, and where it is impossible for an individual to stand guard over the property personally for twenty-four hours a day. (It is in this sense that context can often make a difference to legality, as we suggested earlier). Moreover, if the effect of banning booby traps is to lead to people to defending property either in person or by employing security guards then does this not contradict the earlier objection that life is more important than property? For if actual people are required to protect property from invaders then isn’t it possible that they could be risking their lives in order to do so?

When it comes to the inability of a booby trap to exercise judgment in the force it dispenses, there is no particular reason to suggest that such a trap is any more or less dangerous than an armed person. At least the outcome of activating a trap is likely to be certain (and, to a careful intruder, may be easier to circumvent), whereas a relatively “trivial” trespasser may fare worse against a frightened homeowner brandishing a shotgun in the dark. Indeed, the very nature of crime is that, should the aggressor and the defender confront each other, neither is likely to be able to exercise the most rational and well considered of judgments. Moreover, it is quite possible that a person could underestimate the threat posed by a trespasser, and fail to stop what goes on to be a serious intrusion, whereas a booby trap may have taken care of the matter swiftly. At the very least, therefore, there is no self-evident reason for assuming that the (possibly limited) ability of a human to judge a situation will produce better outcomes than an automated device.

In any case, some security measures such as barbed or razor wire are employed routinely, albeit with some legal regulation. These too lack the ability to assess the nature of a particular threat and yet will dispense the same cuts and scrapes regardless. Should these devices also be banned on account of this?

Finally, it is true that innocent people or parties lawfully present on a property (such as invited guests) could activate a booby trap and, yes, a property owner who set such a trap would be liable for any injuries caused to the innocent party.However, as mentioned earlier with regards to a property owner’s liability for causing “disproportionate” injuries to a minor aggressor, this possibility is knowable in advance to the property owner. He must weigh the gravity of its risk when deciding upon which kind of security system to employ (and if he does proceed with installing a booby trap then he may well call upon complementary mechanisms to reduce the possibility of the trap targeting innocent persons). It is no reason for outlawing the use of booby traps per se.

Warning Notices

Walter Block suggests that any liability for injuring a trespasser with a booby trap is either mitigated or escaped entirely if the property owner places warning signs on the protected property – but that, conversely, the property owner is guilty of an aggressive act if he fails to erect such signs. It is not entirely clear whether a warning notice is required only because of the gravity of the harm that may be dispensed. Block seems to limit the requirement to “booby traps that can kill”. However, as we have already dealt with the issue of proportionality above, we will focus this discussion on whether a trespasser’s knowledge of the trap’s existence should make a difference to the legal outcome, regardless of the level of force dispensed.

It is, of course, likely to be in the best interests of a property owner to deploy warning signs if they serve to enhance the deterrent effect of a booby trap. After all, the preferred outcome for the property owner is for the prospective aggressor to simply leave the property alone, and so if the presence of a sign or notice is sufficiently dissuasive in this regard then so much the better. Indeed, the deterrent effect of measures such as barbed or razor wire relies on people being able to see how awfully sharp it is. This again, however, is an empirical question, and it may be the case that deterrence is more effective if the potential aggressor is unable to predict the dangers that may be lurking within a property he has set his sights on rather than having them advertised on a large notice. If deterrence is better served by concealment then why shouldn’t a property owner be free to omit the provision of warning signs?

That aside, however, the difficulty with Block’s stance is in trying to determine the principle upon which a warning notice serves to transform the springing of a booby trap from an unlawful act to a lawful one. In other words, why should a warning notice make a difference? There seem to be only three possibilities:

  • The warning notice serves to transform the nature of the booby trap from aggressive to defensive;
  • The warning notice, by imparting knowledge of the trap’s presence to the trespasser, breaks the chain of causation between the setting of the trap and the injury caused to the trespasser;
  • Block’s own justification is the invocation of “an implicit contract we have with each other to warn each other of booby traps that can kill.”

Regarding the first option, it is clear that the unilateral imposition of force does not change from being aggressive to defensive merely because a warning is provided. If a person comes up to you and holds a gun up to your head and warns you that he is going to shoot before proceeding to do so then he would still be liable for murder regardless of his warning. Instead, the quality that distinguishes a defensive act from an aggressive act is whether the force repels aggression. So once an aggressive act has actually started then force that is deployed in order to stop it is ipso facto defensive and no further justification for the use of it is required. The purpose of any warning, on the other hand, would have been to dissuade a prospective aggressor from carrying out an aggressive act before it began. It has no bearing upon the defensive nature of force dispensed in response to an aggressive act that has already begun.

Indeed, if it is incumbent upon the setters of booby traps to erect warning notices then should those who utilise methods of personal defence also be required to deploy such notices? Should, for example, a person concealing a handgun be forced to wear a sign round his neck drawing attention to this fact? Should a strong man be required to broadcast to all and sundry that he is hiding excessively powerful biceps under his coat that could floor you within seconds? As we indicated earlier, it may be a good idea for any defender to furnish a warning prior to an act of aggression in the hope that it will be deterred. If an attack seems imminent then the simple revelation of a personal firearm may be sufficient to repel the problem. But once an attack has actually begun, the force used to bring that aggression to a close is, by its nature, defensive, and so nothing more is required on the part of the defending individual.

The only possible justification for the purpose of warning signs in order to categorise a booby trap as defensive is an evidential one – the sign could serve to make the boundary of a property objectively clear so that any reasonable person would have realised that he is trespassing by the time the trap is triggered. (Indeed, any enforcement of property rights requires clear boundaries as it is impossible and, therefore, unjust to force people to act within limits that are not readily perceptible). If, therefore, it would be impractical or unduly burdensome to erect a comprehensive boundary such as a wall or fence, then warning signs (or signs designating the area as private property) may serve as a sufficient alternative. Similarly, if you suspect that a person is about to rob you, then revealing a gun and saying “stop” imparts to that person awareness of the potentially aggressive nature of his actions. Should he turn around and leave you alone (or provide a legitimate justification for his actions) then the situation is defused. If, however, the individual ignores this warning and proceeds farther then it is, by now, reasonably clear that his intentions are genuinely aggressive and that he isn’t just a shifty looking person. In other words, it is in your interests to demonstrate that your deployment of force is the last resort to deter an aggressor so as to leave a judge or jury in no doubt that the force you deploy is defensive. Nevertheless, if this evidential problem does not exist and it is agreed that the trespasser was an aggressor, then furnishing a warning is not a conceptual requirement for justifying the use of defensive force.

The defensive nature of the force dispensed by a booby trap is enough to dispose of the remaining two justifications for the requirement to furnish warning signs – that is, that a warning sign breaks the chain of causation, or is required as the result of an “implicit contract”. If the force deployed is, by its nature, defensive, then it is lawful, and so that is the end of the matter. However, even if we ignore this and presuppose that a booby trap is prima facie aggressive, neither of these further two rationales for requiring warning signs would successfully implicate the property owner.

When it comes to the chain of causation the question we are asking is whose action should constitute the legal cause of the injury resulting from the booby trap – is it the action of the property owner for setting the trap in the first place or is it the action of the trespasser for activating it as a result of his trespass? If the latter prevails, then the property owner would not be liable for setting the trap.

We have dealt in detail with the problem of legal causation in a previous essay on libertarian law, so we will make only a summary here. The general rule is that a person’s action will constitute a legal cause of aggression if that action brings about a reasonable expectation that a physical invasion to another person will result. If, on the other hand, an action requires a reasonably unexpected, intervening cause in order for the invasion to occur then causation is not established. Thus, causation is praxeological – it concerns whether the configuration of means that brought about the invasion to the victim can be attributed to the action of the defendant.

In many cases this will be clear cut. If I hold a gun up to your chest and fire then it is obvious that, once the trigger was pulled, nothing else was required in order to bring about a reasonable expectation that the bullet will enter your body. Difficulties arise, however, when there are intervening causes that are needed to assist in producing the aggressive outcome. For instance, if a person drops a hammer on you from the top of a tall building then isn’t gravity the actual “cause” of the hammer’s descent to the ground? The answer is no – once the aggressor has let the hammer go then the intervention of gravity was part of the reasonably expected course of events (or, in other words, gravity became the means with which the perpetrator caused the aggressive act). The perpetrator put into effect a configuration of means – the hammer and gravity – which, without his act, would not have resulted in the injury to your body. Similarly, causation is not broken when the actions of innocent intermediaries are required in order to complete the act of aggression. For instance, if you poison a pot of tea in a cafe, the fact that a waiter then had to serve the tea to a customer and the customer had to choose to consume it will not absolve you from liability for poisoning the customer. All of these later actions were part of the reasonably expected course of events to the extent that the actions of the waiter and the customer became the means of poisoning the customer. Indeed, the notion that the customer was the legal cause of his own poisoning would be to suggest that he committed suicide, which is clearly not the case.

Causation is not established, however, when a reasonably unexpected, intervening cause is required to bring about the invasion to the plaintiff (or, to put it another way, the defendant failed to employ means that were sufficient to bring about a reasonable expectation of the invasion). Basically, this means that merely “setting the scene” within which a later, unexpected cause arises is not enough. For example, let’s say a workman places a hammer at the top of a building and a bird picks it up, flies off with it and drops it on a passer by below. In this instance, it is true that the incident would never have occurred had the workman not brought his hammer to the top of the building. But the intervention of the bird was unexpected enough so that no one could suggest that the workman’s actions sufficiently employed means to bring about the hammer’s descent onto the head of a person below. Similarly, if you boil a pot of tea to be served to a customer and the waiter then proceeds to throw it into the customer’s face, it is true that the incident would have never occurred had you not prepared the tea. But the intervening act of the waiter was sufficiently unexpected so that no one could hold you liable for the scalding injuries to the customer.

With all of this in mind, what effect would a warning notice of a booby trap have upon causation for the injuries sustained to a trespasser as a result of that trap?

If a warning notice is provided, then we can certainly agree with Block that this would absolve the property owner from liability. Now that the trespasser knows that the trap is right in front of him and that it may activate if he proceeds any farther then it is his action – his decision to trespass – derived from his conscious appreciation of the factors involved that becomes the legally relevant cause of his injuries. In other words, he has undertaken a voluntary assumption of risk by proceeding with the trespass. The situation is akin to a waiter serving poisoned tea to a customer but, immediately prior to the customer drinking the tea the waiter informs him that the tea is poisoned. If the customer then shrugs his shoulders and proceeds to drink the tea regardless then his informed choice breaks the chain of causation, and the death is now a suicide.8 Moreover, like booby traps, visible, defensive methods such as barbed or razor wire also have the potential to cause serious injury. If you injure yourself from climbing over barbed wire then the legal cause is not the property owner laying the wire in the first place – it is your decision to try and climb over it when you know it is there. If these devices are permitted then so too should booby traps accompanied by warning signs.9

However, there seems to be no compelling case for the latter half of Block’s stance, and that the chain of causation would still be broken even if the trespasser was not greeted by any warning signs. The reason for this is that, as we indicated earlier, in order to be liable the property owner needs to bring about a reasonable expectation that the injury to the trespasser will occur as a result of his actions. But it is difficult to suggest that he has done anything more than merely set the scene. For when a stranger is confronted by your property boundary the reasonably expected course of events is that he will turn around and leave you alone. It is not that he will decide to break the law and enter your property.10 Thus, the trespasser’s decision to cross the property boundary in and of itself serves to break the chain of causation between the laying of the booby trap and the injury to the trespasser.

It may be true, as we suggested earlier, that the property owner is a sadist who desperately hopes and prays for a trespasser to invade his property so he can watch with glee as the trespasser blows himself to smithereens. But, as deplorable as this may be, liability results from actions, not from mere wishes or flights of fancy – as far as the law is concerned, what goes on in a person’s mind is entirely his own business. If, however, the property owner erected huge “Welcome!” signs with glittering arrows pointing right towards the direction of the booby trap (or if he invited a private guest onto the property) then matters would be different. All of these actions on the part of the property owner now create the reasonable expectation that a person will cross the boundary of the property and into the danger zone.

This stance seems to be true if we look not just at booby traps but at the presence of other “dangers” that may be lurking on your property. This could be anything from the storage of chemicals all the way down to inadequately maintained lighting or a broken handrail on the stairs. A property owner configures the inside of his property according to the level of risk that he is willing to bear because he, and his invited guests, are the only people who should be there. He and his guests may be comfortable with handling chemicals or perfectly happy to walk up and down stairs that have no handrail, and his right to private property entitles him to make these choices. Why should he have to provide external notices advertising his lifestyle preferences for the benefit of those who would breach this right?

Indeed, more widely, it seems difficult to derive, from the principle of non-aggression, a general obligation to make the external appearance of your property bear any relationship to what is inside of it for the benefit of people who have no lawful business being there. In fact, the opposite is likely to be the case precisely because which persons should be inside and outside of a property are likely to be different. One’s house may be a complete tip on the inside but we may prefer to keep the front garden tidy so as not to upset the neighbours. Moreover, the one right that strangers do have is that any dangers you may be nestling are confined to within your property’s limits and do not spill out to injure people outside the property. Naturally, therefore, nothing on the outside of the property is likely to be as “dangerous” as anything on the inside.

Finally, Block’s own justification for the use of warning signs is an “implicit contract” to warn others of “booby traps that can kill”. Unlike the detractor to whom he is responding, Block is correct to say that implicit contracts are a valid concept, but he is wrong to suggest that they can be invoked in this instance.

The first problem is that, on the face of it, it seems rather bizarre to suggest that liability for blowing someone up with a booby trap is attributed to a breach of contract. Surely, if the property owner is liable, the case is a straightforward one of aggression – i.e. murder, assault or battery? As we shall see this is, in fact, the case, and any attempt to suggest that there is contractual liability is to misunderstand the nature of property relations that arise under contract.

The purpose of a contract is to vary, by agreement, the pre-existing property rights and obligations of the parties because it is in their mutual interest to do so. Liability under contract arises if one of the parties then breaches these agreed obligations.11 For instance, if you wish to buy a car, then the pre-existing property rights are that you own the money and the dealer owns the car. By contract, you agree to vary these property rights so that you become the owner of the car in exchange for the dealer becoming the owner of the money. If, therefore, you drive away with the car without handing over the money then you become an aggressor because you have absconded with what is now the dealer’s money. But the dealer’s title to that money would never have existed in the absence of the contract – it is only because you agreed to vary the pre-existing property titles that you become liable.

In the case of a trespasser, matters are completely different. For the trespasser comes to the situation already in full possession of his right to self-ownership. If, therefore, any violence is inflicted upon him, then it is a straightforward incursion of his pre-existing right. The implicit contract theory, however, suggests that the trespasser arrives at the property without his right to self-ownership, and that ownership over his body is in within possession of the property owner – in the same way as when title to a car you want to buy is first in the possession of a car dealer. The trespasser then bargains (implicitly) for his right to self-ownership. As a result of the mythical contract, the parties “agree” that, if the property owner fails to erect warning signs, the trespasser gains his right to self-ownership from the property owner. Thus, any subsequent injury from a booby trap on the property is a breach of the “implicit contract” and the trespasser may sue. If, on the other hand, warning signs are erected, then the property owner retains “title” over the trespasser’s body, and the subsequent activation of the booby trap is therefore lawful.

The only plausible basis on which this could be made to work is if the act of trespass somehow causes the trespasser to cede ownership over his body (or some lesser, but effective right over his bodily integrity) to the property owner – ownership which the trespasser can then magically get back again through the implicit contract if the property owner fails to erect warning signs. But apart from the fact that such reasoning is enough to make one’s head spin, a trespasser, or any aggressor, does not cede any aspect of his self-ownership to his victim during an act of aggression. The legal rationale for any defensive force that you impose upon an aggressor is not the transfer of any “title” over that aggressor’s body – it is the enforcement and protection of the titles to your own person and property.12

Of course, a later remedy may require the aggressor to compensate you by forfeiting money or other items of his own property, but the purpose of this is to serve justice after the fact, not to furnish protection as it happens. Thus, the convoluted distribution of property titles needed in order to make an “implicit contract” operable in this instance is entirely fictional.

Second, regardless of whatever pre-existing property relations exist between trespasser and property owner, it is difficult to construe any type of “contract” in this instance anyway. An implicit contract exists when any express statement by the parties of their intention to contract is absent, but one can, nevertheless, infer such an intention from their actions. A good example is sitting down in a restaurant and ordering from the menu. The waiter never says “you contract to pay the prices listed if we serve you the corresponding items of food” and the customer never says “yes, I agree to pay to those prices once you have served us the food.” However, when judging the actions of the parties according to their customary and conventional context, it is clear that the restaurant isn’t gifting food to the customers and that the parties intend to contract for payment of the meal.13

Without any inference of intention to contract, implicit contracts cannot be invoked in situations just because you happen to think that people should owe each other obligations or should behave in a certain way. When it comes to a prospective trespasser appearing at the boundary of someone else’s property, it is difficult to construe any intention for them to contract as a result of their actions. The only intention that can be sensibly interpreted from the actions of the property owner is to be to be left alone, a right to which he is already entitled and so does not have to bargain for. It is completely fictional to treat hostile parties seeking to preserve their existing property rights in the same manner as convivial parties who wish to change their property rights in order to furnish each other a mutual benefit under a contract. Thus, any requirement to erect warning signs of booby traps must be justified by carving it as a general obligation that all human beings owe each other; it is not akin to the synthetic obligations that particular parties create for each other through a contract, and any attempt to suggest that there is a “contract” is to pretend that something exists when it clearly does not.14


Thus, after examining booby traps against a number of objections, there seems no reason to impose an outright ban on them from a libertarian perspective. The decision to deploy them should be left to the property owner and the most that can be said is that he will be liable in individual cases where the injured party is either innocent or a minor aggressor. There seems to be no justification for banning them outright.

In closing, it is worth stating this conclusion in layman’s terms: if you trespass upon someone else’s property then you do so at your own peril. This is not a particularly radical principle, and is one that has been watered down only by our current, statist legal systems. Rather, it is likely to find support far beyond the small circle of libertarians.

*     *     *


1Indeed, libertarian scholars who address the deterrent effect of punishment for aggression nearly always conclude that the perpetrator must forfeit something greater than that which he took from the victim in order for deterrence to be effective, even though, as we mentioned, deterrence should be incidental to a just punishment formula rather its primary rationale. See, for instance, Murray N Rothbard, Punishment and Proportionality, Ch. 13 in Id. The Ethics of Liberty, New York University Press (1998), 85-96; Walter Block, Toward a Libertarian Theory of Guilt and Punishment for the Crime of Statism, Journal of Libertarian Studies, Volume 22 (2011), 665-75.; N Stephan Kinsella, Punishment and Proportionality: The Estoppel Approach, Journal of Libertarian Studies 12:1 (Spring 1996), 51-73.

2Unfortunately, our contemporary legal systems are wont to disagree with this principle.

3For an alternative rationale to that presented here, see Walter Block, The Human Body Shield, Journal of Libertarian Studies, Volume 22 (2011): 625-30. Cf. Carl Jakobsson, The Negative Homesteading Theory: Rejoinder to Walter Block on Human Shields, Ibid. 99-102. Space precludes us from examining Block’s notion of “negative homesteading” in detail. Suffice it to say that, as we demonstrate below, it is not necessary to invent a tenuous new concept in order to conclude that C may lawfully shoot B.

4By “objective viewpoint” we mean from the point of view of a reasonable person standing in the victim’s shoes as opposed to the subjective interpretation of the actual victim. To illustrate with an extreme case, if A hallucinates that B is aggressing against him then this subjective perception by A would not serve to legitimise any force he applies against B, however sincerely A may believe it to be defensive.

5See, for instance, Walter Block, Coase and Demsetz on Private Property Rights, Journal of Libertarian Studies, Vol. 1, No. 2 (1977), 111-115; Gary North, Undermining Property Rights: Coase and Becker, Journal of Libertarian Studies, Volume 16, No. 4 (Fall 2002), 75-100. For a detailed comparison with the Rothbardian approach to crime, see Laurent Carnis, Two Irreconcilable Theories of Justice: Social Engineering vs. Ethics of Property, Journal of Libertarian Studies, Volume 22 (2011), 399-436.

6With regards to b) Block raises the possibility of “relatively innocent people who break into cabins in the woods in order to save their lives” – a favourite scenario for those who object to the non-aggression principle in the first place. Suffice it to say here that, while hypothetical scenarios can help to clarify a particular ethical principle, the validity of such a principle does not depend upon its ability to satisfy imaginative instances that are fashioned deliberately to elicit an emotive response; it merely needs to resolve the problems that actually arise in the real world. To paraphrase a legal saying, imaginary cases make bad law.

7By “innocent” we mean those people who, from an objective point of view, are not posing an aggressive threat to the property. Such innocence, however, is likely to emanate from the failure to make a property boundary clear, an issue we will examine below.

8Whoever poisoned the tea may, however, be guilty of attempted murder. It should also be noted that the voluntary assumption of risk must, indeed, be voluntary – i.e. the individual must have the opportunity to withdraw from the scene with his own person and property entirely intact. It cannot be used to prevent a person from exercising his existing rights over his own property. For instance, say a thief holds a gun up at you and shouts “hand over your money or I will shoot!” It is true that you now know the consequences of choosing to keep your money. However, as every option made available to you will subject you to aggression then the assumption of risk is no longer voluntary. Thus, the chain of causation isn’t broken, and the thief would be liable for murder if he proceeds to shoot you.

9There is, moreover, the ontological question of whether concealment and the element of surprise are requirements for a device to be classified as a “trap” in the first place. If you know about a danger that is right in front of you and you are able to avoid it then it is difficult for you to suggest that you were entrapped.

10It is no answer to this to suggest that the laying of the booby trap indicates that the property owner does, in fact, possess the reasonable expectation that someone will trespass. People routinely take precautions against devastating but unlikely events, such as purchasing insurance against fire or accidental death. Foresight of a mere possibility is not equivalent to the reasonable expectation of its occurrence.

11Block’s mentor, Murray N Rothbard, developed the libertarian theory of contracts that defines them as instruments that transfer title to property. See Rothbard, Ch. 19.

12Similarly, if you have consensual sexual intercourse with someone then the legality of this act isn’t derived from some kind of “title” that to you have gained to your partner’s body. Instead, the furnishing of consent bars your partner from later claiming that your act was aggressive.

13Block’s defence of implicit contracts by reference to the attempt of a barista to charge $1 million for a cup of coffee that a patron has been served is also mistaken. The reason why such a contract would fail is that special attention must normally be drawn to unusually onerous terms in order to prove that the customer intended to bear that burden (and, in any case, given that price is a critical contractual component, even an implicit contract is likely to require notice of a more regular price to be accessible to the customer prior to contracting, as it would be on a menu). In the absence of such attention, the contract would be voided on the grounds that the price of $1 million was not agreed (or, in legal jargon, that the parties failed to incorporate the price into the contract). It has nothing to do with any implicit contract to charge only a “reasonable” price. The difference is important because, under Block’s approach, there is a valid contract for a cup of coffee in exchange for a “reasonable” retail price, and so the barista would be able to sue the customer to pay this “reasonable” price (an amount which never existed in the minds of the parties). Under the correct approach, however, the contract is void ab initio, and so the barista would be entitled to recover only the actual, pre-existing costs he incurred in producing the coffee, returning him as close as possible to the status quo ante. The distinction might be negligible in the case of a cup of coffee, but could be significant with a more expensive good or service.

14For the sake of completion, some legal systems recognise “implied-in-law contracts” or “quasi-contracts” which can be invoked in the absence of any evidence of intention. These, however, respond to specific circumstances between specific parties (normally unjust enrichment) and are not used to impose upon you general obligations that you owe to the entire world.