Chapter 3 - You Own Yourself and Nothing Else

Summary

Chapter 3 argues that the Lockean proviso entails two further conclusions embraced by social anarchists. First the chapter argues that, contrary to what right-libertarians typically maintain, the Lockean proviso implies that no one owns any natural resources. This is because any appropriation of such resources would leave others worse off in a way that the proviso does not allow. By contrast, the chapter argues that the proviso is necessarily satisfied when it comes to each agent’s own body ‒ at least, if self-ownership is interpreted in the way proposed in Chapter 1. Thus, while people do not own any external resources, they can easily come to own themselves via acts of self-appropriation. The chapter then goes on to defend at greater length the interpretation of the Lockean proviso proposed in Chapter 1. Finally, it concludes by discussing what the proposed position implies vis-à-vis the rights of children.

The granary is full; the national treasury is substantial. But the starving and frozen are everywhere. It is the result of the private ownership of property.

Chu Minyi, “Universal Revolution”

Chapters 1 and 2 have introduced the social anarchist position and the moral tyranny constraint. Chapter 1 proposed that social anarchism should be understood as a set of theses that includes (i) the consent theory of legitimacy (which, in turn, implies philosophical anarchism), (ii) the Lockean proviso, (iii) the anarchist self-ownership thesis, (iv) the denial that there is any existing private property, and (v) an endorsement of luck egalitarianism as the moral principle regulating the permissible use of unowned external objects (what was there called “the anarchist conclusion”). Chapter 2 then introduced the moral tyranny constraint and argued that this theoretical desideratum entails the first and second anarchist theses as well as luck egalitarianism’s responsibility component. This conclusion represents the first step in the book’s broader project of demonstrating the coherence of the social anarchist position – that is, that the position’s posited theses are not a set of arbitrarily selected (and potentially conflicting) moral principles, but, rather, stand in relations of logical entailment to one another. However, even if both the consent theory of legitimacy and the Lockean proviso follow from the moral tyranny constraint, that still leaves three additional theses that do not obviously stand in any logical relation to these aforementioned propositions.

This chapter will demonstrate that both the self-ownership thesis (as articulated by ASO in Section 1.3) and the rejection of private property follow from the Lockean proviso (and a few other plausible premises). This argument turns conventional libertarianism on its head in two respects. First, libertarian philosophers generally take the self-ownership thesis to be a foundational commitment while viewing the proviso as an auxiliary thesis that one might adopt to avoid the unsavory implications of unlimited appropriation. This prioritization is evidenced by the fact that natural rights libertarians universally endorse the self-ownership thesis but only a proper subset endorses the proviso, with many rejecting it as an ad hoc restriction on the power to acquire property.¹ However, this chapter suggests that the proviso should be treated as the more basic commitment with self-ownership ultimately following from the proviso. Second, it is typically assumed that the Lockean proviso allows for the appropriation of a significant quantity of natural resources.² Against this assumption, the chapter will argue that the proviso actually precludes almost all appropriation of resources (while still allowing people to appropriate their own bodies). The chapter will thereby take a right-libertarian premise and use it to derive a conclusion favored by social anarchists, namely, that people own themselves and nothing else.

The argument proceeds as follows. Section 3.1 argues that the Lockean proviso entails the social anarchist rejection of private property. Specifically, it argues that the proviso, at least as is has been specified in Section 1.2, is stringent to the point where it will not be satisfied by practically any act of initial appropriation, thereby precluding the conversion of natural resources into property. Section 3.3 then argues that, although the proviso is almost never satisfied when it comes to natural resources, it is necessarily satisfied when it comes to each person’s own body – at least, if appropriation is taken to only establish the weaker ownership rights posited by ASO. Thus, the section concludes that the proviso entails that persons own themselves, but only in the sense specified by ASO. Given that much of the chapter’s argument rests on the specific interpretation of the Lockean proviso provided in Chapter 1, Sections 3.3 and 3.4 will provide an extended defense of this interpretation. Finally, Section 3.5 will discuss what the chapter’s conclusion implies vis-à-vis the rights of children.

3.1 The Proviso and Private Property

To see why the proviso entails the rejection of external private property, recall how it was interpreted in Section 1.2:

The Lockean Proviso – A person appropriates some unowned resource via some suitable action if and only if (a) her established claims would not leave anyone worse off under conditions of full compliance and (b) no subsequent waiving of those claims would leave others worse off under conditions of full compliance (where, in both cases, the baseline for comparison is the closest possible world where the appropriator did not exist).³

At first glance, the proviso might seem to preclude practically all appropriation, even without the additional restrictions that Condition (b) places on appropriators. This is because most unowned resources are both useful and scarce; that is, a non-appropriator would benefit from the use of those resources and there is not an available substitute that would allow her to acquire that same benefit at an equal or lower cost.⁴ Seemingly, the appropriation of such resources would leave others worse off_FC, as they would now be unable to obtain the benefit in question without incurring a greater cost under conditions of full compliance. Thus, any appropriation of these resources would violate the Lockean proviso.

However, in response to this suggestion, proponents of the proviso will note that one must consider the net effects_FC of appropriation, not just the costs_FC that it imposes on non-appropriators. To do this, one must attend to the various ways in which appropriation benefits_FC non-appropriators. For example, by precluding fully compliant non-appropriators from using a resource, a person’s appropriation might enable her to improve the resource in a way that is ultimately to their benefit (Schmidtz 1994). Or, alternatively, appropriation can prevent fully compliant people from destroying some resource, thereby allowing future non-appropriators to use and benefit from it (Schmidtz 1990). In such cases, although appropriation imposes certain costs_FC on others by forbidding their free use of the resource, it will actually leave them better off_FC on net, with this fact undermining the prior quick argument that most appropriation violates the Lockean proviso.

There are two things to be said in response to this defense of appropriation. First, even when one factors in the conservation and improvement of resources that appropriation enables_FC, there will still be many cases where appropriation harms_FC people on net. For example, the appropriation of beaches and other scenic locations will often violate the proviso, as fully compliant people will suffer due to not being able to access these spaces and will not benefit from any development made possible by compliance with the established rights. More generally, the proviso precludes the appropriation of any land if there is a single person who would both benefit from moving across that land and would not benefit from the development of that land. Similar remarks apply to consumable resources: While appropriation might leave many excluded parties better off_FC (because appropriation would allow for these resources to be either preserved or improved in the full-compliance world in ways that are ultimately to these non-appropriators’ benefit), appropriation will not occur if there is at least one person who would not benefit_FC in this way and, instead, incurs a net cost_FC in virtue of the owner’s claim against her using the resource. For example, suppose that the appropriation of a fishpond prevents full compliers from overfishing and depleting the fish stock (as they otherwise would have). Such appropriation might leave most people better off_FC, as they are better off purchasing fish from the owner for years than eating for free for a few weeks but running out of fish later. However, if there is even one person who does not benefit_FC in this way – for example, because she only wants to consume fish in the short term – then the appropriation of the pond will not satisfy the Lockean proviso.

Additionally, the proviso entails that one could not appropriate any object where a person would provide some benefit to another in exchange for the latter not using the object in question. In such cases, the appropriation of the object would strip_FC this second person of her bargaining power, as she would no longer be able to use the object in the world of full compliance. She would, thus, end up worse off in this world because she would not receive the benefit that she would have been paid absent appropriation.⁵ In this way, the proviso entails that a significant portion of natural resources cannot be appropriated even if one considers only Condition (a)’s contention that the claims established by appropriation must not leave others worse off_FC.

To arrive at the anarchist’s conclusion that practically all appropriation fails to satisfy the Lockean proviso, however, one must appeal to Condition (b) of the proviso. This condition asserts that, in addition to the established claims not leaving anyone worse off_FC, it must also be the case that no subsequent waiving of those claims would leave anyone worse off_FC. To see why this condition precludes almost all appropriation, consider the case of an explorer who discovers a waterfall and attempts to appropriate it. A few days later, a hiker arrives who wants to spend every morning swimming at the base of the waterfall. If the explorer’s act of appropriation succeeded, then she has a set of claims against the hiker swimming, where it is assumed that each day’s swim is a distinct action, and the explorer has a distinct claim against each. As it turns out, full compliance with these claims would actually be to the hiker’s benefit: Although not swimming each day imposes a cost upon the hiker, it also keeps her from unknowingly polluting the only available water source with her sunscreen, thereby making the water forever taste of soap. Because the hiker prefers never swimming to drinking soap-flavored water, she would end up better off on net if she were to fully comply with the explorer’s full set of (hypothetical) claims. Thus, the explorer’s appropriation would satisfy an unamended statement of the Lockean proviso that merely asks whether the explorer’s full set of established exclusionary claims leaves others worse off_FC (i.e., a proviso that includes Condition (a) but not Condition (b)).

However, Condition (b) of the proviso holds that one must ask whether the hiker would be left worse off given any possible pattern of waiving of the explorer’s posited claims. Suppose, for example, that the hiker wanted to take a post-arrival swim and the explorer decided to waive her claim against this one action. Such waiving would leave the hiker worse off in the full-compliance world, as she would end up with soapy tasting water (due to her swim) and would not get to swim on any of the other days. Given that a possible pattern of post-appropriation waiving would leave the hiker worse off_FC, the explorer’s attempted appropriation of the waterfall does not satisfy the restated Lockean proviso and, thus, does not succeed.

This case helps to illustrate why the proposed interpretation of the proviso entails that practically all purported appropriations of external resources violate the proviso. Note that, for any posited appropriation of a resource, there is a possible world where the appropriator waives all her posited claim rights – functionally treating the resource as though it were still unowned – except for those claims that would impose the greatest costs on full compliers. For the proposed specification of the proviso to be satisfied, it must be the case that every non-appropriator is no worse off_FC in this world than she would be in the world where the appropriator did not exist to appropriate the resource in question. Given that the appropriation of practically any resource would violate this constraint, this specification of the proviso entails the anarchist thesis that practically no one has acquired – or will acquire – private property over external things via acts of initial appropriation.⁶ Specifically, Condition (b) of the proviso undermines the previous libertarian reply that appropriation often benefits_FC non-appropriators by enabling the improvement and preservation of natural resources. For, even if appropriation does establish claims that have this beneficial_FC effect, it also gives appropriators the power to waive these claims while leaving in place only those claims that impose costs_FC on non-appropriators. Given that an appropriately specified proviso must preclude the possibility of appropriators acting in this way, even much beneficial_FC appropriation will violate the proviso. Thus, one arrives at the anarchist contention that there has been (practically) no successful appropriation of external resources.⁷

3.2 The Lockean Proviso and Self-Ownership

Section 3.1 has argued that the Lockean proviso, as specified in Section 1.2, entails the absence of private property rights.⁸ This section will argue that, while the proviso may entail that there is no ownership of external things, it entails that persons can easily acquire ownership of the self – and, more specifically, self-ownership of the kind articulated by ASO. In this way, it aims to provide a novel ground for the self-ownership thesis while simultaneously demonstrating the coherence of the anarchist position (by showing that its various theses stand in the appropriate relations of logical entailment to one another).

So why does the proviso allow for ownership of the self when it also entails that there is no ownership of external natural resources? To answer this question, recall that the proviso holds that an act of appropriation succeeds if its established claims – and any possible subsequent waiving thereof – would not leave others worse off_FC relative to the world where the appropriator did not exist to appropriate. Further, note that the truth of the antecedent of this conditional can be determined by applying the following nonexistence test to the various costs_FC incurred by non-appropriators: such costs_FC pass the nonexistence test if and only if they obtain in the appropriation world but not the counterfactual world where the appropriator did not exist. For example, if P’s appropriation gives her a claim against Q eating some fruit that Q would have enjoyed in the world where P never existed, then Q suffers a cost_FC that passes the nonexistence test. This, in turn, implies that P’s appropriation leaves Q worse off_FC relative to the nonexistence baseline (assuming that there are no offsetting benefits_FC that similarly pass the nonexistence test) and, thus, P’s appropriation violates the proviso.

By contrast, an imposed cost_FC fails the nonexistence test if and only if it would equally obtain in the nonexistence world. For example, if P’s appropriation establishes a claim against Q eating some fruit, but Q would not have been able to eat that fruit in the nonexistence world due to it being out of her reach, then the costs_FC of non-enjoyment fail the nonexistence test. And, crucially, because costs_FC that fail the nonexistence test obtain in both the appropriation world and the baseline for comparison, they will not contribute to non-appropriators being worse off_FC in a way that would violate the proviso.⁹ Thus, an act of appropriation will satisfy the Lockean proviso if all of its imposed costs_FC (and all incurred costs_FC more generally) fail the nonexistence test. Or, to slightly restate this point, an act of appropriation satisfies the Lockean proviso if its established claims – and any possible waiving of these claims – would not impose costs_FC that pass the nonexistence test.

This test has little bearing on the appropriation of external natural resources, as there does not appear to be any case where the costs_FC imposed by exclusionary claims would equally obtain in the world where the appropriator never existed. In other words, when it comes to external appropriation, the imposed costs_FC – that is, all of the various costs_FC discussed in the previous section – still pass the nonexistence test, with the associated acts of appropriation thereby violating the Lockean proviso. Thus, the nonexistence test does not undermine the previous section’s denial that people have acquired or will acquire external property.

By contrast, when it comes to people appropriating their own bodies, the nonexistence test entails that the proviso is necessarily satisfied – at least, if self-appropriation is taken to establish the claims posited by ASO (i.e., claims against any actions that initiate bodily contact without generating unique supplemental benefit). To see why this is the case, consider the scenario where P’s body is unowned (due to her never having previously appropriated it) and Q is in desperate need of a new kidney. Suppose that P then self-appropriates, thereby acquiring a claim against Q that Q not take one of her kidneys. In this case, P’s self-appropriation leaves Q worse off_FC relative to the world where P has no such claim: Absent such a claim, a fully compliant Q would have taken one of P’s kidneys, thereby avoiding the pain and suffering of kidney failure (while a fully compliant Q would now suffer these costs given P’s claim against this action). However, as far as the Lockean proviso is concerned, the question is not whether Q is left worse off_FC relative to the world where some alternative moral facts obtain. Rather, the question is whether Q is left worse off_FC relative to the world where P did not exist – that is, whether the costs_FC she incurs pass the nonexistence test. And, notably, these costs_FC fail this test, as in the counterfactual world where P never existed, a fully compliant Q would be just as disadvantaged as she would be in the self-appropriation world where she complies with P’s established claims. Specifically, in both worlds, she does not get the kidney and suffers the associated costs. Thus, these costs_FC do not count when assessing whether P’s self-appropriation leaves Q worse off_FC in a way that would violate the Lockean proviso. This, in turn, implies that the establishment of a claim against kidney harvesting via self-appropriation does not entail a violation of the proviso.

One might be tempted to conclude that this result generalizes such that a person establishing any right to exclude others from her body does not leave others worse off_FC in a way that violates the Lockean proviso. Were this the case, then self-appropriation that generated the classical self-ownership right against all bodily contact would satisfy the proviso. However, this is a bit too quick, as there are many cases where establishing a general right against contact will impose costs_FC that pass the nonexistence test. For example, consider the case where P’s unowned body stands blocking the only entrance to Q’s office. Further, suppose that if Q is late for work, then her wages will be docked. Given that P’s body is unowned, Q can permissibly shove it to the side, thereby allowing her to enter the building and be on time for work. But what happens if P suddenly self-appropriates? Assuming that self-appropriation establishes a classical right against nonconsensual contact, it follows that P now has a claim that Q not shove her aside. This, in turn, entails that a fully compliant Q would be unable to access her office and would incur the associated cost. Thus, P’s self-appropriation leaves Q worse off_FC than she would have been otherwise.

Further, P’s appropriation leaves Q worse off_FC even after one applies the nonexistence test. In this case, the test asks whether a fully compliant Q would equally suffer the costs of being late for work if P did not exist. And, unlike in the kidney case, the answer here is no: While Q would be late for work in the full-compliance world, she would not be late in the full-compliance world where P does not exist (as there would be no one blocking her path). Thus, the costs imposed by P’s self-appropriation pass the nonexistence test, which, in turn, entails that P’s self-appropriation leaves Q worse off_FC relative to the nonexistence baseline; that is, P’s self-appropriation violates the Lockean proviso.

Why is it that the costs_FC imposed in the kidney case fail the nonexistence test while the costs_FC imposed in the doorway case pass this test? The explanatory difference here is that, in the kidney case, the cost of compliance for Q is limited to the loss of benefits derived from bodily contact without any loss of supplemental benefit. Note that, in this case, the only reason that Q is worse off_FC in the appropriation world relative to the non-appropriation world is that, absent appropriation, she would benefit_FC from the contact that she would make with P’s body (specifically P’s kidneys). When P then makes it such that a fully compliant Q cannot touch her body, that leaves Q worse off_FC than she would have been otherwise. However, the nonexistence of P equally makes it such that a fully compliant Q cannot touch P’s body and derive the associated benefits. Thus, Q is no worse off_FC in the appropriation world than she is in the world where P does not exist, with the costs of P’s self-appropriation thereby failing the nonexistence test.

By contrast, in the doorway case, Q shoving P to the side would uniquely generate supplemental benefit not caused by the physical contact itself, namely, Q getting paid her full wages. P’s self-appropriation then denies Q this supplemental benefit_FC (by giving P a classical self-ownership claim against Q shoving her), thereby imposing costs_FC on Q.¹⁰ Further, because the benefit is supplemental – that is, it is not derived from contact made with P’s body – the nonexistence of P would not equally impose these costs_FC, which is to say that these costs_FC would not have obtained in the world where P did not exist. Thus, they pass the nonexistence test, with P’s appropriation thereby leaving Q worse off_FC in a way that violates the Lockean proviso.

In other words, the costs_FC imposed by a claim against kidney harvesting fail the nonexistence test because they are correlative of a denial of a benefit_FC that is solely derived from bodily contact (i.e., non-supplemental benefit_FC). By contrast, the costs_FC imposed by P’s claim against being shoved in the doorway case pass the nonexistence test because they are correlative of a denial of unique supplemental benefit_FC. And, importantly, this result generalizes: The costs_FC imposed by an agent’s act of self-appropriation will fail the nonexistence test – that is, her self-appropriation will satisfy the Lockean proviso – if and only if her self-appropriation does not establish a claim against any person taking an action that uniquely produces supplemental benefit beyond those benefits that result from contact with the self-appropriator’s body.

There are two different conclusions that one might draw from the preceding discussion. First, one might conclude that the proviso largely precludes the possibility of self-appropriation. Those who favor this approach would insist that self-appropriation establishes the classical right against any nonconsensual bodily contact. They would then concede that the Lockean proviso is not satisfied in the countless cases where compliance with (any post-waiving pattern of) the established exclusion rights would preclude the realization of unique supplemental benefit. This, in turn, would imply that very few – if any – persons possess the self-ownership rights established by initial appropriation.

Fortunately for those attracted to the self-ownership thesis, one can reach an alternative conclusion by rejecting the classical assumption that self-appropriation establishes rights against all bodily contact. Instead, one would posit that appropriation only realizes the weaker exclusion rights referenced by ASO, that is, rights against all and only those actions that both result in bodily contact and do not uniquely produce supplemental benefit. Thus, when assessing whether P’s self-appropriation satisfies the Lockean proviso, one only needs to consider whether P establishing these more limited ASO rights would impose costs_FC that pass the nonexistence test. Further, given that P imposes such costs_FC on Q if and only if she establishes claims against people acting in ways that do uniquely produce supplemental benefit for Q, it follows that P’s self-appropriation necessarily satisfies the proviso. This, in turn, implies that all persons can freely self-appropriate and establish self-ownership rights of the kind articulated by ASO.¹¹

Of course, this result does not entail that one must accept ASO. As just noted, one could still choose to endorse the classical interpretation of self-ownership so long as one is willing to accept the conclusion that practically all acts of self-appropriation violate the Lockean proviso and, thus, almost no one owns themselves. However, first, it is unclear why self-appropriation must be taken to establish the classical set of rights against all bodily contact. Second, one might contend that which rights self-appropriation establishes is, at least in part, a function of which rights satisfy the Lockean proviso. On this approach, one does not start with a particular interpretation of self-ownership and then hope that self-appropriation satisfies the proviso; rather, one affirms a particular interpretation of self-ownership in virtue of the fact that such self-ownership could be established in accordance with the Lockean proviso. In other words, the reason for thinking that self-owners possess ASO rights is because these are the only rights that persons could come to possess. Such a supplemental premise would render the anarchist position coherent, as ASO would then follow from the Lockean proviso.

Finally, note that the foregoing argument does not entail the truth of the self-ownership thesis – that is, the proposition that all persons own themselves in the sense articulated by ASO. Rather, assuming that one accepts ASO, it merely demonstrates that a person owns herself if and only if she has carried out an act of self-appropriation (as all acts of self-appropriation succeed in virtue of the fact that they necessarily satisfy the Lockean proviso). However, if one accepts an account of appropriation such as that proposed in Section 1.3 – namely, that persons appropriate unowned resources by asserting that they own the resources in question (following Rose (1985, 81)) – then it follows that practically all persons own themselves, as almost everyone has, at some point, asserted that they own themselves (e.g., by saying “don’t touch my body!”). Thus, the foregoing proviso-based argument for ASO also provides a novel explanation of why people own themselves: They have successfully appropriated their own bodies in accordance with the proviso. This supplemental justification puts the self-ownership thesis on firmer philosophical footing, albeit at the cost of weakening the rights it assigns to each person.¹²

3.3 Comparing Baselines

The arguments of Sections 3.1 and 3.2 have leaned heavily on the specification of the proviso in Section 1.2. In particular, both arguments rely on its proposed baseline for comparison, that is, its contention that appropriation must not leave others worse off_FC relative to the world where the appropriator did not exist. Chapter 1 did not provide a defense of this specification, as it was primarily concerned with explicating other aspects of the anarchist position. However, now that the specified baseline has been shown to have significant implications, it is worth defending it at some length. Specifically, this section will defend the specification from the objection that it is unduly restrictive, where a more appropriately permissive specification might avoid the conclusion of Section 3.1 that practically no appropriation satisfies the Lockean proviso. To do this, it will consider the most promising alternative specification of the proviso – one famously endorsed by Nozick – and argue that this rival view is implausible. Further, this section will argue that the reasons for rejecting Nozick’s specification also support the anarchist’s proposed nonexistence baseline. It will, thus, conclude that one ought to favor the proposed baseline.

According to Nozick’s specification, an act of appropriation satisfies the proviso if and only if no one is left worse off – or, presumably, for the reasons discussed in Section 1.3, worse off_FC – than they would have been in a world without any appropriation at all.¹³ This specification is attractive to those who want a more permissive proviso that allows for appropriation precluded by the anarchist’s proposed specification, as it licenses appropriation in the many instances where non-appropriators would be harmed_FC by an individual’s appropriation but would still benefit_FC on net from the system of established private property as a whole. For example, suppose that peanut farmer P appropriates some unowned field, where this appropriation enables her to produce a large amount of peanuts. Further, suppose that this appropriation will worsen_FC the position of neighbor Q, as Q both previously enjoyed using the field and is allergic to peanuts (i.e., she has no interest in consuming the produced crops). In this case, P’s appropriation leaves Q worse off_FC than she would have been had P never existed to appropriate the field. However, it does not leave her worse off_FC than she would have been in the world where no appropriation ever occurs, as she benefits_FC extensively from others’ appropriation. For example, she benefits_FC when other farmers appropriate land and thereby become able_FC to grow crops without interference. Thus, P’s appropriation would satisfy Nozick’s specification of the proviso but not the proviso as it has been interpreted here. Given this result, some proponents of the proviso might contend that Nozick’s specification is superior to the anarchist one, as the former allows for appropriation in cases such as the one just described.

The problem with this contention is that the purpose of the proviso is to ensure that appropriation is justified; however, it does not appear that Nozick’s specification is able to play this justificatory role. This point has been expressed by Daniel Attas, who argues that specifications like Nozick’s – that is, specifications that compare how the entire established set private property rights affects Q to how she fares in a world without appropriation – are “completely off the point. The proviso is a requirement of particular appropriations. Particular appropriations have to involve counterbalancing gains in order to be justified … . [Thus,] a promise of increased benefits of the general system. … cannot justify [P] owning [an appropriated resource]” (2003, 359). However, while Attas is right to assert that it is the particular act of appropriation that must be justified, Nozick’s defenders might counter that systemic benefits_FC do justify particular appropriations. It is, thus, worth considering why one might think that the entire system plays this justificatory role. It will then be argued that these apparent grounds for affirming Nozick’s specification are philosophically untenable.

The suggestion here is that defenders of Nozick might advance the following baseline argument to defend the idea that systemic benefits justify individual appropriations:

1. 1. If an action – in tandem with various other actions – brings about a state of affairs that is non-inferior to the relevant baseline for comparison, then that action is justified.

2. 2. If an appropriation satisfies Nozick’s proviso, then it (in tandem with various other actions) brings about a state of affairs that is non-inferior to the relevant baseline for comparison.

3. 3. Thus, if an appropriation satisfies Nozick’s proviso, then it justified.

Admittedly, Premise 1 does not fit easily with Nozick’s signature view that the justice of a state of affairs is a function of the justice of the actions that bring it about, as it reverses the dependency relation between the justifiability of states of affairs and the justifiability of the actions that bring them about. However, without this premise, it is not clear how one could establish that appropriations are just by appealing to the harmless_FC effects of the entire established system of private property. Additionally, the premise can be supported by appealing to cases where the comparative non-inferiority of a resultant state seemingly justifies the actions that brought it about. Consider, for example, the case where a surgeon saves a patient’s life by a process that includes cutting open her chest. In this case, the incision seems justified because it, in tandem with other actions, brings about a state of affairs – namely, the patient continuing to live – that is non-inferior to the relevant baseline, namely, her death. This result seems to support the general claim asserted by Premise 1. One could then apply this general claim to the analogous case where a proviso-satisfying act of appropriation, in tandem with other appropriations, brings about a system of property rights that is non-inferior to the absence of such a system.

The defense of Premise 2 would then point out that Nozick’s specification of the proviso ensures that the entire set of private property rights does not harm_FC anyone – where the absence of harm_FC renders that system non-inferior to the relevant baseline for comparison (namely, the world where no appropriation occurs). Specifically, Nozick’s proposal would preclude any appropriation that tipped the balance such that the entire system of property rights worked to some person’s detriment_FC. Thus, his specification of the proviso ensures that all appropriations preserve the comparative non-inferiority of the system of property rights relative to the absence of any such rights.

The problem with the baseline argument is its assumption that the world without any appropriation is the relevant baseline for comparison to the appropriation world. Notably, the argument glosses over the question of which alternative state of affairs is the relevant comparison point when making judgments of non-inferiority. To answer this question, consider a modification of the surgery case where a nurse embeds a small metal sphere inside the patient’s chest during the operation. In this case, the action of inserting the sphere brings about a state of affairs (the patient living with a sphere in her chest) that is non-inferior to the alternative baseline where no surgery occurs and the patient dies. However, given that inserting the sphere into the chest was clearly not justified, Premise 1 will be false if this no-surgery state of affairs is the relevant baseline for comparison. Thus, to preserve the soundness of the baseline argument, one should seemingly hold that the relevant baseline state of affairs is one where the surgery occurs but no sphere is inserted into the chest – perhaps because the surgical nurse never existed carry out this action. Such a proposal delivers the correct result by blocking the implication that the sphere insertion was justified.

Similarly, consider the case where two parents throw their child a birthday party but a rude guest shoves birthday cake in her face. While this action upsets the child, suppose that she is glad that she got to have the party on net, even factoring in the cake incident (though she would have preferred a party where the incident did not occur). In this case, the guest brought about an outcome that is non-inferior to the comparison world where no party ever took place. Thus, if this is the baseline for comparison, then Premise 1 entails that her action is justified. However, given that the action is clearly not justified, some other baseline for comparison must be posited. And, again, it appears that the more plausible baseline for comparison is the counterfactual world where the rude guest did not (exist to) smash cake on the child.

These results can be generalized as follows. When considering whether the actions in the prior cases were justified, the initial assessments employed what might be called a compensation baseline, where this baseline was defined by taking the actual world and removing from it some conjunction of actions that, together, benefit a person on net (e.g., the complete surgical procedure or the entire birthday party), even though at least one of those actions actually harms the person (e.g., the sphere insertion or the cake smashing). However, in both cases, it was argued that Premise 1 of the baseline argument (“if an action – in tandem with various other actions – brings about a state of affairs that is non-inferior to the relevant baseline for comparison, then that action is justified”) is false if its use of the term “relevant baseline” refers to a compensation baseline. However, note that Premise 2 of the argument (“if an appropriation satisfies Nozick’s proviso, then it (in tandem with various other actions) brings about a state of affairs that is non-inferior to the relevant baseline for comparison”) is true only if “the relevant baseline for comparison” refers to the world that lacks the entire advantage-generating conjunction of appropriations that have been (and will be) carried out by many people across time. The problem here is that this baseline is a paradigmatic compensation baseline. Thus, barring equivocation, Premise 2 is true only if Premise 1 is false – which is to say that the baseline argument is necessarily unsound and cannot support the contention that appropriations that satisfy Nozick’s proviso are justified. This, in turn, implies that Nozick’s specification of the proviso cannot fulfill its theoretical function and should therefore be rejected.

By contrast, the following revised version of the baseline argument is seemingly sound:

1. 1’. If an action brings about a state of affairs that is non-inferior to its corresponding nonexistence baseline, then that action is justified.

2. 2’. If an appropriation satisfies the anarchist proviso, then it brings about a state of affairs that is non-inferior to its corresponding nonexistence baseline.

3. 3’. Thus, if an appropriation satisfies the anarchist proviso, then it justified.

Specifically, one could defend Premise 1’ by appealing to the sphere and birthday cases, where, in each case, non-inferiority relative to a nonexistence baseline – that is, the baseline where the agent did not exist to carry out the action in the first place – would seem to justify the agent’s realization of a particular state of affairs. Strictly speaking, the states of affairs produced by unjustified actions in these cases were shown to be not non-inferior (i.e., inferior) to a nonexistence baseline. In other words, these cases do not function as counterexamples to Premise 1’. Further, if one adjusts these cases such that the action brought about a state that was non-inferior to the nonexistence baseline, then the action seems justified as a result. For example, suppose that the sphere prevented blood clots and thereby ensured the patient’s survival. Or, in the birthday case, suppose that hitting the child with cake was the only way to keep them from eating it and having a terrible allergic reaction. Together, these results suggest that Premise 1’ is true. Given that Premise 2’ is true as a matter of definition, one can then infer the proposition asserted by 3’ and thereby conclude that the proposed nonexistence specification ensures that appropriation is justified. In this way, the baseline argument ends up supporting the anarchist specification of the proviso rather than Nozick’s.

3.4 Defending and Emending the Nonexistence Baseline

The previous section suggested that Nozick’s baseline specification is the most obvious more permissive alternative to the proposed anarchist specification of the proviso. It then argued that the best apparent argument for Nozick’s proviso actually supports the anarchist specification, with proponents of the proviso thereby having reason to favor this specification despite the fact that it entails surprising conclusions that they would otherwise reject. However, given the importance of the nonexistence baseline, more needs to be said in its defense. Specifically, this section will argue that there is a supplemental reason for endorsing the proposed specification, namely, that it follows from the moral tyranny constraint. Next, it will argue that the nonexistence baseline is superior to an alternative counterfactual inaction baseline. Finally, it will present a slightly technical emendation of the proposed baseline so as to bring the proviso into full compliance with the moral tyranny constraint.

To begin, note that there are many libertarians who reject the Lockean proviso, as they take it to be an auxiliary theory that can be costlessly excised from the core set of propositions endorsed by libertarian property theorists.¹⁴ Given the existence of such proviso skeptics, proponents of the proviso need to provide an adequate justification for building the proviso into a theory of property. Fortunately, Section 2.4 provided just such a justification, namely, that the proviso follows from the moral tyranny constraint. However, whether this conclusion holds depends on how one specifies the content of the Lockean proviso. Thus, there is reason to interpret the proviso in such a way as to ensure that it does, in fact, follow from the moral tyranny constraint.

While Chapter 2 did not specifically demonstrate that the constraint entails the nonexistence specification, that gap can now be filled in here. Or, more precisely, it will be argued that a slight generalization of constraint entails the specification in question, where those who accept the constraint ought to accept the generalization as well. Recall that the moral tyranny constraint holds that a theory of duties is acceptable only if it precludes persons from unilaterally, discretionarily, and foreseeably leaving others worse off under conditions of full compliance. Additionally, recall that these qualifiers were built into the constraint because there is seemingly nothing problematic about a theory allowing persons to unforeseeably, nondiscretionarily, or non-unilaterally leave others worse off_FC. For example, suppose that a person makes a choice that, as a matter of pure bad luck (i.e., unforeseeably), leaves her much worse off than others. Further, suppose that she would not have ended up worse off had she made a different choice. In this scenario, a luck egalitarian theory of justice holds that she is entitled to equalizing transfers from others – a conclusion that seems unproblematic even though it entails that the person was able to leave others worse off_FC than they otherwise could have been. Similarly, there seems to be nothing problematic about the fact that luck egalitarianism entitles her to transfers if her advantage-destroying action was nondiscretionary, for example, because her action was a mere reflexive movement rather than an exercise of agency. Nor is there anything problematic about a moral theory permitting some person P to leave another person Q worse off_FC if the only reason that Q ends up worse off_FC is that she made a particular advantage-destroying choice in light of P’s action (i.e., Q’s loss_FC of advantage is a function of both P’s choice and Q’s subsequent choice such that Q is fully able to avoid that loss_FC).

While Chapter 2 opted for a more modest statement of the constraint that enumerated the qualifications restricting when a theory cannot allow persons to leave others worse off_FC, the posited qualifications suggest a more general statement of the moral tyranny constraint: A moral theory is acceptable only if it precludes a person from taking any action such that (a) this action leaves others worse off_FC and (b) the person is morally responsible for this action. With respect to Condition (b), note that each of the qualifications listed by the original moral tyranny constraint expresses an apparent necessary condition of moral responsibility. Seemingly, P is morally responsible for Q’s predicament only if P could have reasonably foreseen that Q would end up in this predicament. Similarly, P is morally responsible for that outcome only if she could have avoided bringing about that outcome (i.e., it was brought about by a discretionary choice). And, if Q’s predicament is brought about by Q’s own actions carried out in light of P’s actions, then that, too, seems to negate the P’s responsibility, as Q is responsible for her own predicament, with that responsibility exculpating P. Thus, it appears that moral tyranny requires moral responsibility – that is, the constraint should condemn a theory that allows P to leave Q worse off_FC in some situation if and only if P is also responsible for leaving Q worse off_FC in that situation.

If one accepts this more general moral tyranny constraint, then one must adjust the proviso if one wishes to sustain the desired entailment relation between the two propositions. Specifically, the proviso must prohibit all and only those appropriations where the appropriator is responsible for leaving others worse off_FC. And this, in turn, requires adopting the posited nonexistence specification of the proviso. To see why the proviso must be specified in this way, consider the case where person O both watches person B pour a bucket of sand in H’s house and declines to do anything to help clean up the sand. In this case, it is B who is responsible for leaving H worse off rather than O. But what explains this fact? The answer cannot appeal strictly to counterfactual choices that B and O could have made, as H would have been better off had either of the two chosen differently (i.e., had B not dumped the sand or had O not declined to clean it up). Rather, it seems that the best way to determine who is responsible for leaving H worse off is to compare the world where H incurs this cost to the world where various agents never existed. Given that H’s house would still have had sand in it had O not existed, H’s predicament cannot be attributed to O or her choices. By contrast, the proposed comparison would not vindicate B in this way. Thus, the nonexistence comparison appears to adequately demarcate when a person is responsible for someone else incurring a cost: O is responsible for imposing a cost on H only if H is better off in the closest possible world where O does not exist.

In short, the moral tyranny constraint insists that theories of duties not enable people to both leave others worse off_FC and be responsible for leaving them worse off_FC in this way. It, therefore, entails the Lockean proviso if and only if the proviso strictly precludes each agent from being responsible for leaving others worse off_FC via acts of appropriation. Further, given that an agent is responsible for leaving others worse off_FC only if they would be better off_FC in the world where she did not exist, the proviso must preclude any and all appropriation that leaves someone worse off_FC than she would have been in the closest possible world where the appropriator did not exist. Thus, one arrives at the conclusion that the moral tyranny constraint entails the proviso if and only if the proviso is specified in the way proposed in Section 1.2, namely, with the baseline for comparison being the world where the appropriator did not exist.

There are two objections to this argument that are worth considering. First, one might argue that, when assessing whether an agent is responsible for some cost, the relevant comparison is not the world where the agent did not exist but, rather, the world where she refrained from carrying out some set of actions that imposed that cost.¹⁵ In other words, O is responsible for a cost incurred by H only if H would have been better off had O refrained from exercising her agency in the situations in question – or, to restate this consequent a bit more precisely, there is some set of actions S such that H would have been better off had O refrained from carrying out every member of S. This proposal has the apparent advantage of comparative evaluative simplicity, as one need only consider a possible world where some set of actions did not occur rather than the more distant possible world where O did not exist at all. Additionally, it seems to equally deliver the correct results in the sand case: H would have been better off if B had refrained from pouring the bucket of sand in H’s house, but there is no action on the part O such that H would have been better off had O not carried out that action. Given that (a) the inaction comparison is able to equally demarcate responsible parties from non-responsible parties in this test case and (b) it is easier to apply than the nonexistence comparison, one might conclude that it should replace the latter comparison in the proposed necessary condition of responsibility.¹⁶ And, this, in turn, would entail that the Lockean proviso, if it is to follow from the generalized version of the moral tyranny constraint, must be specified such that the inaction world is the relevant baseline for comparison.

However, there are four reasons for favoring the nonexistence comparison over this proposed inaction comparison. First, it is not clear that the inaction comparison is simpler than the nonexistence comparison. Note that the former comparison is already more syntactically complex than the nonexistence comparison (where syntactic complexity is a standard metric for assessing theoretical simplicity) (Baker 2016). Additionally, contrary to initial appearances, the inaction comparison does not seem to be any easier to apply to specific cases. Note that, in order to exculpate someone, one must consider every possible subset of the actions she has carried out and compare the actual world with the possible world where every member of that subset was not carried out – where this possible world will often be quite distant from the actual one, for example, the world where the agent did nothing at all.¹⁷ Thus, carrying out the rival inaction comparison turns out to be significantly more epistemically demanding than merely comparing the single nonexistence world to the actual world.

Second, the inaction comparison seems to deliver incorrect results in certain cases. For example, suppose that O is moved against her will and placed in a doorway. Minutes later, a fire starts inside the building, but O declines to move, thereby blocking the doorway with her body. This, in turn, results in H sustaining a serious injury due to not being able to escape the fire. According to the inaction comparison, O is not responsible for H’s injury: H would not have been better off had O refrained from exercising her agency in this situation, as this counterfactual world is identical to the actual world (due to O not exercising her agency in either). By contrast, the nonexistence comparison delivers the intuitively correct result that O is responsible for H’s injury. Given that H can freely pass through the doorway in the world where O does not exist, she will avoid injury and thereby end up better off relative to the actual world. Thus, the nonexistence comparison appears to have greater extensional adequacy than the inaction comparison.

The doorway case suggests a third worry about the inaction comparison, namely, that it may prove difficult to draw a defensible metaphysical distinction between action and inaction or the exercise of agency and the absence of such exercise. If the comparison is to yield any determinate judgment, each action must have a counterpart nonaction (as opposed to some rival action, which might generate confounding effects). However, it is not clear what would count as a nonaction given a natural account of what actions are – for example, intentional positionings of the body across space and time – as standing in one place is equally an intentional spatiotemporal positioning of the body. Absent such an account, the inaction comparison will lack determinate content.

Finally, the inaction comparison does not provide adequate theoretical support for exculpatory judgments relative to the nonexistence comparison. Briefly, the nonexistence comparison supports the conclusion that a person is not responsible for some state of affairs because it functions as a premise in the following argument:

1. 1. A person P is responsible for some state of affairs S only if her choices are part of the explanation of why S obtains.

2. 2. If S would equally obtain absent P making any choices at all, then P’s choices are not part of the explanation of why S obtains.

3. 3. If S equally obtains in the world where P does not exist, then S would equally obtain absent P making any choices at all.

   Thus, if S obtains in the world where P does not exist, then P is not responsible for S.

By contrast, the inaction comparison could only be substituted into this argument by changing the argument as follows:

1. 1’. A person P is responsible for some state of affairs S only if her choices are part of the explanation of why S obtains.

2. 2’. If S would have equally obtained had P not carried out any subset of her actions, then P’s choices are not part of the explanation of why S obtains.

   Thus, if S would have equally obtained had P not carried out any subset of her actions, then P is not responsible for S.

While this argument would still be valid, it would be much weaker, as Premise 2’ is contestable: P’s choices could still explain why S obtains even if S would equally obtain if any conjunction of P’s actions were replaced by their inaction counterparts because P’s choice to not carry out some rival action(s) might explain why S obtains. Indeed, it was just such a choice by O in the doorway case that seems to explain H’s injuries.

Premise 2’ could be strengthened if one adjusted the argument as follows:

1. 1”. A person P is responsible for some state of affairs S only if her actions are part of the explanation of why S obtains.

2. 2”. If S would have equally obtained had P not carried out any subset of her actions, then P’s actions are not part of the explanation of why S obtains.

   Thus, if S would have equally obtained had P not carried out any subset of her actions, then P is not responsible for S.

The problem with this revision is that Premise 1” is open to contestation by those who contend that P might be responsible for S in virtue of some of her omissions explaining S (even if her actions do not explain S). By contrast, it seems much harder to dispute Premise 1 (and the identical Premise 1’), as it makes reference to P’s choices rather than merely her actions. Thus, the nonexistence comparison should be favored over the proposed inaction comparison.

While this first objection to the nonexistence comparison does not succeed, there is a second objection that can only be addressed by emending the proposed comparison and the associated specification of the Lockean proviso’s comparative baseline. Specifically, this objection notes that the nonexistence comparison will wrongly exculpate people (i.e., declare them not responsible) in cases where B has provided prior benefit to H. For example, consider the case where wind blows sand into H’s house, B cleans it up, but then, a few days later, pours a bucket of sand into H’s house. In this case, B seems responsible for H’s predicament in virtue of her second action. However, the nonexistence comparison would say that B is not responsible, as H would not be worse off had B never existed. To avoid this bad result, the comparison must be modified such that B is held to be responsible for leaving H worse off via some action only if H would be better off in the world where B never existed and all of the costs and benefits that B had previously provided to H – that is, all of the other costs and benefits for which B is responsible – were provided in some other way.¹⁸

In short, when assessing whether B is responsible for leaving someone worse off (or worse off_FC), one must carry out the more complex comparison presented just prior.¹⁹ Further, given that the proviso must be specified so as to ensure that it only precludes appropriations where the appropriator is responsible for leaving others worse off_FC, it follows that its baseline for comparison must correspond to this emended comparison. Specifically, the proviso must hold that appropriation succeeds if and only if its established claims – or any subsequent waiving of those claims – would not leave anyone worse off_FC relative to the world where the appropriator did not exist but all of her prior imposed costs_FC and benefits_FC still obtained. That said, this emendation does not seem to undermine the arguments presented previously. The establishment of – and most-harmful selective waiving – of claims over external resources will still almost always leave others worse off_FC than they would have been in the nonexistence world where they still possessed the prior past benefits_FC and costs_FC imposed by the appropriator. In fact, this adjustment actually makes the proviso slightly more restrictive for reasons that will be discussed in the next section. Similarly, self-appropriation still necessarily satisfies the proviso. The fact that the baseline for comparison now includes all of the costs_FC and benefits_FC previously produced by the self-appropriator does nothing to change the fact that all costs_FC imposed by her self-appropriation will be equally imposed by her nonexistence in that comparison world. Thus, the emended proviso still supports the anarchist contention that people own themselves and practically nothing else (with only those who have yet to self-appropriate being excluded from the set of self-owners).

3.5 Appropriation and Children

With the foregoing argument in place, it is now possible to discuss what the anarchist position entails vis-à-vis children. The primary thing to note is that it implies that children are not self-owners until they self-appropriate, where such self-appropriation requires the cognitive capacities discussed in Section 1.3. Specifically, it was suggested there that persons self-appropriate by asserting that they own themselves. Thus, persons cannot be self-owners if they lack the capacities to make such an assertion (the linguistic capacity needed to make assertions more generally, the cognitive capacity needed to conceptualize ownership, etc.). Given that infants and young children lack these capacities, it follows that they are not self-owners and that their bodies therefore qualify as unowned natural resources. There are two worries that might be raised about this implication.

First, one might worry that the absence of self-ownership unacceptably permits people to mistreat children in various ways, as children lack important claims against bodily contact of the kind possessed by adult self-owners. However, there are three things that can be said in response to this worry. First, practically no theories of rights, libertarian or otherwise, take infants and young children to be self-owners. Thus, it is not a unique problem for the anarchist position that it, too, denies self-ownership rights to young children. Second, as was discussed in Sections 1.6 and 1.7, the fact that someone does not possesses a (self-)ownership claim against a person taking some action does not imply that she lacks any claim against the person taking that action. Rather, she might have a distributive claim against that action (or, perhaps, some other variety of claim, though, for the sake of parsimony, no other kinds of claims have been posited here). One can then apply this observation to the case of children who have not yet self-appropriated: although these children lack self-ownership rights, they will still have a robust set of claims against mistreatment, as they will have a claim against any uses of their body that would leave them worse off than others absent some sanctionable choice on their part (where it is assumed that children lack the requisite capacities to choose sanctionably).²⁰ Granted, they have this same claim vis-à-vis all other natural resources and not just their respective bodies; however, given that they will be particularly affected by how people interact with their bodies, their bodily distributive claims will likely be much more restrictive than their other distributive claims (i.e., their claims against uses of their bodies will far outnumber their claims against uses of any other object). One can, therefore, expect that the anarchist position will entail that children have a claim against practically all actions that intuitively seem like child abuse or mistreatment.

There is a third quick reply that can be made to the worry that the anarchist position entails that children lack self-ownership rights. Specifically, one might argue that this result actually counts in favor of social anarchism, as it has the intuitively attractive implication that paternalistic bodily contact – that is, contact that benefits the child – is permissible. Note that ASO entails that self-owners have a claim against paternalistic interference with their bodies (so long as that interference does not uniquely generate supplemental benefits). Thus, if children were self-owners, one could not clothe them, for example, without infringing upon their rights. By contrast, if children merely possess distributive claims vis-à-vis their bodies, then it will typically be permissible to make contact with their bodies when that contact is to their benefit.²¹

The second worry that one might have about the proposition that the bodies of young children qualify as unowned natural resources is that it implies that parents can appropriate their children’s bodies before the children develop the requisite capacities to self-appropriate. Indeed, this concern is raised by Susan Moller Okin (1989, 79–85) as part of her critique of Nozick’s entitlement theory of justice. A quick reply to this concern is that parental appropriation violates the proviso, as there are many post-waiving patterns of the established claims that would leave the child worse off under conditions of full compliance. For example, parental appropriation would give parents a claim against the child putting food in her own mouth, where compliance with this claim would leave the child much worse off. However, one might worry that this reply is, in fact, too quick, as it neglects the way in which the proviso’s nonexistence baseline interacts with parental appropriation. Note that this specification – at least, the unemended version – entails that parental appropriation violates the proviso if and only if some subset of the established claims leaves the child worse off_FC relative to the world where the appropriating parent did not exist. However, if a child’s parent(s) did not exist, then the child would not exist either, with the cost_FC of nonexistence seeming to equal or even exceed whatever costs the child would incur by complying with her parents’ ownership rights over her body.

One way of responding to this worry is to argue that parental appropriation actually does leave the child worse off_FC relative to the nonexistence baseline (i.e., the world where the appropriating parent(s) – and, by extension, the child – did not exist). For example, one might maintain that being a moral slave to one’s parents is a “fate worse than death.” If it is better to not exist than have to comply with any arbitrary parental ownership claims vis-à-vis one’s own body, then parental appropriation still violates the proviso. Alternatively, one might simply respecify the baseline of the proviso to avoid this arguably marginal problematic implication. Specifically, one could hold that the relevant baseline is the closest possible world where the appropriator does not exist but where all of the non-appropriators under consideration do. Thus, the nonexistence world would be one where the child’s parents did not exist but she was somehow conjured into existence or engineered in a lab. And, given that the child is worse off_FC in the world where her parents own her body than she is in this respecified comparison world, it follows that parental appropriation violates the Lockean proviso.

An easier response, however, is to simply appeal to the more precise formulation of the proviso presented in the prior section. There it was argued that the proviso must be emended such that the comparison world is the one where the appropriator did not exist but all of her previously produced costs_FC and benefits_FC still obtained. Given that these benefits_FC include the child getting to experience life, the relevant baseline for comparison would be the world where the parents did not exist but the child still existed. Thus, parental appropriation of a child’s body would violate the emended proviso, as the established claims would leave the child worse off_FC than she would be in the appropriate baseline for comparison. This result, in turn, implies that no one can appropriate a child’s body prior to her developing the relevant capacities needed to self-appropriate. In this way, the anarchist position avoids the unacceptable implication that adults or parents can come to own a child’s body.

3.6 Conclusion

This chapter has attempted to demonstrate that the Lockean proviso simultaneously entails that most people own themselves in the sense articulated by ASO (as it is trivially easy to self-appropriate) and that there is no external private property. In this way, it has attempted to demonstrate the coherence of the anarchist position while also addressing some of the most obvious objections that might be leveled against the foregoing argument. It has also provided an extended defense – and slight emendation – of the Lockean proviso, thereby bolstering the starting premise on which the chapter’s argument rests. Finally, it has completed the explication of what anarchism implies vis-à-vis children that was started in Section 1.3. There it was argued that children lack self-ownership rights because they lack the capacities needed to self-appropriate. This chapter has now explained why this lack of self-ownership does not leave them vulnerable to permitted mistreatment or the appropriation of their bodies. In this way, it has demonstrated that the anarchist position can be employed to provide a well-grounded and extensionally adequate account of the rights of children.

Finally, it is worth noting that this discussion of children’s rights is another illustration of how the component parts of the anarchist position complement each other by jointly entailing attractive results. The anarchist premise that people acquire self-ownership via self-appropriation helped to explain why practically all adults – despite varying cognitive capacities – have the same set of rights while young children do not possess these rights. The anarchist rejection of external property in favor of luck egalitarian distributive claims then helped to ensure that these children still have rights against mistreatment and abuse (while permitting paternalistic bodily contact). And the anarchist interpretation of the Lockean proviso protected children from having their bodies appropriated before they had a chance to self-appropriate. In this way, the distinct components of the anarchist position come together to help answer notoriously difficult philosophical questions about the moral equality of persons and the rights of children.

The fact that the anarchist theses jointly entail attractive conclusions in this way also reveals an additional sense in which the anarchist position is coherent: Its separate theses hang together in the sense that affirming only some of these theses but not others will negate certain attractive implications and often generate unfortunate ones in their place. For example, rejecting the Lockean proviso – as radical right-libertarians are inclined to do – raises difficult questions about why adults are not able to appropriate the bodies of young children prior to their achieving self-ownership.²² One can perhaps solve this problem by contending that children are self-owners from birth, perhaps in virtue of the fact that they will eventually develop certain capacities later on. However, one must then address the apparent implication that parents are forbidden from making even paternalistic contact with their children’s bodies. Additionally, by grounding self-ownership in scalar cognitive capacities, one faces the difficult challenge of explaining human moral equality (as was discussed in Section 1.3). Similarly, if one denies that people possess distributive claims over unowned things while accepting the other anarchist theses, then it is difficult to explain why child abuse and other forms of mistreatment are wrongful, as one can no longer appeal to the anarchist’s posited distributive claims as part of this explanation. Thus, there is additional reason to affirm the entire set of anarchist theses as opposed to some proper subset – that is, reason beyond the fact that these theses both stand in entailment relations with one another and are also jointly entailed by the moral tyranny constraint – namely, that these theses jointly deliver favorable results that do not follow from the conjunction of any proper subset of the theses.