Chapter 3 argued that no one has successfully appropriated external natural resources because (practically) no acts of appropriation satisfy the Lockean proviso. This chapter will provide an alternative route for reaching this conclusion by arguing that those who accept a consent theory of legitimacy must also concede that there have been no successful acts of appropriation. Specifically, it will contend that property ownership is a form of legitimacy and, thus, has the same necessary conditions as legitimacy, namely, consent. Given that no one has actually consented to the establishment of property, it follows that there is no existing private property, as social anarchists contend.

To reach this conclusion, Section 4.1 will begin by introducing a slightly modified notion of legitimacy called territorial legitimacy. Next, Section 4.2 will argue that anyone who endorses a consent theory of legitimacy should also endorse a consent theory of territorial legitimacy. Section 4.3 will then argue that property ownership in land entails territorial legitimacy (and, thus, has consent as its necessary condition). Further, Section 4.4 will argue that there is no relevant distinction between land and objects such that consent becomes a necessary condition of appropriating any natural resource – a result that, in turn, entails that no one has any private property rights over such resources. Sections 4.5–4.7 consider and reject three objections to the proposed argument. However, whereas these objections do not succeed, Section 4.8 will note that a consent theory of property acquisition is in tension with the claim in Chapter 3 that persons can easily appropriate themselves. It will then consider three possible ways to resolve this tension. Finally, Section 4.9 will conclude with a discussion of the relationship between philosophical anarchism and the anarchist conclusion – that is, the conclusion that the permissible use of external resources is governed by distributive claims rather than property claims.

4.1 Territorial Legitimacy

The crucial step in this chapter’s argument is to show that anyone who has property rights over some tract of land has the same normative power as a legitimate state, where the term “legitimacy” is used to refer to this power. For, if this is the case, then the consent theory of legitimacy entails that land ownership can only be established with the consent of others – that is, the initial appropriation of land has consent as its necessary condition. To see why property rights entail legitimacy, recall the definition of legitimacy offered in Section 1.1: Person P is a legitimate authority with respect to another person Q when P has the power to determine what obligations Q has via the issuing of edicts. In other words, if a legitimate P at time t issues the edict that Q must ϕ, then Q is obligated to ϕ at t. Additionally, it will be helpful to label the set of duties imposed by a legitimate authority the duty bearer’s political obligations.

Notably, this account makes legitimacy an interpersonal relation that might obtain between any two agents. Thus, one cannot simply assert that a state is legitimate tout court; rather, it must be specified which persons are subject to its legitimate authority. However, this notion of legitimacy is very different from the power that actually existing states claim to possess. As Simmons notes, one of the primary rights claimed by states is the jurisdictional control right to impose and coercively enforce laws upon all people within its claimed territory (Reference Simmons2016, 4–5). According to Simmons, the legitimacy of states should be understood as bounded by the borders of their territory such that “only those persons within a state’s claimed territories are claimed as subjects of that state’s authority, as bound by its laws” (Reference Simmons2016, 31).

Note that this makes the actual power claimed by states both weaker and stronger than unbounded legitimacy as defined just prior. On the one hand, it is weaker because states only claim the right to regulate conduct within their territory. By contrast, an unbounded legitimacy relation obtains irrespective of the location of the person(s) subject to the legitimate authority. On the other hand, the power claimed by states is stronger than unbounded legitimacy in that it is the power to specify the obligations of anyone who is within their respective territories. An alien – who, by definition, is not a person with respect to whom a state is legitimate in the unbounded sense – is, nonetheless, taken by the state to be obligated to comply with its edicts upon entering its territory. Further, the state takes her to be obliged to comply with edicts that were issued prior to her entry. If the state passed a law in January prohibiting drug use, an alien who enters the territory in February would, in the eyes of the state, be obliged to refrain from using drugs (despite the fact that she was not in the territory when the edict was issued).

To put this point more precisely, states claim the power to specify the conditional obligations of others, where such obligations are those that obtain only on the condition that the obliged party is within a given state’s territory.Footnote ¹ This power will be called territorial legitimacy, where some person P is territorially legitimate with respect to person Q if and only if there is some bit of territory T such that, if P issues an edict of the form “if Q is within T, then Q must ϕ,” then Q is obligated to ϕ if she is within T. States can then be understood as claiming that they possess universal territorial legitimacy; that is, they are territorially legitimate with respect to all people.

Section 4.2 will argue that it is territorial legitimacy that should concern consent theorists; that is, if one takes consent to be a necessary condition of unbounded legitimacy, then one should also take consent to be a necessary condition of territorial legitimacy. Section 4.3 will then argue that having property rights is equivalent to possessing territorial legitimacy. Thus, consent theorists should also take consent to be a necessary condition of the initial appropriation of private property. The remaining sections will then address some objections to this argument, discuss how it bears upon the self-ownership thesis, and explicate the relationship between philosophical anarchism and the anarchist conclusion.

4.2 A Consent Theory of Territorial Legitimacy

As noted in Section 4.1, territorial legitimacy is weaker than unbounded legitimacy, as the latter is a power to specify subjects’ unconditional obligations while the former is the power to specify conditional obligations (where such obligations obtain conditional on a subject’s location). Given this difference, a consent theory of legitimacy does not necessarily entail a consent theory of territorial legitimacy. Indeed, there is no logical inconsistency in simultaneously holding that (a) a person must consent if another is to have the power to specify her unconditional obligations and (b) the weaker power to specify conditional territorial obligations does not require such consent. The question, then, is whether a person who endorses Proposition (a) should also endorse a consent theory of territorial legitimacy.

There are a number of reasons to answer this question in the affirmative. First, from a purely dialectical standpoint, proponents of consent theory are typically concerned with the moral standing of actually existing states. Given that such states insist that they have territorial legitimacy rather than unbounded legitimacy, consent theorists will want to make consent a necessary condition of the former as well as the latter. Additionally, consent theorists would not want their thesis to be rendered irrelevant by the conditionalization of an unbounded legitimacy claim. For example, suppose that a monarch denies that she needs consent to oblige others because she is not legitimate in the unbounded sense but, rather, is simply territorially legitimate vis-à-vis the entire Milky Way Galaxy. Unless consent theorists are willing to admit that their thesis loses all relevance in this case – that is, when what is asserted is not “Q is obligated to ϕ when P says she must ϕ,” but, rather, “Q is obligated to ϕ if she is within the Milky Way and P says she must ϕ while in the Milky Way” – then the consent theorist should also take consent to be a necessary condition of territorial legitimacy.

More importantly, the grounds for adopting a consent theory of legitimacy equally support endorsing a consent theory of territorial legitimacy. Consider, for example, Simmons’ prominent argument by elimination for a consent theory of legitimacy (Reference Simmons2001).Footnote ² Simmons begins by introducing three possible categories of moral requirement under which political obligations might fall. First, there are natural duties, which are “moral requirements which apply to all [persons] irrespective of status or of acts performed” and are “owed by all persons to all others” (Reference Simmons1979, 13). Additionally, because the duties are owed to all others, the content of those duties must be general, making no reference to particular persons or institutions (Reference Simmons2001, 47).Footnote ³ By contrast, special obligations are owed by particular people to other particular people and arise from the actions of individuals. For example, the moral requirements generated by acts of promising are special obligations, as (a) only a proper subset of people are obliged to carry out the promised action; (b) only a proper subset of people are owed this action (where the content of the obligation makes specific reference to these people); and (c) the moral requirement to act did not previously exist, but, rather, came into existence via the actions of the involved parties.

Simmons divides special obligations into two subcategories: those that are voluntary and arise via intentional acts of consent (e.g., promissory obligations) and those that do not come about via voluntary action (e.g., filial obligations) (Reference Simmons2001, 45). Simmons then argues by elimination, arguing, first, that political obligations cannot be natural duties, and, second, that they cannot be nonvoluntary special obligations. He, therefore, concludes that political obligations must be voluntary special obligations, that is, they have consent as their necessary condition.

Simmons makes two arguments to support his claim that political obligations are not natural duties. First, he argues that political obligations are particular in a way that natural duties are not: Political obligations are owed to only one state, with the content of one’s obligation specifically referencing that state (Reference Simmons2001, 47). One might have a natural duty to support states that are just or aid states in desperate need, but one will owe this duty to any state that meets the relevant posited criteria. By contrast, political obligations are owed to only one particular state. Thus, political obligations cannot be natural duties.

Second, Simmons argues that natural duties come in two varieties. Negative duties are requirements that agents refrain from acting in certain ways. By contrast, positive duties require positive action by the agent. Simmons contends that, while negative natural duties are perfect and, thus, allow for little to no discretion in terms of how they are carried out, positive natural duties are imperfect and allow people a degree of discretion over how to discharge those duties (Reference Simmons2001, 48). For example, an agent is typically permitted to refrain from discharging positive natural duties if discharging those duties would impose a significant cost on her (47). Given that political obligations do not allow for such discretion – including those that demand the agent carry out some positive action – it follows that they cannot be natural duties and must, instead, be special obligations.

Having shown that political obligations are special obligations, Simmons still needs to show that they are of the voluntary variety as opposed to the involuntary variety. Here, again, he employs an argument by elimination, considering the most plausible theories of involuntary special obligation and rejecting each in turn. While Simmons’ arguments against these theories are too numerous to reconstruct here, he provides sufficient conceptual resources for constructing an abbreviated version of his argument that eliminates whole classes of theories without having to consider them individually.Footnote ⁴ Specifically, nonvoluntary special obligations can be divided into two kinds: those that are grounded in the provision of benefits – that is, have the provision of benefits as their necessary condition – and those that can obtain even absent any such benefit being provided. Simmons rejects the possibility of the latter, contending that it is implausible to hold that anyone might owe involuntary duties to an agent from whom they receive no benefit (Reference Simmons1979, 158).Footnote ⁵ As a supporting example, he considers the case of a fur trapper living in isolation so deep within the interior of a territory that the state is not able to provide her with any benefits such as security or defense (159). Simmons contends that, given her circumstances, the trapper is not obligated to comply with the laws of the state (e.g., its gun control laws) (159). Further, if one accepts that the trapper has no such obligation because she does not receive any benefits from the state, then it follows that receipt of benefit is, at least partially, a ground of having a (nonvoluntary) special obligation.

This leaves only one remaining competitor to voluntary special obligations: the class of nonvoluntary special obligations grounded in the provision of benefits. These obligations purportedly arise when the receipt of benefits leaves the recipient indebted to the provider. However, Simmons argues that the receipt of benefits fails to give rise to such special obligations. First, he notes that the mere receipt of benefits (as opposed to the acceptance of benefits) cannot ground obligations, as it is implausible to think that a person who explicitly refuses some benefit but has it forced upon her owes a special obligation to the provider of that benefit (Reference Simmons2001, 56). Further, even when benefits have been accepted, Simmons insists that this acceptance still cannot ground special obligations. Specifically, he argues that, even insofar as an indebted person owes some return to her benefactor, she does not owe whatever the benefactor demands (56). Rather, she merely owes some “fitting return” that is adequately “responsive to the benefactor’s needs” (56). Thus, benefaction cannot generate political obligations, as such obligations are “content-specific,” that is, they demand specific performances over which the obligor has little to no discretion (56–7). In this way, Simmons rules out the possibility of any sort of nonvoluntary political obligations; that is, one must be a consent theorist about political obligations/legitimacy.

Does this argument for a consent theory of legitimacy also commit its proponent to a consent theory of territorial legitimacy? The answer to this question will depend on whether any of the premises or inferences described previously would be compromised if one were to replace all references to the political obligations of unbounded legitimacy with references to the conditional obligations of territorial legitimacy. To put the question a bit more precisely: Must such conditional obligations also fall under the category of voluntary special obligations? Or could they be natural duties or involuntary special obligations? The answer to these questions appears to vindicate a consent theory of territorial legitimacy. Quick consideration of each step of Simmons’ argument suggests that, just as political obligations are voluntary special obligations, so, too, are the conditional obligations established by territorially legitimate authorities. Thus, territorial legitimacy has consent as its necessary condition.

First, note that, like political obligations, territory-specific conditional obligations are special obligations, as they are owed only to some person(s) rather than to all persons. There is a slight disanalogy between political obligations and territorial obligations, as the former are typically thought to be owed to a single political authority (the state) while the latter can be owed to multiple authorities, each corresponding to some distinct bit of territory. However, as with political obligations, territorial obligations are owed to a proper subset of people rather than being owed to all people.Footnote ⁶ Additionally, the territorial obligations that states claim to impose are both perfect and often demand positive actions – a conjunction of properties that is incompatible with an obligation being a natural duty. Thus, Simmons’ argument that the obligations associated with unbounded legitimacy cannot be natural duties applies equally to the conditional obligations established by territorially legitimate authorities.

Similarly, the fact that the obligations associated with territorial legitimacy are conditional fails to exempt them from the second step of Simmons’ argument, which denies that there are nonvoluntary special obligations grounded in something other than the receipt of benefit. Given that territorial obligations are special obligations, Simmons’ conclusion implies that they also cannot be both nonvoluntary and grounded in something other than benefit receipt. This implication can be further supported by appealing to a modified version of Simmons’ fur trapper case. Simmons’ contention is that the state is not legitimate with respect to the trapper because it does not provide her with any benefits. However, suppose, instead, that the state merely asserts that it is territorially legitimate such that the trapper must comply with its laws if she is within its territory. Such a weakening of the state’s asserted power does nothing to make it more plausible that the trapper must comply with its gun control laws. Thus, if territorial obligations are to be nonvoluntary, they must be grounded in the receipt of benefits.

Finally, consider Simmons’ rejection of the receipt of benefits as a ground for political obligation. Simmons suggests that benefaction cannot generate moral requirements that have the kind of specific content that characterizes political obligations. Similarly, the specificity of territorial obligations precludes them from being grounded in the receipt of benefits: They are requirements to act in the specific way(s) dictated by the territorially legitimate authority (when one is within the territory) without any of the discretion characteristic of benefaction-grounded obligations. Thus, the conditional obligations imposed by territorially legitimate authorities must be special obligations of a voluntary kind. Given this result, any consent theorist who bases her view on Simmons’ argument by elimination should also accept a consent theory of territorial legitimacy.Footnote ⁷

4.3 The Absence of Appropriation

Sections 4.1 and 4.2 introduced the concepts of legitimacy and territorial legitimacy and argued that, if one is a consent theorist about the former, one should also be a consent theorist about the latter. The current section builds upon this conclusion by arguing that the holder of private property rights in land is territorially legitimate with respect to all other persons. Thus, consent theorists are committed to the conclusion that the establishment of property rights in land has consent as its necessary condition.

To begin, note that to have a private property right over some tract of land is to have a bundle of rights including the right to use the land, the power to transfer the land (i.e., all of the listed rights and powers), the right to exclude others from that land, and the power to waive these rights.Footnote ⁸ This last-mentioned power is notable because, when paired with the right to exclude others from the land, it follows that the right-holder has the power to determine the conditions under which others are permitted to use the land. If P has property rights over some tract of land, then P has a waivable right to exclude Q from that land, where she can specify the conditions under which the right is waived. Thus, P has the power to make it such that Q rightfully occupies and/or uses that land if and only if Q complies with some edict issued by P. For example, P might declare that anyone who wishes to use the land – where such use includes standing/walking upon the land – must wear red. She would then have a conditional right against Q that Q wear red if Q is on the land. Further, because this conditional right has a correlative conditional obligation, it follows that Q is obligated to wear red if she is within P’s property. Thus, P has the power to establish conditional obligations for Q via the issuing of edicts, where the antecedent of the obligation is Q being within the bounds of some geographic territory. In other words, P has the power of territorial legitimacy described in Section 4.1.

Given that the consent theorist is committed to the proposition that territorial legitimacy has the consent of all claimed subjects as its necessary condition, she is consequently committed to the proposition that property rights in land have the consent of all claimed subjects (namely, all other people) as their necessary condition. This, in turn, implies a commitment to the proposition that the acts of initial appropriation that establish such property rights have the consent of all others as their necessary condition. Further, given that no one has, as a matter of empirical fact, consented to appropriation, consent theorists must maintain that such initial appropriation has not occurred, which, in turn, implies that all land remains unowned. Thus, consent theorists must deny that there are any existing property rights in land, with one anarchist thesis thereby entailing another.Footnote ⁹

4.4 Land, Resources, and Artifacts

Note that the aforementioned conclusion only applies to land-based property rights. But what about property rights over objects and resources? Are such entitlements also ruled out by a consent theory of legitimacy? If not, then consent theorists are committed to a much less radical position than the anarchist contends, as there could still be ownership of any resources aside from land (setting aside the proviso-based argument of Section 3.1). However, there is reason for thinking that a lack of property rights in land entails that all resources and objects are similarly unowned. Specifically, it appears that all property rights over objects entail the power of territorial legitimacy. To see why, recall that P is territorially legitimate with respect to some territory if and only if P issuing the edict that Q must ϕ if she is within the territory entails that Q is obligated to ϕ if she is within the territory. But what counts as a territory? And what counts as being within that territory? The answers to these two questions, it will be argued, reveal that to use an owned object or resource is to be within its owner’s territory.

With respect to the former question, territory might be thought of as a portion of physical space. However, this isn’t quite right – at least, insofar as this portion of physical space is understood as being a fixed spatial region. Consider, for example, the territory claimed by the United States Federal Government: From a cosmic perspective, this territory is moving extremely rapidly in a corkscrew-like motion as the Earth simultaneously rotates and orbits the Sun, which is, itself, in motion. Thus, to make sense of territorial claims, territory must be understood as space defined in relation to some bit of mass such as a planet. Specifically, a territory is a portion of the surface of some massive object (it may also extend above and/or below the surface, but this is more controversial).

The answer to the question of what counts as being within a territory is more straightforward: To be within a territory is to be in contact with that territory. For example, when a person walks onto land claimed by the United States, she makes contact with the relevant surface region of the Earth, thereby qualifying her as being within that territory. One might also contend that those who tunnel under or fly above the surface are within the territory. However, the suggestion here is that to hold such a view is to presuppose the more controversial view (noted just prior) that the region below and above the surface is part of the territory as well. Thus, to remain neutral regarding whether one should hold this more expansive notion of territory, one might say that Q is within some territory T if – but not necessarily only if – she is in contact with the relevant surface region of the relevant massive object.

Given this account of being within a territory, P is territorially legitimate with respect to Q if, when P issues the edict that Q must ϕ if she is in contact with the surface of some massive object, Q is obligated to ϕ if she is making such contact.Footnote ¹⁰ Once territorial legitimacy has been recast in this way, the apparent distinction between property rights in land and property rights over objects collapses. To have property rights vis-à-vis some object is to be able to declare that Q must ϕ if she is in contact with the object and thereby make it obligatory that Q ϕ-s if she touches the object. Thus, the owner of that object is territorially legitimate with respect to all other persons.

Put somewhat differently, the territorial legitimacy of the land owner (or head of state) entails that she is able to issue conditional exclusion orders barring others from making permissible noncompliant contact with some surface region of one of the very large objects that we call planets. Similarly, the object owner is able to issue conditional exclusion orders barring others from making permissible noncompliant contact with the surface region of the smaller objects that rest on what we call planets. Of course, there is a size difference between large objects (planets) and the smaller objects resting on those large objects, but, for these purposes, this difference does not seem morally salient. One might think of objects as microplanets that differ in size – but not in kind – from the macroplanets that we generally associate with territory.

Given the lack of a principled distinction between massive objects of different sizes, all property rights can be understood as belonging to a single kind (as opposed to there being distinct kinds of property rights over land vs. resources and objects). Specifically, a property right over any item includes a right to conditionally exclude others from coming into contact with some bit of mass – and, thus, bestows territorial legitimacy upon the rights-holder. Therefore, consent theorists cannot merely deny that there is any owned land; rather, they must also deny that there are any owned resources or artifacts. This conclusion represents a libertarian reason for rejecting external private property altogether.

4.5 Initial Appropriation and Obligation Imposition

There are a few different objections that libertarian defenders of private property might raise against this argument. First, they might appeal to a set of existing objections to the claim that initial appropriation requires consent. Specifically, a number of libertarian philosophers have raised objections to the popular argument that (a) initial appropriation imposes obligations on others and (b) one can impose obligations on others only if they consent to being so obliged. For example, Gerald Gaus and Loren Lomasky (Reference Gaus and Lomasky1990), Simmons (Reference Simmons2001), and Hugh Breakey (Reference Breakey2009) have all objected to Contention (b) by arguing that there are many examples of nonconsensual obligation imposition; thus, there is nothing problematic about imposing obligations via initial appropriation. Gaus and Lomasky appeal to the case of the outstanding professor whose excellent performance unproblematically imposes an obligation upon the head of her department to sign off on a merit-based pay raise (Reference Gaus and Lomasky1990, 492). Simmons cites (among others) the case of people who make use of a tennis court and thereby nonconsensually – but unproblematically – oblige others to not use it (Reference Simmons2001, 220). And Breakey presents a number of seemingly unproblematic cases of duty imposition, including the case of the person who tells another a secret and thereby obliges her not to tell anyone, as well as the case of the person who occupies some bit of physical space and thereby imposes obligations on others not to invade that space (Reference Breakey2009, 622–3).

Alternatively, Bas van der Vossen (Reference Van der Vossen2015) has argued against Contention (a) by positing that no one ever imposes new obligations upon others, as acts like initial appropriation merely change the practical requirements of other people’s already existing conditional obligations. To illustrate this point, van der Vossen suggests that, for any given agent Q, the following conditional statement would be obligatory (i.e., Q is obliged to make the conditional statement true): If some person P has hair, then Q does not touch P’s hair without permission. Given the existence of this conditional obligation, it follows that P does not impose any new obligations on Q when she grows out her hair; rather, she merely changes the practical requirements of Q’s preexisting conditional obligation, where these requirements follow from the conjunction of the obligation and empirical facts about the world (69–70). Similarly, van der Vossen contends that each person has a conditional obligation to treat other people as property owners if those people carry out acts of initial appropriation. Thus, when people engage in such acts, they do not problematically impose new obligations but, rather, change the practical requirements of that conditional obligation (74).

However, even if one concedes the objections to both Contentions (a) and (b), the argument of this chapter is still sound. Recall the previous contention that a (territorially) legitimate authority is not merely a person who has the power to impose obligations; rather, she is a person who has the power to specify the content of people’s obligations via the issuing of edicts. Thus, those who affirm the prior claim that legitimacy – and, consequently, initial appropriation – requires consent need only maintain that this particular method of obligation imposition has consent as its necessary condition, without having to defend Contention (b)’s much broader claim that all obligation imposition requires consent. Given this limited commitment, one might fully concede that playing tennis or telling people secrets imposes obligations upon others without their consent while simultaneously maintaining that one cannot nonconsensually impose obligations on others via the issuing of edicts.Footnote ¹¹ Indeed, while a person’s choice to play tennis may oblige others to stay off of the court, it seems highly implausible that she can oblige them to stay off of the court simply by ordering them to do so – unless, of course, they agree to comply with her orders.

In other words, those who endorse a consent theory of legitimacy need not insist that the generation of obligations always has consent as its necessary condition. Rather, they need only affirm the more modest thesis that a certain sort of power to generate obligations has consent as its necessary condition, namely, the power to oblige others within a certain geographic space via the issuing of edicts. Thus, even if the arguments against Contention (b) succeed in showing that one can impose obligations without consent, that does not negate the claim that the more specific power of legitimacy – and, consequently, the possession of property rights – has consent as its necessary condition.

Further, note that if these arguments did succeed in showing that initial appropriation requires consent, this result would, in turn, entail the falsity of consent theories of legitimacy. As was argued in Section 4.3, to possess property rights is simply to be a (territorially) legitimate authority; thus, if one can obtain property rights without consent, then one can also become a legitimate authority without consent, which is to say that the consent theory of legitimacy is false. However, if this is the conclusion of Gaus et al.’s arguments, then they have seemingly proved too much, as those who criticize Contention (b) typically have other commitments that entail the consent theory of legitimacy. For example, practically all of these critics are libertarians who believe that anyone can acquire private property, where such ownership gives them both a permission to use their property as they like and an immunity from the nonconsensual loss of this permission. However, if a state is legitimate, it can enact laws that impose regulations on persons’ property, thereby obliging them to only use their property in a certain way. Given that a legitimate state can strip people of their permissions in this way, the only way to preserve the claim that persons are immune from the nonconsensual loss of such permissions is to maintain that states are legitimate only if they have received everyone’s consent. Thus, while libertarians might think that they are rescuing property rights by attacking Contention (b), they are actually opening the door to states having the power to nonconsensually regulate or even transfer away people’s property claims.Footnote ¹²

A similar reply can be made to van der Vossen’s denial that initial appropriation imposes obligations on others. Suppose that one fully accepts his claim that an agent never imposes new obligations on others and, instead, merely realizes the antecedents of their already-existing conditional obligations.Footnote ¹³ Given this assumption, P can be understood to be territorially legitimate with respect to Q if and only if, for any given action ϕ , Q has the conditional obligation to ϕ if P issues the edict that Q must ϕ within her territory and Q is within said territory. Indeed, this is how territorial legitimacy was defined in Section 4.1. In other words, the consent theory of territorial legitimacy can be understood as insisting that Q’s consent is a necessary condition of her having this particular set of conditional obligations. Further, the contention has been that the possession of property rights entails that this same set of conditional obligations obtains – and, thus, that initial appropriation has consent as its necessary condition.

This restatement of consent theory helps to clarify why van der Vossen’s argument cannot function as an objection to this chapter’s conclusion. Even if he is correct in claiming that initial appropriation does not impose any novel obligations, the claim being advanced by the chapter is that the conditional obligations entailed by initial appropriation obtain only if consent has been given. Thus, consent would still be a necessary condition of initial appropriation even if such appropriation imposes no new obligations.

4.6 The Propertarian Objection

There is a second possible objection that right-libertarians in particular might raise against the foregoing argument. According to this objection, while property ownership entails territorial legitimacy, libertarians who think that state legitimacy requires consent need not concede that there is no private property. This is because their claim does not entail or presuppose the more general proposition that anyone must receive consent to be a territorially legitimate authority. Rather, only states must obtain consent because that is the only way for them to acquire the power of territorial legitimacy in a context of already-established property rights. The idea here is that, for any given region, private individuals arrived first, appropriated, and thereby became territorially legitimate with respect to – that is, owners of – certain holdings. Given their title to the land, later-arriving states are unable to establish ownership/legitimacy over individuals’ already-claimed territory – unless, of course, the latter agree to transfer their claims to former, thereby ceding to states the power to conditionally exclude people from the relevant territories. Thus, despite the fact that both property owners and legitimate states possess an identical moral power, only states require consent to acquire said power, as they uniquely face the challenge of establishing rights over a territory that encompasses other people’s already-claimed property. Call this the propertarian position.

There are four problems with this proposal. First, note that it makes the truth of the consent theory of state legitimacy a contingent matter, as the theory will be true only if certain empirical claims about the history of state formation and property formation are also true. Specifically, property claims over some territory must predate the formation of the state that claims that territory. While this temporal relation will certainly obtain in some instances, there are likely many regions that were uninhabited prior to a state claiming them, with people only establishing residence and state-sanctioned property claims (i.e., obtaining legal property rights rather than natural ones) in those regions later on.Footnote ¹⁴ Thus, the propertarian can, at most, claim that certain states are not legitimate with respect to certain unconsenting individuals. Indeed, there will be many people for whom the edicts of the state are morally binding even absent any form of consent having been given. Further, one cannot know whether a particular individual is obliged to comply with the edicts of a state without first conducting an elaborate empirical investigation of the history of that state and that individual’s claimed property to see which territorial right was established first. Thus, the propertarian position represents a significant retreat from the claim favored by most libertarians, namely, that any state must acquire consent if it is to be legitimate.

Second, propertarianism imposes an additional theoretical cost on libertarians because it entails that there could be nonvoluntary legitimate states in a world without previously acquired private property. Note that, in order to countenance the formation of private property without consent, the propertarian has to maintain that the only situation in which state territorial legitimacy has a person’s consent as its necessary condition is when the territory in question overlaps with her already-established private property. Thus, in the absence of such property, the propertarian would have no theoretical resources to deny the legitimacy claims made by states. Perhaps some libertarians will bite the bullet and concede that some people might have the power to impose obligations on others via edict in such a world; however, most libertarians – particularly those attracted to a standard consent theory of legitimacy – will likely reject this conclusion.

A third and related point is that the propertarian position is dialectically weak. Note that the position’s core claim is that there is an incompatibility between the existence of private property owners and a state being legitimate with respect to those owners. The propertarian then posits that there are such owners and rejects state legitimacy as part of a modus ponens argument. However, one might equally accept one of the many arguments for state legitimacy and employ the propertarian incompatibility premise as part of a modus tollens argument: Given that there are legitimate states and such states are incompatible with libertarian entitlements, it follows that there are not any such entitlements. Further, because the propertarian claim is that the only reason that states are illegitimate is because legitimacy is incompatible with private property, propertarians would have no independent grounds for denying the posited premise that there are legitimate states. Granted, they could appeal to their positive reasons for endorsing the existence of property rights. However, these reasons would then have to be weighed against the reasons for thinking that states are legitimate, with there being a nontrivial chance that the latter prove weightier, thereby negating the propertarian position.

Finally, note that propertarianism is incompatible with any consent theory of legitimacy grounded in considerations other than the conflict between state legitimacy and preexisting private property. As discussed just prior, the propertarian must concede that there is no problem with legitimacy obtaining without consent, as she insists that initial appropriation does not have consent as its necessary condition. On her view, consent is only needed when late-arriving states want to govern a territory, where some portion of that territory is already owned. However, this claim contradicts any position that holds that legitimacy requires consent irrespective of facts about property. For example, both Simmons’ position as described in Section 4.2 as well as other influential arguments against nonconsensual legitimacy (e.g., those advanced by Huemer (Reference Huemer2013) or van der Vossen (Reference Van der Vossen2019)) would contradict the propertarian claim that, in the absence of property, there can be legitimacy without consent. Given that such legitimacy allows for the establishment of arbitrary perfect, special obligations, Simmons would still insist that it has consent as its necessary condition. Similarly, van der Vossen (Reference Van der Vossen2019) observes that libertarians often reject nonconsensual legitimacy on the grounds that it amounts to an unacceptable form of moral subordination. Given this commitment, such libertarians would still seemingly reject nonconsensual legitimacy as an unacceptable form of subordination even when it does not conflict with preexisting private property rights. Thus, most libertarian consent theorists would not want to adopt propertarianism, as that would require rejecting their grounds for endorsing consent theory. Or, to put this point a different way, all of the arguments that they have developed for consent theory will bear against propertarianism, thereby calling its plausibility into question.

4.7 Commonsense Distinctions

Finally, property-sympathetic libertarians might object to the chapter’s thesis by appealing to a family of existing arguments for the nonidentity of property ownership and state legitimacy. What these arguments all have in common is that they rest on various commonsense claims about the relationship between private property and state territory. However, it will be argued that these arguments are either flawed or unacceptable to those who endorse core libertarian theses, that is, the people at whom the foregoing argument is directed.

The first commonsense argument is one that has been put forward by Lea Brilmayer (Reference Brilmayer1989, 15), Allen Buchanan (Reference Buchanan, Buchanan and Moore2003, 234), and Cara Nine (Reference Nine2008a, 149). It begins with the premise that, if person P buys a tract of land L within the borders of state S, P owns that land but S is still legitimate with respect to L, as S retains the right to regulate conduct within L, collect taxes on L, etc. If territorial legitimacy and ownership are identical powers as argued previously, then both P and S are owners of L. However, there cannot be multiple (non-joint) owners of L. Thus, ownership and territorial legitimacy are distinct relations. Specifically, on this view, legitimacy is a jurisdictional power to impose and enforce rules in L; by contrast, ownership exists within jurisdictions as a distinct power. Only if one accepts this distinction can one make sense of how all of the land within a territory belongs to the state (and, on a democratic view, those it represents) while certain tracts of that land are owned by private individuals (Buchanan Reference Buchanan, Buchanan and Moore2003, 234).

The first problem with this argument is that it is not clear how the jurisdictional power is supposed to differ from property ownership, particularly given the foregoing argument that P’s ownership entails the power to impose rule regulating conduct in L. While perhaps the above-mentioned philosophers might posit that ownership is in some sense subordinate to – and, thus, distinct from – legitimacy, that would seemingly just assert that which needs to be demonstrated. Second, note that libertarians cannot appeal to this argument because they reject its starting premise. On their view, ownership entails exclusive control of the owned thing by the owner. In other words, if P owns L, it cannot be the case that S retains the right to regulate and tax L. Admittedly, this understanding of property entails the counterintuitive result that the citizens of legitimate states cannot own property. However, libertarians would insist that this is the correct conclusion and that, while such citizens perhaps have legal property rights afforded to them by the state, such rights are not genuine moral ownership rights of the kind that they endorse. Indeed, the incompatibility of ownership and legitimacy is part of the reason that libertarians insist that legitimacy requires consent, as noted in the previous section.

A second quick argument also made by Buchanan holds that “property in land is conceptually and morally distinct from the right to territory … because land is not the same as territory” (Reference Buchanan, Buchanan and Moore2003, 232). Specifically, he maintains that land is “a geographical concept” while territory “is a geographical jurisdiction,” where a jurisdiction is domain in which an authority gets to make and enforce rules (232–3). However, while it is true that land is strictly a geographic notion, the ownership of land gives one the ability to make rules governing those within that region, as discussed previously. Thus, while the concept of territory might have normative implications that the concept of land lacks, it does not follow that the ownership of land is conceptually distinct from the possession of territorial legitimacy.

A third argument proposed by Nine (Reference Nine2008b) contends that ownership and legitimacy are distinct because they serve different functions. Nine argues that the rights associated with ownership protect people’s ability to “pursue their own conception of the good” (961). By contrast, the rights established by legitimate states protect actions and relations of the kind that “makes possible the establishment of justice,” for example, the enforcement of laws and the protection of the commons (961). This difference in function entails that ownership is not reducible to legitimacy (and vice versa).

However, libertarians who wish to defend the existence of property would have to reject this proposal for a number of reasons. First, most would deny that the proposed examples of justice-upholding actions are properly carried out by states rather than property owners. For example, anarcho-capitalists like Huemer (Reference Huemer2013) contend that law enforcement should be carried out by private companies and purchased on the market. And David Schmidtz (Reference Schmidtz1990) contends that private property rights are a way of protecting the commons from degradation and preserving resources for latecomers. Second, libertarians take the relations between people and their holdings to be relations of justice. Given that property rights protect the upholding of these relations, libertarians would deny that territorial rights are justice-promoting in a way that property rights are not. Finally, many libertarians would reject Nine’s background assumption that rights “are social tools that we use to protect and encourage the realization of certain values” (Reference Nine2008b, 961). For most natural rights libertarians, rights express pre-social facts about the justice and permissibility of various actions. They would, thus, deny that property rights have a function (in the relevant sense), and, by extension, would deny that property rights have a function distinct from territorial rights.

The final argument to be considered is one independently advanced by both Margaret Moore (Reference Moore2004) and Buchanan (Reference Buchanan, Buchanan and Moore2003). Drawing from democratic theory, this argument contends that a state’s territory differs from property in that the former belongs to the people of that state, where this belonging constrains state authority. For example, Buchanan argues that this belonging relation entails that exercises of state authority must be for the benefit of its citizens (Reference Buchanan, Buchanan and Moore2003, 234–5). Similarly, Moore argues that this relation precludes the state from transferring parts of its territory to other states, particularly if the inhabitants of those regions do not want such a transfer to occur (Reference Mellor, Collins, Hall and Paul2004, 141–2). These restrictions set territorial legitimacy apart from ownership, which lacks any such constraints upon its associated rights of use, exclusion, and transfer.

Insofar as libertarians endorse a consent theory of legitimacy, they cannot employ this argument to fend off the preceding argument against private property. Specifically, such libertarians cannot consistently accept the proposed democratic restrictions on state legitimacy, as they would not see a legitimate state as subject to any restrictions beyond those imposed by the consenting parties. According to consent theory, if all parties agree to obey the laws of the state, then those parties would be obligated to obey territorial transfer laws mandating that they now obey the laws of a different state. Similarly, they would be obliged to obey laws that were not to their benefit. Of course, a state’s subjects may have insisted on prohibitions on territorial transfer and/or non-beneficial legislation as a condition of their consent, thereby giving their state only partial legitimacy. However, the fact that it is possible for a state to have full legitimacy free of such constraints negates the claim that territorial legitimacy is limited in a way that ownership is not. Thus, libertarians sympathetic to consent theory could not appeal to this argument to reject the conclusion that property rights entail territorial legitimacy.

Even if one concedes that legitimacy is bounded in the sense described previously, the argument against property still goes through. Note that, in order to succeed, the argument does not have to demonstrate that legitimacy and the powers afforded by ownership are identical. Nor does it have to prove that property rights endow their possessors with a power identical to some alternative power that states possess (where it is Moore’s and Buchanan’s contention that this latter power is weaker than – and, thus, nonidentical to – property owners’ territorial legitimacy). Rather, the argument merely needs to show that ownership entails territorial legitimacy, as one can then infer from the consent theory of territorial legitimacy that ownership has consent as its necessary condition (given that, if A implies B and B implies C, then one can validly infer that A implies C). In other words, even if Moore and Buchanan are correct that states are not territorially legitimate because their powers are constrained in various ways that territorial legitimacy is not, it would still follow that property owners are territorially legitimate. Thus, property ownership would still have consent as its necessary condition irrespective of whether or not states are territorially legitimate.

4.8 Consent Theory and Self-Ownership

While the foregoing objections do not succeed, the conclusion that property has consent as its necessary condition creates an apparent problem for the anarchist contention that persons are able to acquire ownership of their own bodies through appropriation. If the appropriation of any bit of matter grants the owners the power of territorial legitimacy, then self-owners possess territorial legitimacy, as the bodies they own are just bits of matter like any other resource. Thus, consent theory seemingly entails that self-appropriation has consent as its necessary condition – a result that precludes the kind of easily achievable self-appropriation that was posited in Chapter 3.

The quick response to this objection is to note that the anarchist position affirms ASO as opposed to the classical self-ownership thesis – that is, that self-owners have claims against only certain forms of contact, namely, contact that does not uniquely generate supplemental benefit. Because ASO’s posited exclusion rights are much more limited than those posited by the classical self-ownership thesis, anarchist self-owners are correspondingly limited when it comes to their ability to issue edicts that conditionally oblige others. In the case of classical self-ownership, if self-owner P issues the edict that Q must ϕ if she is in contact with P’s body, then Q is obligated to ϕ if she is making such contact (as P has the power to waive her right to exclude Q from her body on the condition that Q ϕ-s). By contrast, this is not true if P has only the exclusion rights posited by ASO. Given that P only has a right against Q using P’s body in ways that do not uniquely generate supplemental benefit, she would have only the following power of self-sovereignty: If P issues the edict that Q must ϕ if she is in contact with P’s body, then Q is obligated to ϕ if she is making such contact and such contact does not uniquely generate supplemental benefit. Given that this is a much weaker power than territorial legitimacy, it is not clear that the basis for positing a consent theory of territorial legitimacy will also support the claim that self-sovereignty has consent as its necessary condition. Anarchists might thereby avoid any contradiction between consent theory and the self-ownership thesis by denying that self-sovereignty requires consent in the way that territorial legitimacy does.

What if this quick reply does not succeed? Suppose it turns out that self-sovereignty is, in fact, relevantly similar to territorial legitimacy such that it has others’ consent as its necessary condition as well. In this case, some anarchists may be tempted to simply abandon the self-ownership thesis and maintain that the permissible use of all resources – including bodies – is governed by distributive claims rather than property claims. This would not be an entirely intolerable outcome, as most of the anarchist position would remain intact: One would still end up with a rejection of private property and an endorsement of luck egalitarian distributive rights, both of which have been shown to follow from the moral tyranny constraint and other core libertarian theses such as the consent theory of legitimacy. However, as discussed in Sections 1.4 and 1.9, the self-ownership thesis has many attractive implications; thus, retaining the thesis would enhance the anarchist position’s plausibility. Additionally, an anarchist position that included the thesis would better reflect the commitments of self-identified anarchists who typically prize bodily autonomy.

Fortunately, there are a number of possible ways to resolve the tension between consent theory and self-ownership, even supposing that one rejects the quick reply. Specifically, this section will critically discuss three additional ways to avoid the posited contradiction, beginning with the least promising proposal and concluding with the most promising. That said, each proposal has its own theoretical advantages and disadvantages, and reasonable people might disagree about which is best when it comes to resolving the contradiction between the two anarchist theses.

The first way to avoid a contradiction between self-ownership and consent theory is to weaken the latter such that consent becomes a necessary condition of a person establishing legitimacy rather than a necessary condition of legitimacy tout court. One would then need to reject the anarchist contention that self-ownership is acquired rather than native – that is, one would have to hold that people enter existence with self-ownership rights already in their possession (either at birth or when they first attain moral personhood). Together, these adjustments would allow one to maintain that persons possess self-ownership rights even absent others’ consent due to the fact that these rights are given rather than established; by contrast, initial appropriation would require consent, as it bestows territorial legitimacy vis-à-vis some bit of matter on persons who did not previously possess this power. In this way, the anarchist could preserve the self-ownership thesis while still rejecting private property on consent theory grounds.

However, both of these proposed adjustments come at a cost. First, treating self-ownership as native rather than acquired gives rise to the theoretical problems discussed in Section 1.3 and Section 3.6. And, second, it is not clear why consent would be a necessary condition of establishing legitimacy but not a necessary condition of legitimacy itself. Perhaps one might hold that what is problematic is people having discretionary control over what powers they have. Thus, while consent is not needed to justify a person’s territorial legitimacy, her making herself legitimate requires such justification. The problem with this suggestion is that it seems to be the power itself that requires consent, even when it is not subject to a person’s control. Suppose, for example, that a monarch insisted that she has possessed territorial legitimacy over some region since birth. Few consent theorists would think that this makes her claimed power any less problematic than if it were acquired. This suggests that a distinction between established and native territorial legitimacy cannot support a more limited consent theory that only applies to the former; rather, such a weakening of the theory would be ad hoc and arbitrary.

A more promising way to resolve the tension between consent theory and acquired self-ownership is to further weaken the self-ownership thesis such that it not only assigns the more limited set of exclusion rights articulated by ASO but also does not assign self-owners the power to waive those rights conditional on certain future acts being performed. Without this power, self-owners cannot conditionally permit others to make contact with their bodies and, thus, cannot impose conditional obligations on others via the issuing of edicts. For example, standard ASO assigns self-owner P the power to waive her right against Q using P’s body in a way that does not uniquely generate supplemental benefit on the condition that Q wears red, where such conditional waiving obliges Q to wear red if she touches P in a way that does not uniquely generate supplemental benefit. By contrast, the even-more-limited version of ASO would deny that P has this power. This, in turn, entails that Q is obliged to not make the specified use of P’s body irrespective of what P says, thereby stripping P of her power of self-sovereignty. Given that the weakened version of ASO does not entail either self-sovereignty or territorial legitimacy, self-appropriation would be fully compatible with the conjunction of consent theory and the fact that no one has consented to anyone else’s self-ownership.

There are three potential objections to this proposal, though replies can be made to each. The first objection is that the theoretical choice to exclude the power to conditionally waive from the bundle of self-ownership rights is both ad hoc and unmotivated. Why deny persons this power aside from the fact that it makes self-ownership compatible with consent theory? In response to this challenge, one might appeal to various objections that have been made to the conditional exchange of goods and services. For example, I have elsewhere argued that anarchists’ opposition to limiting others’ freedom leads them to assert that conditional exchange is morally objectionable (2020b). If this thesis is correct, it would represent a principled, anarchist-friendly reason for denying that persons have the power to waive their self-ownership rights conditional on others acting in some way.Footnote ¹⁵

The second objection is that the proposal threatens to undermine this chapter’s foregoing argument against the existence of private property, as the property proponent could make the same argumentative move and posit that property owners lack the power to conditionally waive their property rights. Given that property owners are territorially legitimate only if they can waive their claims to exclude, the acquisition of non-waivable exclusionary property rights would not require consent, thereby sidestepping the argument presented in Sections 4.1–4.4. Admittedly, this move would succeed in its aim of defusing this chapter’s consent theory-based argument against private property. However, it will be unacceptable to the defenders of property who would seek to deny this argument. Additionally, Section 3.1’s argument from the Lockean proviso would still be sound even if exclusionary property rights are not waivable. Thus, the anarchist could still maintain that no one owns any external natural resources (though she would have only one argument to support this claim rather than two).Footnote ¹⁶

Finally, one might reject this proposal on the grounds that non-waivable exclusion rights are unattractive, as they do not give self-owners enough moral control over their bodies. Consider, for example, the person who wishes to spar in a boxing ring with a friend or be kissed by someone who made her laugh. Absent a power to waive her self-ownership rights, she could not give others moral permission to make such contact with her body. Such activities would thereby wrong her – and this would be true even if she held all of the attitudes and performed all of the actions associated with giving consent. Such a result would be highly implausible, thereby ruling out this strategy for rescuing self-ownership from consent theory.

In response to this objection, note that the proposed weakening of ASO does not deny self-owners the power to waive their rights. Nor does it deny them the power to waive their rights conditional on some already obtaining fact or even most future occurrences. Rather, it merely denies them the power to waive their exclusion rights conditional on the otherwise-excluded party performing some future action. Thus, in both the boxing and kissing cases presented just prior, the person would be able to waive her claims against bodily contact. Granted, there would still be some cases where this proposal would not allow for conditional waiving of the kind favored by libertarian-minded people. For example, sex work, hired surrogacy, and the purchase of kidneys or blood plasma would violate the self-ownership rights of the persons whose bodies are being used – and this would be true irrespective of whether they agreed to sell their bodies in this way. However, this is a relatively small theoretical cost to pay for rescuing the self-ownership from the consent theorist’s argument against property ownership.Footnote ¹⁷

The final proposed way to make ASO compatible with consent theory is to weaken the latter such that P can acquire the power of territorial legitimacy – or, more precisely, self-sovereignty – vis-à-vis Q if and only if (a) P acquires consent from Q or (b) she would not be responsible for leaving Q worse off_FC regardless of how P exercised her acquired power. This added disjunct (b) would then be satisfied by self-appropriation for reasons similar to why self-appropriation satisfies the Lockean proviso (as discussed in Section 3.2). To see this, consider a case where Q’s ϕ-ing would make contact with P’s body and thereby generate x units of advantage for Q without uniquely generating any supplemental benefit. Further, suppose that P exercises her power of self-sovereignty and issues the edict that, if Q ϕ-s, she must also ψ, where ψ-ing imposes some cost of y on Q. Given that P’s established claim is conditional, Q’s choice to either ϕ or not ϕ will determine what her future full compliance involves, and, thus, what costs_FC she incurs. In the full-compliance world where Q declines to ϕ, the cost_FC she incurs is simply x, as she foregoes_FC the advantage she would have had if she had made the specified contact with P’s body. By contrast, the full-compliance world where Q ϕ-s is one where Q also ψ-s, with Q thereby avoiding the cost of x but incurring the cost of y.

Why, then, is P not responsible for leaving Q worse off_FC when she issues her edict? Recall from Section 3.4 that P is responsible for imposing some cost_FC on Q only if Q’s incurred cost_FC survives the nonexistence comparison – that is, Q would not suffer that cost in the closest possible full-compliance world where P does not exist. Further, note that the cost_FC that Q would incur if P had never existed is equal to x, as P’s nonexistence would preclude Q from obtaining the x units of advantage that she would otherwise acquire by using P’s body. Thus, the costs_FC of Q not ϕ-ing are equal to the costs_FC she incurs in the world where P does not exist – that is, the former costs_FC do not survive the nonexistence comparison. This, in turn, implies that P is not responsible for Q incurring a cost_FC of x when she does not ϕ.

What if a fully compliant Q does ϕ? Given that her ϕ-ing entails that she will then ψ and incur a cost_FC of y, there are two possibilities: either she ends up no worse off_FC than if she did not ϕ (because x ≥ y) or she ends up worse off_FC (because x < y). If ϕ-ing leaves her no worse off_FC, then P is not responsible for leaving her worse off_FC, as she is no worse off_FC in the world where she ϕ-s than the world where P does not exist. For, as noted in the previous paragraph, Q is no worse off_FC in the world where she does not ϕ than she is in the nonexistence world; thus, if she is no worse off_FC in the world where she ϕ-s than she is in the world where she does not ϕ, then, transitively, she is no worse off_FC in the world where she ϕ-s than she is in the world where P does not exist.

Suppose, instead, that Q is worse off_FC if she ϕ-s relative to the world where she does not due to x being less than y. In this case, Q may end up worse off_FC than she would be in the world where P does not exist. However, recall from Section 3.4 that an additional necessary condition of P being responsible for leaving others worse off_FC is P having unilaterally left them worse off_FC. In other words, any avoidable costs_FC that Q incurs due to her subsequent choices do not count when assessing whether P is responsible for leaving her worse off_FC (i.e., whether Condition (b) is satisfied). Given that Q could avoid the supplemental costs_FC of ϕ-ing by not ϕ-ing, it follows that P is not responsible for these costs_FC. Thus, irrespective of which choice Q makes, P is not responsible for her incurred costs_FC, that is, P is not responsible for leaving Q worse off_FC. Further, given that P’s edict was described generically, it follows that a self-sovereign P will never be responsible for leaving Q worse off_FC by issuing an edict, with self-sovereignty thereby necessarily satisfying Condition (b) of the revised consent theory of self-sovereignty. Thus, self-sovereignty is compatible with the previously proposed revised consent theory.

There are two objections to this proposal that are worth addressing. The first worry is that the move to append Condition (b) to standard consent theory is ad hoc and, more problematically, in tension with the motivations for adopting consent theory in the first place. Here, one might reply by appealing to the claim in Chapter 2 about the relationship between consent theory and the moral tyranny constraint. Specifically, Section 2.4 suggested that the reason for endorsing consent theory is that legitimacy satisfies the moral tyranny constraint only if it has consent as its necessary condition. Because legitimate authorities are able to foreseeably and discretionarily leave others worse off_FC via the issuing of edicts, they must be denied the ability to unilaterally do so; that is, consent must be a necessary condition of this power to oblige. However, Section 3.4 argued that the moral tyranny constraint should be generalized such that it condemns moral theories that allow people to leave others worse off_FC while being responsible for leaving others worse off_FC in this way. And it has now been shown that self-sovereign agents are never responsible for leaving others worse off_FC in this way. Thus, there is no moral tyranny-related reason for insisting that consent is always a necessary condition of self-sovereignty and territorial legitimacy more generally. Rather, Condition (b) is an appropriate addition to standard consent theory, as, like Condition (a), it allows for (territorial) legitimacy so long as that power does not violate the moral tyranny constraint.

The second objection is that this amendment effectively reduces consent theory to the Lockean proviso. Once one makes non-worsening_FC a disjunctive sufficient condition of legitimacy (and self-sovereignty) alongside consent, the resultant principle starts to closely resemble the proviso with its contention that an act of appropriation establishes (self-)ownership rights if and only if the established claims and the possible subsequent waiving of those claims would not leave others worse off_FC. After all, both consent theory (in its revised form) and the proviso are satisfied if the established rights of exclusion – when paired with the power to waive those rights – do not leave others worse off_FC. Thus, one might worry that this third compatibilist proposal renders the argument of this chapter redundant (even if it does allow for self-appropriation).

Two things can be said in response to this objection. First, one might reply that even if the Lockean proviso and the proposed revision of consent theory do not differ importantly in their content, this does not diminish the dialectical significance of the foregoing argument. Even if one ends up accepting a consent theory that is quite similar to the Lockean proviso, one does so because standard consent theory threatens to negate the existence of all private property including the modest self-ownership rights posited by ASO. In other words, the proposed revision represents the conclusion of an independent argument against private property that has the standard consent theory of legitimacy as its starting premise. This is of particular dialectical significance given that those who favor private property seemingly need to endorse a consent theory of legitimacy for the reasons discussed in Section 4.5: Absent such a theory, people are liable to have various permissions to use their owned property stripped from them by state regulation. Thus, one can understand the foregoing argument as a way of pressuring both consent theorists and property rights proponents into first giving up private property and then accepting a revised version of consent theory that approximates the Lockean proviso.

Second, note that there is an important difference between the compatibilist revision of consent theory and the Lockean proviso, namely, that the latter does not make consent a sufficient condition of appropriation. However, the foregoing discussion also suggests that this is something of an oversight and that the proviso should really take a disjunctive form with both consent and non-worsening_FC serving as individually sufficient and disjunctively necessary conditions of appropriation. Thus, one might take the proposed compatibilist position to represent a synthesis of the proviso and standard consent theory. Specifically, each principle can be understood as an attempt to address initial appropriation’s moral tyranny problem, with the proviso aiming to preclude culpable worsening_FC while consent theory aims to preclude any sort of unilateral worsening_FC. Yet, when considered in sequence, each principle appears to be overly stringent. The proviso is too stringent because it precludes consensual appropriation that leaves the consenter worse off_FC – an outcome that is unproblematic vis-à-vis the moral tyranny constraint. And standard consent theory is too stringent because it demands consent even when a successful appropriator would be unable to leave anyone worse off_FC irrespective of how she exercised her established powers. By contrast, the posited compatibilist version of consent theory allows for appropriation given either consent or non-worsening_FC, thereby correcting for the respective overreaches of the proviso and standard consent theory. It is, thus, importantly distinct from the Lockean proviso (in addition to the fact that it is derived from standard consent theory).

4.9 Philosophical Anarchism and the Anarchist Conclusion

Before concluding, it is worth drawing attention to how the argument of this chapter – or, more precisely, the thesis that ownership entails territorial legitimacy – helps to contextualize what Section 1.6 called the anarchist conclusion. When it comes to discussions of state legitimacy, consent theorists are philosophical anarchists who insist that there are no existing legitimate states (while allowing that there could be a legitimate state under the appropriate conditions).Footnote ¹⁸ However, this position raises a natural objection, namely, that it seems to problematically imply that there is no obligation to comply with just laws. While much legislation is immoral, states also pass laws mandating that persons act in ways that morality seems to require, for example, laws that forbid murder. Thus, upon first encountering philosophical anarchism, many worry that it negates obligations to comply with these laws – that is, obligations to act in the way that morality requires – or to cooperate with state authorities who are enforcing these laws.

In reply to this worry, philosophical anarchists typically note that there are many independent grounds for obligations aside from the edicts of the legitimate state. For example, Simmons (Reference Simmons1999) draws a distinction between justification and legitimacy. On this account, a legitimate state is one that has the exclusive right to impose novel obligations upon its subjects as a result of some sort of special relation that they stand in with respect to one another (746, 752).Footnote ¹⁹ By contrast, a justified state is one that a person has reason not to undermine (and perhaps even reason to support) in virtue of its moral quality (753). Additionally, Simmons introduces a notion that might be called justified action. The idea here is that the moral character of some specific action carried out by a state gives others reason to not interfere (and possibly assist) with that action. Thus, even if a state is neither legitimate nor justified, one might be morally required not to interfere with some moral action such as its deployment of police officers to prevent violence (770).

In other words, the philosophical anarchist reply to the aforementioned worry is that, even if a state is not legitimate, individuals might still have certain obligations with respect to that state and its actors. Similarly, the anarchist who rejects private property can apply this rejoinder to worries about what obligations people have vis-à-vis natural resources in a world without private property. Specifically, even if the anarchist position denies individuals the territorial legitimacy entailed by property rights, it can still assign them distributive rights that determine which uses of natural resources are permissible. Thus, the anarchist conclusion can be understood as an application of Simmons’ philosophical anarchism to the special case of property ownership: Given the absence of legitimate property owners, it appeals to other moral bases to posit a set of distributive obligations vis-à-vis the use of land and resources.

4.10 Conclusion

Thus concludes the anarchist argument against private property. The contention has been that two distinctively libertarian theses – namely, the Lockean proviso and the consent theory of legitimacy – independently entail that persons lack any sort of private property rights over natural resources. Social anarchism can, thus, be understood as an inversion of standard libertarianism: While it starts out with core libertarian premises, it arrives at the opposite conclusion regarding the existence of private property. However, this result does not establish that one ought to instead, accept the egalitarian anarchist conclusion introduced in Section 1.6. Defending this claim will be the task of Chapter 5.