Legal Obligation, Criminal Wrongdoing, and Necessity

Abstract Individuals sometimes do things that they know will violate the terms of a statute. Most scholars deny that such actions are always morally wrong, but a coherent theoretical account of the relationships between 1) moral obligation, 2) legal obligation, and 3) criminal wrongdoing that can robustly classify hard cases has been elusive. This article starts with a Kantian account of the relationship between law and morality, and it proposes two closely related standards: one for legal obligation, and another for criminal wrongdoing. It then tests the plausibility and resilience of these standards by using them to generate illuminating new analyses of classic hypothetical cases involving alleged crimes committed under circumstances of necessity. These analyses offer reason to believe that the standards proposed in this article can anchor a Kantian theory of criminal responsibility that is simultaneously rigorous and humane.


Introduction
Over half a century ago, H.L.A. Hart identified three central yet unresolved questions in legal philosophy: "How does law differ from and how is it related to orders backed by threats? How does legal obligation differ from, and how is it related to, moral obligation? What are rules and to what extent is law an affair of rules?" 1 Hart identified important shortcomings in theories advanced by Austin, Kelsen, and others-for example, he was right to point out that any correct legal theory must be able to distinguish between law and private coercion 2and his alternative approach catalyzed a larger movement. But 20th century positivism has since proved itself vulnerable to the intractable questions Hart undertook to sidestep or resolve. 3 Hart conceived of legal obligation as a sui generis form of obligation, irreducible to moral or prudential parts. But when subsequent scholars have tried to enactments by legitimate states count as laws, and he therefore characterizes Kant's philosophy as authoritarian. 12 By contrast, Robert P. Wolff concludes that statutes can obligate us as laws only if they accord with the judgments of individuals upon whom they are imposed-an account in tension with Kant's examples, and one that seems to undermine the law's authority to resolve disagreements about rights. 13 Some scholars seek middle ground by sanding down the uncompromising angles of Kant's moral concepts in the legal domain. Thomas Hill, for example, argues that our moral obligations to obey juridical laws need not have the same "content and stringency" as the laws themselves, that juridical laws merely have "weight" in moral deliberations, and that they may be morally disobeyed if an agent does so respectfully. 14 But this account cannot be reconciled with the basic features of Kant's conceptual apparatus. First, morality is not an external perspective from which we can weigh juridical laws. Juridical laws are-by definitionmoral laws. 15 Moreover, our duties under juridical laws are 'perfect duties', which identify specific acts or omissions as unconditionally obligatory. 16 By contrast, ethical laws unconditionally require us to have certain ends-our own moral perfection and the happiness of others-but we must pursue those ends only through actions that do not violate our perfect duties, including those under juridical laws. 17 Fidelity to Kant's basic conceptual architecture thus requires roughly the opposite of Hill's approach.
Without the shared concepts that unite the ethical and legal domains, Kantian moral philosophy would have no more structural integrity than an engine built out of parts calibrated to two different systems of measurement. Therefore, to preserve the possibility of a truly unified theory of law and ethics, this article embraces the unconditionality of Kant's basic moral concepts. It then goes on to prove that they do not preclude nuanced and plausible resolutions in hard cases. 95:1 Archiv für Geschichte der Philosophie 36 at 45. 18. As such, it is part of a growing Kantian literature devoted to plotting a course between the twin threats that Jacob Weinrib has dubbed "anarchism and quietism" without sacrificing conceptual rigor, and with it the unique promise of Kantian theory when it comes to the coherent integration of our public and private duties. Weinrib, supra note 11 at 22. Charles Fried, for example, recently argued that judicial review transmutes the public use of reason Kant thought essential into boundaries within which a state's authority remains supreme. See Charles Fried, "Defining and Constraining the Sovereign: 'The Most Difficult of All Tasks'" in Finkelstein & Skerker, supra note 11 at 81. Robert Alexy has lately characterised Kant as an inclusive legal non-positivist for whom "extreme injustice is no law." Robert Alexy, A novel interpretation of Kant's theory of legal obligation is at the heart of this article. While Kantians broadly agree that individuals are legally bound by the exercise of the state's legislative authority, fewer have addressed the question of what exactly constitutes an exercise of that authority. I will aim to show that the legislative authority is the authority to create new laws-categorical imperatives-and that this is not the same thing as the authority to obligate us with words alone. In most cases, lawmakers can create new laws by linking their legislative commands to threatened punishments for disobedience. Such legislation presents individuals with two alternatives: freedom under laws and the loss of freedom. Faced with these alternatives, individuals are rationally required to preserve their freedom regardless of their personal preferences. I will aim to show that only my proposed account of legal obligation can ensure that we are all equally bound by laws. It also has a fruitful further implication: legislation fails to establish new legal duties in circumstances in which freedom under laws is not an available alternative for an agent. Part I of this article outlines two foundational ideas in Kant's legal philosophy -freedom and a civil condition-and briefly describes the way in which they justify and limit the state's authority to establish new legal duties. Part II proposes a Kantian standard for legal obligation: lawmakers can impose new legal duties on us only by providing a universally dispositive reason to obey. Part III proposes a Kantian standard for criminal wrongdoing that operationalizes the intuitive Kantian idea that criminals choose to exempt themselves from the general requirements of public law, and then illustrates its application to simple cases. Finally, Part IV applies these proposed standards to a pair of classic hypothetical cases involving alleged crimes committed under conditions of necessity. The nuanced and plausible new analyses that result from this exercise offer good reason to believe that the pair of standards proposed in this article can anchor a Kantian theory of criminal law that is simultaneously rigorous and humane.

I. Freedom and a Civil condition
Two key ideas in Kant's legal philosophy underpin the argument of this article: 1) freedom, and 2) a civil condition. Freedom is "independence from being constrained by another's choice." 19 As Arthur Ripstein has explained, this conception of freedom differs from those according to which individuals are freer just in case they can do a larger number of things, or pursue their happiness more effectively. 20 Instead, Kantian freedom is fundamentally interpersonal: a person is free just in "Kant's Non-Positivistic Concept of Law" (2019) 24:4 Kantian Review 497 at 504. Arthur Ripstein has developed a sweeping and mostly persuasive account of private law that successfully integrates Kant's a priori legal principles with the authority of lawmakers and courts to make binding empirical judgments. See Arthur Ripstein, Private Wrongs (Harvard University Press, 2016). 19. Kant, supra note 8 at 393 (6:237). Unless otherwise specified, I will always use the word 'freedom' to refer to external freedom rather than to the complementary Kantian concept of internal freedom of the will. 20. See Ripstein, supra note 7 at 32-33. case no other person or group can use that person's body or any other resources that belong to that person for another's purposes. 21 The premise of Kant's argument for state authority is that we have a pair of natural duties in a state of nature. First, we each have an internal duty (that is, a duty that we owe to ourselves) of "rightful honor": we must not allow others to use us as mere means to their ends. 22 In the context of Kant's legal philosophy, this means that we have a duty not to permit others to deprive us of our freedom. We likewise have an external duty (owed to others) not to wrong others by depriving them of their freedom. 23 Taken together, these two duties amount to a universal "innate right" of freedom, to the extent that it can coexist with the equal freedom of others. 24 Because freedom is understood as independence, our innate right of freedom amounts to a Hohfeldian right that others refrain from engaging in actions that 'coerce' us, meaning actions that override our independence with respect to our bodies or anything else that belongs to us. We also have a strictly correlative duty not to engage in actions that violate the innate right of any other person. Unlike specific property rights or contractual expectations, which we must choose to acquire, we have our innate right of freedom simply because we are human. 25 Kant then writes that our innate right of freedom "already involves the following authorisations, which are not really distinct from it (as if they were members of the division of some higher concept of a right)": 26 1) equality (independence from non-reciprocal obligations); 2) independence (being one's 'own master' so long as one does not wrong others); 3) association (interacting with others so long as others are not thereby wronged). 27 These "authorisations" can likewise be redescribed in terms of Hohfeld's fundamental legal concepts. 28  understood as a Hohfeldian immunity: we are each immune to attempted exercises of power that would impose non-reciprocal obligations on us, and we each have a correlative disability with respect to imposing such non-reciprocal obligations. Importantly, this immunity is (like all aspects of our innate right of freedom) inalienable: my duty of rightful honor makes me legally incapable of "throw[ing] myself away" by choosing to take on new legal duties that do not equally bind others. 30 In a state of nature, our innate right can extend only as far as our own bodies. For example, individuals in a state of nature have a correlative duty not to wrest an object from a non-threatening person's grasp or displace that person, since these acts would violate another's innate right. 31 But although we each have a right to be free of bodily coercion in a state of nature, we are necessarily insecure in our enjoyment of it, since we have no assurance that others nearby will not choose to coerce us. Perhaps worse, in a state of nature our mere proximity to others can wrong them by physically threatening them in a way that is inconsistent with their innate right, which will at best limit our domain of liberty with respect to association. 32 While John Locke thought the state of nature was merely inconvenient because rights were difficult to defend, Kant argues that the very features of the concept of a right make it the case that acquired rights of property, contract, and status are impossible without a state. First, a person can acquire a right to an external thing only by making a choice, and our innate equality disables any person from unilaterally imposing new correlative duties on others. 33 Second, disputes about the boundaries of acquired rights necessarily arise, and in a state of nature they cannot be resolved because no one is obligated to defer to anyone else's judgment. 34 Third, we lack equal assurance that others will respect our rights in a state of nature, and we are not truly independent of the choices of others-and therefore have no acquired rights, properly speaking-if our enjoyment of our putative rights depends on others' discretionary self-restraint.
Because individuals in a state of nature therefore cannot acquire property rights, they are unable to rightfully pursue many projects that they could physically pursue. A farmer who attempts to grow a crop of corn on a field, for example, is not independent of the choices of others in this pursuit, because the farmer's neighbors remain at liberty to uproot the growing crop. Our incapacity to constrain others from interfering with projects that involve large external objects or long periods of time effectively "put[s] usable objects beyond any possibility of being used." 35 For this reason, and also because our innate right to be free of bodily coercion is otherwise insecure, Kant argues that we have a duty to 30

442
Newhouse ourselves and to those around us to become members of a state. 36 Only as members of a state is it possible for us to enjoy our equal freedom to rightfully pursue the range of projects that we are capable of pursuing without wronging others. A state can create the necessary conditions for our enjoyment of equal freedom by exercising its "three authorities"-legislative, executive, and judicialeach of which are logically necessary to eliminate one of the three barriers to acquiring rights in a state of nature that Kant earlier identified. 37 The legislative authority is the capacity to enact laws that empower individuals to acquire property, contract, and status rights under uniform rules, as well as those laws necessary to establish and maintain state institutions. The judicial authority is the capacity to exercise judgment to authoritatively resolve disputes about rights. 38 The executive authority is the capacity to coerce us in accordance with the law, thus providing us with independent assurance that our rights are secure, and that we are all equally bound by the laws. 39 The state's three authorities, together with the innate right of freedom and certain fundamental principles of acquired right, constitute the idea of a civil condition. 40 Because the state's authorities are necessary prerequisites to our enjoyment of our rights, Kant's argument goes, we must see ourselves as having authorised the state to exercise these authorities by means of the "omnilateral" will-a constructive group agent comprised of the united wills of those individuals subject to the same laws in virtue of their identical purpose of enjoying their rightful freedom. 41 In a civil condition, we can see ourselves as free under neutral laws that reciprocally constrain the sets of actions that we are each at liberty to undertake, because we are rationally required to endorse the state's authority to promulgate, apply, and enforce those laws. 42 A state enables individuals to extend their independence to include acquired rights, such as property, but so long as we do not wrong others, no state has the authority to strip us of our original innate right of freedom, including its authorisations: equality under law, independence, and association. This is the case because the innate right of freedom justifies state authority in the first place, including the authority to create a system of legal property rights. 43 The foundational status of the innate right of freedom relative to acquired rights plays an important part in the arguments that follow, because it means that the "burden of proof" concerning the existence of an acquired right falls on the legal claimant who asserts its existence. 44  exists can "appeal methodically to [the] innate right to freedom" to defend against a charge of wrongdoing until the claimant demonstrates that the acquired right in question has indeed been established by law. 45

II. A Standard for Legal Obligation
In Kantian terms, a crime is by definition a knowing transgression of a duty. 46 In order to evaluate potential examples of criminal wrongdoing, therefore, we must first identify an agent's relevant duties under juridical laws. Kant argued that lawmakers can create new legal duties for us only by supplying the following two ingredients: 1) a legislative command, and 2) an external incentive to obey that command in the form of a coercive response to disobedience. 47 An external incentive is necessary because it provides us with "equal assurance that [others] will observe the same restraint" that we observe when we obey. 48 Without such assurance, our duty of rightful honor would prevent us from acquiring a new legal duty that others might freely choose to disregard.
The necessary and sufficient conditions of equal assurance remain a matter of debate, but Arthur Ripstein's account is particularly influential. Ripstein acknowledges that assurance is an essential prerequisite to legal obligation in both public and private law. In the domain of private right, a lack of assurance entails that "nobody is under any obligations with respect to external objects," and therefore "neither [party to a property dispute] has any rights, properly speaking." 49 In the context of public right, "a parallel point applies": assurance is necessary to make an obligatory public law "effective." 50 I will summarize his account of how the law provides this assurance in both domains and then explain how my own proposal improves on it. I will aim to show, first, that Ripstein is misguided to look to civil remedies to provide the equal assurance necessary to establish property rights. Moreover, although Ripstein later correctly concludes that criminal penalties supply the essential form of assurance necessary to create new legal duties of all types, he fails to identify a key feature of such penalties that makes this possible: their inconsistency with our innate freedom. Without this key insight, Ripstein's account of equal assurance cannot enable us to accurately identify our legal duties.
Ripstein argues that there are two dimensions along which civil remedies provide the assurance necessary to create a duty to refrain from encroaching on the acquired rights of others. First, these remedies assure individuals that if they become victims of wrongdoing, their resources will be restored to them, as nearly as possible. Second, these remedies may discourage potential wrongdoers by 45  Newhouse ensuring that any rights violation "carries potential disadvantages." 51 Ripstein thinks that this deterrent effect is "secondary, but supports the assurance provided by the remedy itself." 52 Deterrence is deemed secondary because "the remedial aspect of [civil] enforcement gives you all the assurance you need: you have what is yours, because if another wrongs you, you will be able to get it back." 53 Ripstein's later remarks on criminal punishment seem in tension with his earlier conclusion that civil remedies provide all the assurance we need. Because a wronged party can choose whether to "stand on his or her own rights," Ripstein concludes that civil liability is not "systematically" imposed. 54 He infers from this fact that criminal penalties are a necessary form of assurance for acquired rights after all, because the state is empowered to impose them on all knowing wrongdoers regardless of the private choices of victims. Only this systematic assurance, he concludes, can satisfy our duty of rightful honor and therefore conclusively establish acquired rights. 55 Notably, however, Ripstein does not suggest that these systematically threatened punishments provide anyone, much less everyone, with a dispositive reason to obey the law. Instead, they merely "compete with the criminal's other incentives," on his view, and presumably are sometimes outweighed by them. 56 I believe that neither of these arguments accurately describes the 'equal assurance' that lawmakers must supply in order to create new legal duties. Ripstein is correct to view civil remedies as essential logical extensions of acquired rights in cases in which they are violated. He is also correct to characterise deterrence as a 'secondary' effect of these remedies. But this is not because deterrence is a merely secondary form of assurance. As an analysis of Kant's equal assurance requirement, this discussion grasps the wrong end of the stick: equal assurance addresses us in our capacity as potential wrongdoers, not as potential victims. As potential wrongdoers, we have a duty to refrain from encroaching on the property of others only because we are assured that all other potential wrongdoers are equally constrained. This is not at all the same thing as assuring us, in our capacity as potential victims, that our resources will subsequently be restored to us if we are wronged. Civil remedies have an important legal function: they preserve private rights over time by creating legal powers and liabilities for parties to a rights violation that has already occurred. But such remedies have neither the purpose nor the effect of supplying the kind of equal assurance that creates new legal duties.
Ripstein later rightly concludes that threatened criminal punishments supply the critical form of assurance, and when he describes them as "systematic," he identifies one of the key features that enable punishments to serve this function. 57 However, when Ripstein suggests that these threats merely "compete with the criminal's other incentives," he gives the impression that it is the discomfort or inconvenience occasioned by criminal penalties that constitutes the necessary equal assurance. 58 If that is indeed what Ripstein means, it would be a problem for his account, because it is the assurance, and not the punishment itself, that must be "equal" if any new legal duties are to be created. Any theory of assurance that relies only on those features of punishment that might be outweighed by other considerations cannot explain how such assurance can be equal, because no incentive that depends on personal preferences for its effectiveness can motivate all individuals to obey, even if it is systematically provided. 59 By contrast, I will aim to show that the state can provide equal assurance by threatening to impose criminal punishments that deprive us of our bodily independence, because bodily independence is an element of our innate freedom. 60 Recall from Part I that a civil condition is a system of equal freedom under law, and freedom is independence from being subject to the choices of others. Such a system cannot function over time unless it distinguishes between objects of choice-things that are external to us and can therefore be freely acquired and discarded-and the bodily sovereignty that enables individuals to acquire external rights in the first place. Individuals can rationally give up property, contract, or status rights without compromising their innate freedom, but the duty of rightful honor disables any person from legally choosing to surrender bodily independence, since that would ipso facto make that person a mere means for others. 61 The inalienability of our innate freedom justifies many existing legal doctrines according to which individuals are legally disabled from consenting to certain kinds of treatment, such as destructive physical violence. 62 In contract law, terms that purport to limit one party's future bodily independence, such as those for slavery or indentured servitude, are legal nullities because they are inconsistent with our innate freedom. 63 Courts decline to order specific performance of more 58. Ripstein, supra note 7 at 319. 59. On persuasive account of the nature and scope of Kantian transcendental freedom, we lack the ability to freely choose an end toward which we are weakly inclined over an alternative toward which we are more inclined on any non-moral basis. But although our innate freedom is inalienable, our legal right to it-which Kant refers to as our "dignity [as] a citizen"-is not absolute. 65 We can never throw away our innate freedom by signing a contract, but we will be stripped of it, at least temporarily, if we commit a crime. 66 Convicts are transformed into "mere tool[s] of another's choice" by punishments that are inconsistent with bodily independence. 67 When Kant writes that "it is impossible to will to be punished," he means that a criminal punishment is by definition the kind of treatment to which an individual could not consent because it is inconsistent with innate freedom. 68 Ripstein aptly describes punishment as "exclusion from the system of freedom," but he never identifies the key role that this insight must play in a successful account of equal assurance. 69 Our duty of rightful honor obligates us to preserve our innate freedom at the cost of any personal preference or private purpose that we may have. 70 We are therefore each morally obligated to avoid a punishment that is inconsistent with our external freedom if, by obeying the law, we can continue to see ourselves as free under laws that we have given to ourselves. Because people are often irrational, no law can empirically guarantee universal obedience. But for assurance to be 'equal' in any sense at all, lawmakers must at least give everyone a dispositive reason to obeya reason that does not depend on our contingent personal preferences or purposes.
Criminal punishments can accomplish this goal, because rational agents who are presented with the alternatives of freedom, on one hand, and the loss of freedom, on the other, have a duty to preserve their freedom. My proposed standard for legal obligation is therefore as follows: The legislative authority creates a legal duty to obey its command by presenting individuals with two alternatives: 1) Obey the legislative command and continue to see yourself as free under laws you have given yourself by means of the omnilateral will; or 2) Disobey and become liable for a punishment that is inconsistent with your innate freedom. 71 When this standard is met, individuals have a duty to obey the legislative command even when the chances of arrest and prosecution are small, because lawbreakers surrender their right to be "beyond reproach," meaning that their innate freedom is no longer legally protected. 72 This proposed interpretation can make sense of Kant's claim that "[t]he law of punishment is a categorical imperative"-in other words, that state officials are legally obligated to deprive criminals of some aspect of their innate freedom. 73 Why did Kant think that state officials have a duty, rather than just a Hohfeldian power, to punish crime? I believe the correct answer has two parts: 1) the state must have previously specified a punishment of this type in order to supply the equal assurance necessary to create the very legal duty that the criminal knowingly transgressed; and 2) because our innate equality requires punishments to be systematically imposed, the same law that obligates citizens to obey ipso facto obligates state officials to punish those who disobey. 74 If we want to know whether a statute has created a legal duty, on this view, we must initially determine whether its prescribed penalty is a consequence that we could rationally choose, "for it is no punishment if what is done to someone is something he wills." 75 Some criminal punishments that meet this standard, such as the death penalty and lengthy incarceration, are very harmful to those on whom they are imposed, but less harmful punishments exist that are inconsistent with our bodily independence. Punishments such as house arrest, curfews, community service work, and/or mandatory in-person instruction are more appropriate punishments for less serious forms of criminal wrongdoing. 76 As legally specified responses to violations of the terms of statutes, all of these punishments are adequate to create legal duties, because they all conflict with our bodily independence and therefore with our innate freedom. 77 As Ripstein has correctly observed, "Kant cannot accept the idea that the criminal law is a series of [contractual] offers, because these are not offers that anyone could rightfully make." 78 To some readers, my proposed standard for legal obligation may appear objectionably Hobbesian. What about the distinctively Kantian ethical obligation to obey the law out of respect for law as such? 79  448 Newhouse obligation, and it is therefore not maximally virtuous to obey the law merely to preserve our own individual external freedom-even though preserving our freedom is itself an end that we are morally obligated to uphold. 80 But we can obey a legislative enactment out of respect for law as such only if that enactment qualifies as a law in the first place.
Recall that a law is by definition a categorical imperative: a command that unconditionally binds an agent. 81 This is true whether the law in question happens to be ethical or juridical in nature. Kantian 'respect' is, by definition, our subjective response when we apprehend a law, and it can move us to obey the law on ethical grounds alone. 82 We can therefore only respond with respect to something that already is a law. Any attempt to elevate a mere hypothetical imperative, such as a desire to avoid the pain and inconvenience of prison, into categorical status by means of the very respect that definitionally constitutes our response to a categorical imperative is akin to pulling oneself up by one's shoelaces. 83 Only if we are each morally-and thus equally-obligated to obey a legislative command in order to preserve our own individual freedom can we be universally legally obligated to obey it as a new juridical law. 84 Once a new law has been successfully created in this way, we each have an ethical duty to obey it out of respect for law as such.
My proposed standard for legal obligation is not inconsistent with the exclusivity of the state's legislative authority. Only the state can create new duties by threatening to punish disobedience, because its exclusive possession of the legislative authority is what enables us to see ourselves as free under laws that we give to ourselves. This freedom under laws is what we are rationally required to preserve through our obedience. By contrast, if a burglar threatens to lock me in a closet until I open my safe for them, there is no alternative open to me that will allow me to regard myself as free. The burglar is simply threatening to violate my independence in two qualitatively different ways, and a Hobson's choice of this sort cannot create a duty to obey them.
We each have a duty to obey exercises of the legislative authority, but the legislative authority is the authority to enact laws-categorical imperativesone at a time. It is not some generalised authority to bind us with words alone. If the legislative authority could bind us with words alone, then an 'external 80. Virtue requires us to do the right thing for the right reason. What I mean to say here is that an individual who obeys the law only to preserve that individuals' own freedom is not acting on the correct unitary moral reason: universal freedom, of which the individual's personal freedom is a wholly included part. incentive' might be a useful or prudent tool for motivating citizens to obey, but it is hard to see why it would be an essential prerequisite to legal obligation, as Kant explicitly insists that it is. 85 The foregoing argument resolves this textual mystery by explaining exactly how punishments can create new legal duties. It also helpfully clarifies what kind of response to wrongdoing is adequate to serve this function: one that is inconsistent with our innate freedom. 86 Most importantly, it ensures meaningful equality under law, which our duty of rightful honor demands.

III. A Standard for Criminal Wrongdoing
From a Kantian perspective, the distinction between civil and criminal wrongdoing arises from the fact that every external 87 action has a dual nature: a physical manifestation (the 'act') and its conjoined principle of action (the 'maxim'). 88 Arthur Ripstein has shown that an action is civilly wrong in virtue of a property of its act: the physical incompatibility of one person's conduct with the freedom of others. 89 Some agents who commit civil wrongs act on unobjectionable maxims, in which case those wrongs are "mere faults" that should not give rise to criminal liability. 90 Civil wrongs legally entitle wronged parties to remedies that restore their acquired rights to them insofar as that is possible. By contrast, a crime is a knowing transgression of a duty. 91 Because knowledge of an act's wrongful nature is what makes it criminal, an action can be considered criminally wrong only on the basis of a property of its maxim, the principle of action that an individual adopts when choosing to undertake it. 92 The criminal law targets maxims because its purpose is to provide equal assurance by constraining our conduct prospectively. 93 In order to function as a deterrent, it must prohibit conduct that individuals can identify in advance by evaluating the maxim on which they are contemplating acting. 94 Arthur Ripstein aptly describes a criminal act as one by which "the criminal seeks to exempt himself from the law" by violating the same binding rules of This standard asks whether two hypothetical states of affairs could coexist: 1) the legality of actions on A's maxim, and 2) everyone's freedom in a civil condition. 98 In other words: Could everyone possibly be free in a civil condition if actions on A's maxim were legal? If the answer is "no," then individuals who act on the maxim of A are seeking to exempt themselves from the necessary preconditions of equal freedom that they are rationally required to endorse for others. Any action on A's maxim is for this reason criminally wrong. Some readers may wonder why this standard compares the legality of actions on certain maxims, rather than the maxims themselves, to the concept of everyone's freedom in a civil condition. This question has three answers, one textual, and two philosophical. As an interpretation of the second prong of Kant's Universal Principle of Right, this standard accounts for Kant's stipulation that an action is right (meaning that it is not contrary to any legal duty) if "under its maxim the freedom of choice of each can coexist with everyone's freedom" in a civil condition. 99 In the context of an existing legal system, to say that an action is 'legal' just means that everyone is free to choose it. Philosophically, sticking closely to Kant's language turns out to be important for two reasons. First, it is metaphysically impossible for a maxim itself to be inconsistent with the external freedom of others. Freedom, understood as physical independence, can never be violated by an abstract principle. But, secondly, we cannot overcome this objection by assuming that an individual's external conduct will necessarily track the subjective principle on which they are acting in order to 95 evaluate that package as a whole. Maxims can include mistakes about the world, or can describe conduct that an agent tries, but fails, to undertake.
Suppose, for example, that an Envious Job Seeker acts on the maxim: "I will remove the sparkplugs from my more qualified neighbor's car in the dark of night, so that my neighbor cannot drive to the job interview tomorrow." 100 Suppose also that the Envious Job Seeker is disoriented in the dark and therefore mistakenly removes the sparkplugs from the envious jobseeker's own, very similar car instead. The Envious Job Seeker's external conduct can coexist with everyone's freedom-nothing that belongs to anyone else has been touchedand so can the Envious Job Seeker's maxim, since the maxim itself is just an abstract principle. However, the legality of actions on the Envious Job Seeker's maxim could not coexist with everyone's freedom in a civil condition: if theft were legal, no one would be capable of acquiring property rights. My proposed standard for criminal wrongdoing can therefore accurately identify attempted crimes as crimes themselves: actions by which wrongdoers seek to exempt themselves from the same legal duties that they are rationally required to endorse for others.
To apply this standard to cases, one must first formulate appropriate maxims. The criminal law punishes only actions involving external conduct (although its criminality will depend on an agent's mental state), so the maxims relevant to this inquiry are "very specific first-order principle[s] of volition and consequently action" 101 with the following structure: I will [act] when [circumstances] in order to [end]. 102 The ends specified in these maxims are what criminal law doctrines refer to as 'motives': desired states of affairs that move an agent to act. 103 Unlike intentions-results that an agent undertakes to cause in the doctrinal sense 104 motives as such have no impact on the legal status of an action. 105 Consider the example of a potential organ donor who refuses to donate a kidney to a nephew because the potential doner actively dislikes the nephew and wishes him dead. This is an evil motive for refusing, but a mere motive, even if it is the death of another, cannot transform a choice into a crime, any more than the possession of an admirable motive can transform a crime into a legal action. 106 Whether or not an action is criminally wrong will instead depend on the agent's intended physical conduct and any results the agent intends to bring about as well as certain relevant circumstances. For example, it is not criminally wrong for an assault victim to coercively defend against an attacker, because the attack is a circumstance that justifies defensive coercion. 107 This is true even if the assault victim is privately motivated by a desire to harm the attacker, so long as the victim's use of force objectively amounts to a proportionate defense. 108 To distinguish intentions from motives, intended results that are imputable to an agent by causation doctrines must be incorporated into a maxim's description of the agent's act. For example, both surgeons and murderers cut people with knives, but the surgeon's intended act can be accurately described as 'operating', while the murderer's intended act should be described as 'killing'. These examples show that a person who formulates maxims for analysis must do so in light of the content of the law, because maxims would be hopelessly complex if they included every detail of an agent's subjectively intended conduct and known circumstances. Only by consulting legal standards is it possible to generate succinct maxims by selecting the legally relevant features of an action. 109 After a maxim is formulated, it must be converted into a 'Legality Proposition', because my proposed standard for criminal wrongdoing compares the legality of actions on an agent's maxim, rather than the maxim itself, to the concept of everyone's freedom in a civil condition. Legality Propositions omit the ends specified in maxims, since these private motives are legally irrelevant (recall that results the agent intends to legally cause are folded into the act definition) and they also abstract from the specific identities of individuals referenced in maxims. Legality Propositions therefore take the following form: . 109. Joachim Hruschka characterizes the maxim-formulation process as an aspect of "first level imputation," which also includes a description of the agent's external act. Hruschka, supra note 88 at 682. This process must take place before any action can be legally evaluated even for civil wrongdoing, since it is only by reference to maxims that we can individuate an action in the context of an ongoing stream of activity by an agent. See Arthur Ripstein, supra note 7 at 381. This process does not render the subsequent application of legal standards trivial, because the status of an action as criminally wrong will often depend on a maxim's combination of legally relevant features.
In the context of a Legality Proposition, the claim that an action is 'legal' is a claim that it does not violate any legal duties, including our natural law duty not to violate others' innate right of freedom. Once formulated, a Legality Proposition must be compared to the concept of everyone's freedom in a civil condition. 110 A conflict between these two ideas shows that these states of affairs could not simultaneously obtain in the world and therefore indicates that actions on the underlying maxim are criminally wrong. For example, consider a Malicious Assailant who acts on the maxim: "I will hit my neighbor's child, who is playing on the sidewalk, with a rock in order to amuse myself." After eliminating the legally irrelevant motive, one can derive this corresponding Legality Proposition: It is legal to hit others with rocks when they play on the sidewalk. This proposition conflicts with the concept of everyone's freedom in a civil condition because that concept includes the innate right of freedom, which puts us all under a legal duty not to physically coerce others who are not themselves violating the rights of others. The Malicious Assailant's action is therefore criminally wrong, and the assailant can be justly punished if the relevant legal jurisdiction imposes a criminal penalty for battery. 111 Next, consider the case of a Regulatory Scofflaw, who knowingly violates the terms of a malum prohibitum penal statute. 112 The Scofflaw might adopt the following maxim: "I will open a restaurant when I have not obtained a licence and doing so will contravene a penal statute in order to make money." This maxim yields the following Legality Proposition: It is legal to open a restaurant when one lacks a licence and doing so contravenes the terms of a penal statute. 113 Comparing this proposition to the concept of everyone's freedom in a civil condition reveals no necessary conflict with any individual right, because restaurant licensure is just one possible strategy a state might adopt to minimise the risk of 110. Recall that a civil condition is a set of necessary laws: innate right, the a priori principles of private right, and the state's three authorities. The concept of a civil condition does not presuppose that individuals always act justly. 111. See for example NY Penal Law §120.00, 70.15. In jurisdictions in which criminal law is exclusively statutory, lawmakers have a duty to protect our innate right of freedom by means of a penal statute, but the test for criminal wrongdoing does not presuppose that an agent lives in a fully just legal order. Nothing in this article precludes the possibility of a system of common law enforcement of malum in se offenses, though I will generally discuss penal statutes rather than common law criminal doctrines. 112. A valid penal statute is a malum prohibitum law if its content does not track a natural law principle of private or public right. 113. This Legality Proposition may at first appear internally incoherent, but that is not the case. An action that contravenes a penal statute is not illegal (i.e., contrary to a legal duty) merely by definition or by logical implication. I will demonstrate later that penal statutes fail to create legal duties in cases involving necessity.
454 Newhouse food poisoning. 114 However, a conflict appears when we focus on the concept of the legislative authority, which is part of the larger idea of a civil condition. The legislative authority includes the capacity to enact positive laws, including penal laws, on the basis of lawmakers' empirical judgments about how to best secure conditions of equal freedom. 115 If it is logically possible that a malum prohibitum penal statute serves the state's purpose of securing everyone's freedom in a civil condition, then individuals must regard it as a product of the omnilateral will even if they personally consider it misguided. 116 In this example, it is logically possible that a restaurant licensing law prevents restauranteurs from wrongfully endangering their patrons' health. As we saw in Part II, illegality is necessarily the effect of a penal statute so long as agents who obey it can continue to regard themselves as free under laws. The Regulatory Scofflaw's Legality Proposition therefore conflicts with the possibility of everyone's freedom in a civil condition because it is inconsistent with the nature of the legislative authority, and the Scofflaw can be justly punished under an applicable law so long as the punishment prescribed is proportionate to the gravity of the offence. 117

IV. Application to Necessity Cases
Part II of this article argued that lawmakers can create new legal duties for us only by providing us with a dispositive reason to obey: one that does not depend on our personal preferences or purposes. The resulting standard for legal obligation can operate in combination with the standard for criminal wrongdoing from Part III to make sense of Kant's famous analysis of a shipwrecked sailor who commits a murder in circumstances of necessity. More importantly, these twin standards underpin an appealing new analysis of a more controversial class of necessity cases involving alleged violations of property rights.
114. For this reason, only knowing violations of malum prohibitum statutes can be shown to be criminally wrong. 115. In addition to penal laws, the legislative authority includes the capacity to enact "permissive laws," which empower us to alter our legal rights and obligations by choice. The argument that follows depends on an important distinction between an unconditional duty and a duty that applies in all circumstances. In Kantian terms, all duties created by statutes are "unconditional" in a specific sense: they apply to all individuals regardless of their personal preferences or purposes. 118 An unconditional duty may or may not apply to individuals in all circumstances, however. Sometimes, the same statute that creates a new legal duty specifies circumstances under which it does not apply. For example, U.S. law requires all individuals to file federal income tax returns every year, unless they have an annual gross income below the exemption amount, or unless they are in one of several other sets of circumstances stipulated by law. 119 The fact that individuals have different duties under tax law owing to their different circumstances is not inconsistent with our innate equality under law, because our external circumstances are not personal attributes. When legal duties apply only to those in certain external circumstances, we can-at least potentially-claim that the law applies equally 'to all those similarly situated' and can therefore bind them unconditionally.
I will aim to show that necessity cases are best understood as a class of cases in which individuals' external circumstances are such that penal statutes fail to present individuals with the usual alternative between freedom under laws and the loss of freedom. Because no statute can provide the equal assurance necessary to create a legal duty to obey a legislative command in circumstances of necessity, duties created by statutes do not extend to those circumstances. If necessity is properly understood as a circumstance-defined limitation on the scope of duties created by legislation, it is fully consistent with the law's essential impartiality, because it takes no account of individuals' preferences or purposes, or of any other personal attribute.
In Kant's most famous criminal law example, a Drowning Sailor bent on selfpreservation kills a non-threatening victim by pushing that victim off a floating plank. 120 The Drowning Sailor's maxim might be formulated as follows: "I will kill a non-threatening person when doing so is my only chance of survival in order to live." The associated Legality Proposition is: It is legal to kill a non-threatening person when doing so is the killer's only chance of survival, and when doing so contravenes the terms of a penal statute.
This proposition, like that of the Malicious Assailant, conflicts with the victim's innate right of freedom. In Kant's words, the doctrine of right would "be in contradiction with itself" if our innate right had no correlative legal duty. 121 We therefore always have a natural law duty not to physically coerce non-threatening others, whether or not this conduct has been forbidden by statute. The Drowning Sailor's natural law duty should have inspired respect for law, and that respect could have moved the Drowning Sailor to refrain from murdering the victim, Kant nonetheless concludes that no penal statute can empower the state to punish the Drowning Sailor, because "a penal law of this type could not have the effect intended." 122 If the argument of Part II is correct, then the "effect intended" by a penal statute is the creation of a legal duty to obey a legislative command, and lawmakers can only accomplish this by supplying equal assurance: an external incentive that creates a dispositive reason to obey. In most circumstances, lawmakers can supply equal assurance by threatening a punishment that is inconsistent with our innate freedom, which our duty of rightful honor obligates us to preserve at the cost of any amount of discomfort or heartbreak that our obedience might entail.
In the Drowning Sailor's circumstances, however, a threatened criminal penalty cannot set up an alternative between freedom under laws, on one hand, and the loss of freedom on the other. Instead, the relevant legislation in Kant's era would (if applicable to the sailor's case) have presented individuals with a different set of alternatives: death by drowning, on one hand, and the death penalty, on the other. The death penalty is the most complete deprivation of freedom that an individual can suffer at the hands of the state. The question, then, is whether the sailor's alternative, death by drowning, is one that they are rationally required to prefer. Kant himself answers this question in the negative, but why would that be the case?
I think the answer lies in the fact that the freedom that we can usually preserve by avoiding a threatened criminal punishment is specifically our external freedom: our physical independence. The Drowning Sailor would have no remaining physical independence after drowning at sea, and likewise no remaining physical independence after being put to death by the state. The Drowning Sailor's good moral character should be worth more to them than life itself. But no external incentive of the type that the state can provide can generate a dispositive reason for them to obey a statutory command against murder, because with respect to external freedom, the outcome is identical under the alternatives that such a statute could generate.
A general penal law against murder therefore does not apply to murders committed in circumstances of necessity, whether this limitation is reflected in the language of the statute or not. Because the statute does not apply in such cases, it cannot empower the state to punish the Drowning Sailor. Such a circumstancedefined limitation on a statute's scope of application does not undermine the essential impartiality of the penal law: it applies equally to all similarly situated individuals regardless of their preferences or purposes. Indeed, a suicidal sociopath, who pushed the victim off the plank out of curiosity, would likewise be unpunishable so long as the sociopath understood the external circumstances to be such that the sociopath's only alternative was certain death.
Next, consider a more controversial case of necessity that arises in the context of an acquired property right. Arthur Ripstein describes a hiker who breaks into an unoccupied cabin in the woods to escape a sudden storm in which the hiker would otherwise perish. As a threshold matter, Ripstein claims without analysis that the "hiker commits a wrong against the owner of the cabin: he uses another person's property without his consent." 123 Ripstein then considers the "harder question" of whether the hiker's conduct is a crime in the Kantian sense of being a knowing wrong. 124 He correctly explains that Kant's unconditional moral concepts preclude us from characterising the hiker's conduct as an example of justified coercion. 125 It would be incoherent to claim that the hiker has a 'right to life' that overrides (or outweighs) the cabin owner's property right, because Kantian rights are "rights to independence, not [rights] to existence," and our reciprocal rights and duties are by definition incapable of coming into conflict. 126 To say that someone has a legal right is just to say that others have an unconditional duty not to violate it. Coercion is only justified (that is, consistent with rights) insofar as it "hinder[s] : : : a hindrance to freedom," and in this case the cabin owner is doing nothing to impair the freedom of the hiker. 127 On the basis of these arguments, Ripstein concludes that the hiker is, like the Drowning Sailor, a criminal wrongdoer who escapes punishment only because the relevant penal statute could not guide the hiker's conduct prospectively. 128 In other words, Ripstein concludes that the hiker has a legal excuse for his crime. Unlike a justification, which "changes legal relationships between persons," he explains, "an excuse is personal to the person excused and cannot change the rights of others." 129 Although superficially plausible, Ripstein's conclusion has an unsettling implication: that a freezing hiker has a moral obligation to die rather than take shelter in someone else's cabin. This implication arises because our legal duties are moral duties: categorical imperatives, just like the ethical ones. 130 Results like this one are worth reconsidering. They make the most attractive feature of Kant's moral philosophy-its unity-look like a defect instead. I will aim to show that Ripstein reaches the wrong conclusion because he starts with a faulty premise. Ripstein is mistaken to assume without analysis that the Freezing Hiker's use of the cabin violated the cabin owner's property right, and this mistake ensnares him in a false dichotomy of justification or excuse. If the hiker's conduct does not Ripstein erred by conceptualizing necessity as a subsequent carve-out in pre-existing legal right. Necessity is more accurately understood as a domain of residual liberty that is shielded by our innate right to bodily independence. This residual domain of liberty has never been, and cannot be, eliminated by positive law, because lawmakers are incapable of supplying the equal assurance necessary to create new duties in circumstances of necessity. If this is correct, then property rights simply do not include the right to exclude uses by others in circumstances of necessity, the Freezing Hiker's Legality Proposition does not conflict with the concept of everyone's freedom in a civil condition, and the Hiker's conduct is not criminally wrong. 132 This intuitively sound conclusion is not inconsistent with any fundamental feature of property rights. The purpose of a system of property rights is to enable the rightful pursuit of projects by equally constraining our conduct in accordance with a set of uniform rules. It is important that these rules be neutral in the sense that they do not take any account of the "matter of choice"-the preferences or purposes of individuals-and that they can be applied by an independent judge to definitively resolve any property-related dispute. Understood as a circumstancedefined limitation on the possible extent of property rights, necessity constitutes an equal and reciprocal domain of residual liberty, and disputes about whether some particular action falls within that domain can be readily adjudicated by a neutral judge without reference to individual preferences or purposes. Suppose, for example, that the Freezing Hiker was suicidal, entered the cabin only to look at the pictures on the walls, and survived due to accidentally falling asleep there until the storm had passed. If the hiker understood the circumstances to be such that entering the cabin was necessary to survival, then doing so would fall within the ambit of our equal and reciprocal domain of residual liberty regardless of the hiker's private purposes.
Perhaps some property owners could rightfully pursue an even wider range of possible projects if the boundaries of their property interests were defined solely in spatial terms, but the above argument indicates that this is not a result that lawmakers could bring about, even if they tried. Moreover, it is implausible that 132. I do not address here the question of whether a scheme of financial reimbursement akin to the civil remedies available under current law is justifiable in such cases. I do not rule out the possibility that such a scheme could conform to the requirements of justice. My argument shows only that the claim that property uses in circumstances of necessity wrong property owners could not do any work in a successful justification for such a scheme. Note also: my analysis does not extend to a different set of cases that Ripstein considers elsewhere, in which the owner of a chattle makes unauthorized use of real property under circumstances when doing so is necessary to retrieve or retain possession of that chattle. See Ripstein, supra note 18 at 130-55. In such cases, the relevant rights of all parties are statutory rather than innate, and Ripstein's explanation of the 'partial privilege' rules adopted in those cases as essential to preserve the systematicity of rights in both realty and movables is plausible. 460 Newhouse the limits of property interests must be defined exclusively by metes and bounds. 133 In practice, these acquired rights often have temporal or circumstantial boundaries as well as spatial ones. For example, the owner of a fee simple timeshare interest in a condo has a right to exclude co-owners from the condo, but only at certain times each year. 134 The owner of a farm that is subject to an easement entitling the neighbors to traverse it to reach the local church cannot rightfully pursue projects that conflict with that use, regardless of what the neighbors' private purposes are for going to church. We do not usually think that non-spatial boundaries render property rights chimerical.
While it is true that property owners may not know the exact limits of their rights in advance of a legal judgment following a use for which necessity is claimed, such uncertainty is common in other property law contexts. For example, there may be doubt about the presence or absence of an easement until a judge has made an authoritative determination. In Kantian terms, a judge who determines that an easement exists has not concluded that one person's right overrides (or outweighs) that of another. Rather, the judge has concluded that the parcel owner's independence does not extend to uses that conflict with the easement holder's right. An easement holder who legally traverses a servient tenement is not engaged in justified coercion. Rather, this action is not coercive and therefore requires no justification.
If necessity is best understood as a residual domain of innate liberty rather than as a right in the Hohfeldian sense, a legal conclusion that the Freezing Hiker is innocent of wrongdoing would not, pace Ripstein, entail a correlative duty to rescue others in dire straits. 135 As a legal matter, a cabin owner who is at home in a locked cabin when the Freezing Hiker arrives at the door is at liberty to ignore the hiker's pleas for assistance. Should the hiker break into the occupied cabin in this revised hypothetical, the cabin owner will be entitled to respond coercively only if threatened with bodily harm by the hiker. While this might prove to be a factually complex legal case, it is imminently capable of authoritative judicial resolution by means of the application of neutral legal standards that impose no affirmative legal duties on the owner of the cabin in question.

Conclusion
This article has proposed a pair of normative standards: one for criminal wrongdoing and another for legal obligation. Both standards have significant textual support as interpretations of Kant, but the primary work of this article has been to demonstrate their philosophical value. This article has shown that the proposed standard for criminal wrongdoing is clear and resilient, and that recognising the 133. Metaphysically, a person is "an absolute unity." Kant supra note 8 at 427 (6:278). By contrast, objects in the natural world are individuated by means of practical reason and can therefore be divvied up for legal purposes in many ways. under-appreciated role of threatened punishments in the creation of legal obligations yields rigorous new analyses of a pair of classic necessity cases. To those who find themselves attracted to Kant's general vision of freedom in a civil condition, this article has also striven to prove that Kant's unconditional moral concepts do not invariably spell doom for broadly popular conclusions about justice in hard cases. Indeed, the standards proposed in this article appear well-suited to anchor a broader Kantian criminal law theory that is as nuanced and plausible as it is rigorous.