This guest post is adapted from an academic essay originally written as a thesis in Political Philosophy at the University of Edinburgh.
Hart famously broke away from Austin's teachings on jurisprudence in rejecting his account of law as “command backed by threats”, claiming it to be overly simplistic. Hart controversially claimed that law does not necessarily require sanctions, contrary to most previous political thinking; “To be obliged is to be liable to a sanction,”1 “Coercion is an essential element of law.”2 I intend to show that in fact, there is not such a great conflict between Hart’s intuitive ideas of law and Austin’s definitions, as first appears. The key is Austin’s, Hart’s, and others’ conception of sanctions in relation to rights, and the threat of nullity thereof. For Hart, nullity is not a sanction.3 I will show that, by analysing and eliminating discrepancies between different philosophers’ concepts of rights (strictly in the legal sense), we can produce a mutually agreeable definition of right – and therefore sanction, and eventually law. This unified definition proves that not only is nullity a sanction, but it is the only possible sanction (in the role of punishment) capable of producing a valid law. By this definition law does require sanctions, but the concept of sanction is made broad enough (or rather, restored to the breadth intended it by Austin and Kelsen) to accommodate common-sense intuitive ideas about law while excluding Hart’s (and others’) otherwise problematic counterexamples.
At the start it is important to make some observations pertaining to the way we talk about rules – specifically, I would like to draw a distinction between descriptive and prescriptive statements. This distinction is sometimes instead drawn using the terms “constative” and “normative” respectively, but I will eschew that language in order to avoid conjuring unintended connotations of moralistic discourse. It is nonsensical to call “rules” either descriptive or prescriptive, because when we discuss rules and laws we are actually talking about implications of the content of statements (which we inappropriately refer to as rules themselves). Statements can be either descriptive or prescriptive: Statements such as “if you do x, then you will gain/lose y rights” are descriptive – threats are these kinds of statements (in the form of pseudo-causal predictions of consequences). Statements such as “you do/do not have the right to x” are prescriptive – sanctions are these kinds of statements (in the form of metaphysical modifications of rights). A command such as “do x” is in itself neither descriptive nor prescriptive; it may however imply or specify a threat of a sanction. A valid rule need not be a command; a command by itself is superfluous as part of a statement forming a rule, if the rule-forming statement contains both the descriptive and prescriptive parts without falling back on implication. Commands on their own do serve a purpose in daily life, but only because of the way we interpret them. We read “Keep off the grass” as implying “Keep off the grass or x”. “Keep off the grass or x” is equivalent to the descriptive “If you do not keep off the grass, you will lose/gain y rights”.
Most positivist philosophical discussion of rules is inextricably bound up with discussion of rights; "The laws (properly so called) which are set by men as political superiors, or by men, as private persons, in pursuance of legal rights…"4 Hart himself, when discussing the attribution of rights, describes them as “a blend of fact and rule”,5 and in a theory he later disowned6 draws a distinction between description and ascription.7 Hart claimed that rights are not descriptive, but are ascribed to human beings. I shall stick with my original terminology, as “prescriptive” is more appropriate a term to apply to that which does the (active) prescribing – the sanction, which is what we are talking about, as opposed to the passive agent – to whom things (rights) are ascribed. I believe Hart was mistaken to abandon this distinction, although he was still not quite on the right track at that point; Hart’s order of dependencies is confused and circular. It is not appropriate to say that rights depend on rules, as modification of rights is necessarily a component of the definition of a rule (see below). Instead rights are simply a part of the description that forms a rule (or law). In complement to the description, a sanction or sanctions form the prescriptive element of a rule, as it brings about changes in rights (i.e. changes in parts of other descriptive statements). These two are necessarily separate in order for the sanctions of one rule to be able to modify another rule – it is logically inconsistent to say that a rule could modify its own definition, but a purely prescriptive sanction can modify the purely descriptive right forming another rule (or law).
More precisely, rather than forming part of a description of a law, it is better to say that rights (as descriptions in themselves) refer to a portion of the descriptive part of a law. In this way, it is possible to make statements about rights that imply (by reference) fuller statements about rules or laws without having to specify them explicitly – an abridgement that occurs regularly in ordinary language. Saying “I have a right to buy this magazine” does nothing other than imply the fuller claim “There is a law such that, if X occurs, then sanction Y.” In this case, X would be “shopper A pays the correct amount for a magazine, that is on sale, In a shop…(etc)” and Y would be something like “shopper A has a right to take the magazine or demand seller B to give him the magazine, and additionally, seller B no longer has the right not to (i.e. has a duty to) let shopper A have the magazine.” It should be clarified at this point that descriptive statements about rights and laws are not truth-apt; they sometimes sound like they are tracking the truth, or predicting facts about the future, but in fact they have no intrinsic truth value (though this may be attributed externally, as in the case of threats coming true – an external prediction that a threat will be carried out may be truth-apt). It is not in any way inconsistent for the implications of statements about rights to be false and the statements themselves to be true.
Kelsen was an influential thinker on the topic of rights and duties, and criticised Austin regarding these issues.8 Kelsen picks up on Austin’s distinction of “primary” and “secondary” rights and duties.9 He redefines rights and duties as “commands”, primary being the behaviour stipulated by the law, and secondary the behaviour stipulated by the sanctions attending the law. Referring to them as commands highlights their prescriptive rather than descriptive nature, however in many ways it obscures their common-sense meaning. If we instead return to Austin’s language (as later writers seem to) and look at them as “rights” and “duties” – that is, as specific states of the agent, rather than mysterious metaphysical entities somehow possessed by the agent – we see they are not in fact fundamentally different from each other, as the primary/secondary distinction may imply. Simply, the primary prescribe the state “before” and the secondary prescribe the state “after” transgression of the law. In fact, rather than “transgression” (which implies only negative laws and hence solely concentrate on law-breaking) we should say “fulfilment of the criteria of the law in order to warrant its sanctions” – we must remember sanctions can be positive as well as negative (really, they are neither – they are just not intrinsically value-charged at all). For instance, fulfilment of the requirement to have two witnesses to a will results in the ability to pass on one’s possessions – in this case, the requirements are before, and the ability granted is after, but both can be expressed in terms of rights and duties. The “before” is simply a right to perform certain actions which would fulfil a given criterion. The “after” is simply a right for the recipient to the will-maker’s possessions, attended by a miscellaneous host of duties applied to other individuals; the authorities have a duty to pass on the possessions to the recipient, etc (where “have a duty” really means “no longer have the-right-not-to” – see below).
Hohfeld produced a definition of the relationships between rights, liberties and duties that seems to have been almost universally accepted for the majority of the 20th century – rights, liberties, powers and immunities each have a corelative and an opposite.10 My definitions are not incompatible with Hohfeld’s, however there is ambiguous overlap between some terms as I use them, and those he defines, so it is necessary for me to distinguish between the two. My definition of “right” is equivalent to Hohfeld’s; however I wish to dispense with a concept of Hohfeldian “liberty” altogether, as it is surplus to requirements in the definition of law; the only common-sense definition we initially require when speaking of an agent’s position is between “where one may” and “where one must”. Rights provide the “may” aspect already, and separately defining liberties would unnecessarily overload the definition – in our descriptive rule we are not talking about whether others have a duty corelative with our agent’s right, as that is covered by the prescriptive sanction separately. I also wish to collapse Hohfeld’s opposites and corelatives into a single reversible reductivist definition; I define “duty” (the Hohfeldian corelative of “right”) as “no right not to” (the inverse of the Hohfeldian opposite of “right”). In dealing with sanctions as modifying rights, it is easier to talk about negative rights than rights and duties separately; in this way we can reduce “where one must” to “where one may not not”.
By performing such a reduction, we can clarify that rights are nothing metaphysical, but simply a description of a state of being where one’s activities or inactivities receive the protection of the law (by threatening sanctions to those who would interfere), the description of which is partly composed of the statement of the right. In fact, rights are often defined in terms of others’ obligations; "If a right is a legal right, it is necessarily a right to somebody else's behaviour, to behaviour to which the other is legally obligated. A legal right presupposes somebody else's legal duty,"11 "A law is a command which obligates a person or persons."12 It is important to note that this so-called duty held by the executors of the sanction, which would be the traditional Hohfeldian corelative to “right”, is actually reducible to a rule in itself, which modifies the rights (both positively and negatively) of the specified executors of a given sanction to another rule. For instance, the rule stating bank robbers must be put in prison does not give a duty to the prison guards per se. Instead if a man robs a bank, the sanction attending the rule he broke creates rights for the prison guards to imprison him, and creates a duty for them to (removes their previous rights not to) imprison him – which is in itself a rule, prescribing a set of sanctions to be carried out in turn if they fail to imprison him.
One of Hart’s central objections to Austin’s command theory is the claim that “many laws do not have sanctions,”13 instead providing facilities or conferring powers.14 We ought to analyse what these “sanctionless” rules have in common with sanctioned rules; it would seem both have a description, phrased in a predictive format, along the lines of “if x conditions are satisfied, then y will happen”. In this case it seems perfectly reasonable to refer to y as the sanctions attending the rule; Hart’s rejection of this statement stems from his inappropriately narrow conception of “sanction” as “punishment”. Kelsen before Hart in fact defined sanctions as either positive or negative: "It is a fact well worth noting that of the two sanctions here presented as typical [disadvantage, advantage], in social reality the first plays a far more important role than the second."15 Hart appears simply to disregard the more subtle option, but this leaves his theory flawed.
Hart also neglects the possibility of nullity as a (punishing) sanction. A law may not have a punishment, the only threat of not following it being the nullity of the offered right/power. But do we assume non-nullity of rights by default? As Bayles remarks, "Hart's argument assumes that in the absence of a special relationship limiting liberty, one has liberty."16 Or is the opportunity to have the law grant us “facilities” or “powers” (which can be collapsed to further rights) in itself a right? This right may surely be revoked, i.e. someone may be disallowed to marry (permanently, rather than an individual failure to complete a marriage ceremony). In this case, we are nullifying a right, and surely this counts as a sanction? Consider the example of metered parking: One is granted (by default) the right to park in metered spaces, and a rule is present that if they do not pay the meter, they will receive a fine. In reality, the fine is not itself a sanction, as parking offences are civil and not legal ones, so the only true sanction is that one may be disallowed to park their car in future if they have unpaid parking tickets – their previous right to park is revoked. The offence itself would constitute “failing to park” – going through the actions of parking (placing the car in a space) but failing to pay. This is the same as failing to make a will or failing to perform a marriage ceremony properly, although the latter failures do not specify further revocation of rights as a result, while the former does. There is instead the failure to gain the further rights that making a will or getting married offer, which is in itself a sanction (although a mild one, as it implies no permanent loss). The distinction here lies in two sorts of nullity sanctions; the failure to produce new rights (which would contribute to components of new rule descriptions), and the revocation of previous rights (starting to fulfil or no longer fulfilling the criteria of a pre-existing rule, thus incurring or losing its sanctions). Under our reductivist conception allowing for negative rights, revocation is reducible to negative fulfilment of criteria; removal of rights and imposition of duties are equivalent, therefore active sanctions and nullity sanctions are simply two sides of the same coin (contra Hart).
Hart talks about two types of “rules”.17 He seems to simply take this language from Austin without stopping to think about what it implies. He tries to isolate two types of rules – “keep off the grass” rules and “if someone does x, in conditions matching y, then z will happen” rules. Kelsen also distinguishes the two, although he argues that sanctionless commands (“norms”) are meaningless: “…this duty may also be presented in the form of a separate norm forbidding the delict… the formulation of such a norm undoubtedly facilitates the exposition of law. But … the only genuine legal norm is the sanctioning norm.”18 This agrees with our conception of sanctions as separate prescriptive (normative) creators of duty, but when reduced appropriately, the first turns out to just be a case of the second (if someone walks on the grass then their rights may be curtailed, or another may gain rights/duties to stop them). It is crucial to expand on what is meant by “then z will happen” – new rights are granted, old rights are revoked (nullity), new duties are created. If duties are read in this system as simply the removal of a pre-existing right to not x, as described above, we see nullity as the sanction, again. The concepts of granting and nullifying rights appear to be all that are required for a model of sanctions. In light of this, we can see Hart is right in asking for rules that create or remove rights, but he is wrong to draw a distinction between those and Austinian rules, and instead should accept nullity as a valid sanction.
It should be noted at this point that we are talking solely about legal rights – “rights which are creatures of law strictly or simply so called.”19 We are not interested in rights or duties from a moral perspective (concept of “wronging” someone20), as that would risk “prescriptivity” (hence my wish to avoid the term “normative”); when discussing what rights one has, we are being descriptive (as parts of rules). Such discussion of rights is closely involved with the concept of claim or advantage; “Rights are most often thought of as claims to something or alternatively as protected options to act,”21 “a legal right is a system of Hohfeldian positions that, if respected, confers dominion on one party in face of some second party in a potential confrontation over a specific domain and that are implied by the legal norm or norms that constitute that system.”22 By Hohfeld’s own narrow definition, “The term ‘a legal right’ refers to any sort of legal advantage.”23 In a context of law, for a law (or any rule) to be valid, the lawgiver must have the ability (do not confuse with Hohfeld’s “power”) to perform the sanctions. The statement “you have a right to x” simply means “you will not be made to not-x”. Sanctions in such a scenario being removal of rights or imposition of duties, the form they take is simply “A is made to x”, where for instance x is “enter and remain in prison”. This sanction upon A, modifying to A’s rights, may in itself constitute a “command” – a directed rule (see below for discussion of address) – for B, such as “if you do not force A to x, then z will occur to you”. This rule is just a descriptive statement, no different to stating “if you drop that vase, it will smash.”
Such laws can in fact form chains, with the conditional effects on one’s rights (the prisoner or the bridegroom) becoming conditions upon another’s (the jailer or the registrar). This chaining carries no risk of recursion or contradiction, as it is not possible for a lawgiver to describe a law with a prescribed sanction to be carried out against oneself – such a law would not be valid, as the lawgiver would not have the ability to see the sanction carried out. All such chains must lead to a not-always-traceable origin – either the lawgiver or the sovereign in some form. In reality, such chains can only be traced by attempting to break them, and they are usually redundant along their whole length. Historically, such chains collapse in situations of mutiny and rebellion. A sovereign’s ability to enforce sanctions diminishes as more sanctions are required (so fewer non-sanctioned individuals are available to carry them out), until a threshold is passed and the sovereign ends up powerless – at this point, none of the sovereign’s laws can logically be valid.
The concept of address seems to be important to most philosophers in reference to rules or at least sanctions; “A sanction is an evil inflicted on a person for not complying with a wish addressed to her.”24 Hart does acknowledge the subject, but claims that laws need not be promulgated or addressed: “It may indeed be desirable that laws should … be brought to the attention of those to whom they apply. … But laws may be complete as laws before this is done, and even if this is not done at all.”25 In defining rules versus laws however, it is logical to claim rules are always addressed to someone and laws can not be (must be addressed to everybody – address to no-one implies address to everyone). While common-sense examples of rules are (unaddressed) blanket cases, they need not be, which is why we use language such as “the rule is, whenever x happens, y” – suggesting there can be rules for specific cases (contra Hart’s gunman argument, see below). As above, a command is simply an implied rule, accompanied by specified or implied sanctions. A rule is valid if the sanctions can be carried out – if the addresser has the power to revoke the rights sanctioned. A rule becomes a law if addressed to everyone; the law is only valid however, if the lawgiver has the power over rights of all addressed, as per the sanction. For Hart however, the notion of authority is already presupposed in his definition,26 so this distinction may not be necessary for him.
It is worth taking a minute to analyse one of Hart’s objections against Austin’s Command Theory developed in his gunman analogy, the argument from persistence: “The gunman does not issue to the bank clerk … standing orders to be followed time after time by classes of persons. Yet laws pre-eminently have this ‘standing’ or persistent characteristic.”27 By his use of language, Hart has here established laws to be some metaphysical entity, independent of both lawgiver and recipient, that ought to be able to somehow “endure” – contrary to our conception of law as relation between descriptions. He then immediately proceeds to say that there must be a “general belief” in a punishment for disobedience, thus shifting the linguistic focus to the recipient. After that, Hart implies that the enforcer of any punishment for transgression must be another individual who does so “voluntarily, i.e. independently of fear of the threat.”28
Here Hart has performed some very sly manoeuvres that ought not go unchallenged. Having first disembodied the lawgiver from his law, and re-attached the recipient as someone with an active relationship to said law, Hart now tacitly implies that for a law to be valid, it must be voluntarily enforced by other individuals (As an aside, these individuals are ones he assumes must be un-coerced by “the” threat, as presumably they would not punish someone if they feared the same for themselves in turn - this implies also that there can be only one coercive factor in play, which is another fallacy). The result is we now have a law that has no lawgiver, or the lawgiver of which is assumed impotent to enforce it directly, relying only on fear in the individual to validate the law. This is fallacious on many levels; the “law” Hart started the argument with is very similar to the one I have described, a description of a set of conditions and their result. It ends however with a law the validity of which is determined by its imagined prescriptive value – how effective it is for prevention of its conditions. Hart has led us from a solid (albeit metaphysical) conception of a law to something relational, yet that no philosopher would describe as a law (which in fact matches closer our idea of sanctions attendant upon laws) – and is now holding that implied definition against Austin’s arguments. Specifically, this relational pseudo-“law” Hart has created can of course not endure if the condition for its validity is that it not be broken (rather, its initial conditions met). He demands that for the law to endure, the threat of the result of the initial conditions be enough to prevent them being reached, but has quietly removed the valid threat element by converting it to something other than a descriptive statement.
If we avoid this chicanery and apply the concept of endurance to our descriptive definition of law, we can see that a law can be said to “endure” for exactly as long as the description holds true (truth-aptness measured by and subject to external factors, as mentioned earlier). This is not a causal description, prone to failure in prediction – it is a conceptual one. For instance let us formulate a law thus: “if agent A (criminal) robs a bank, agent A will lose the right not to remain in prison (which he had previously), and gain the right to remain in prison, and agent B (jailer) will lose the right not to force agent A to remain in prison, and gain the right to imprison agent A (which he did not have previously).” Since the sanction consists of only shifts of rights (which are themselves purely descriptive) rather than physical events (although they may lead to physical events), the descriptive pseudo-prediction (i.e. the law) simply cannot fail (be false). Truth-aptness may be added by the physical events that attend it, in which case the prediction may turn out false, but the law will not be invalidated – for example, if a prisoner escapes, this does not render the law itself invalid. The only way such a law would fail to endure, would be in a situation where the sanction itself is no longer potent – i.e., when the lawmaker or sovereign can no longer enforce the sanction, because so many individuals have been sanctioned by the chain (see above – perhaps in a mass rebellion) that the conceptual description would cease to be true; rights that the failing sovereign would remove or grant are now under another’s authority to maintain or withhold (for instance, the rebel leader).
Such a breakdown would happen spontaneously and silently, the rights themselves being uninterrupted, but the chain of sanction that provides them changing its route. A suitable analogy would be water (rights and duties) from a reservoir (the lawgiver), flowing in many different streams (the chains of sanctions), over the landscape (the law) and into individual farms (those within the law). The shift of power example would be illustrated simply be the water beginning to flow from two reservoirs (the first being less full), and eventually only continuing to flow from the second, without any of areas of ground previously covered suddenly being left dry. The routes of the streams may differ but each point along them does not have a different amount, or sort, of water to before. There comes a point when the final drop has left the first reservoir, and all water now flows from the second, but only an observer at the first reservoir would be aware of this – every point in every stream keeps flowing in exactly the same way as before. So why, in the real world, are we sometimes aware of shifts of rights and laws when power changes hands? Sometimes the handover is not so smooth, and the initial reservoir suddenly ceases to provide water before the new one has established sufficient flow in the correct direction in all the streams; it quickly becomes evident which areas were relying on the first reservoir, as their riverbeds run dry (their rights are no longer corelative to others’ duties).
I wish to also clarify that I am not using a purely negative conception of rights – “A has a right to x” means “A will not be prevented from doing x” only down the chain of authority. This is not the same as saying “no-one may stop A doing x” (which implies “anyone who stops A doing x will face sanctions S”) as that would be a separate law entirely, although the one is often accompanied by the other. Also note that all rights I refer to are rights to/from the state/sovereign/sanctioning body. So-called rights between private individuals are just simplifications of more complex rights to the state. Appropriate to the water example, rights are one-way – the individual has or lacks rights (“rights reside in persons”29), granted by the state/monarch/lawgiver. It is illogical to say the state/monarch/lawgiver has or lacks rights, if there is nobody higher up for them to answer to.
Further to the above discussion on rights and sanctions, it seems in practicality, Hart may be ignoring the rights we assume upon entering society; in which case it would seem logical for him to say there is no loss in nullifying contracts – especially when combined with extremely low-loss examples (i.e. failing to make a will). Operating on the earlier parking ticket example however, we can see that a right of ours (the right to park) previously assumed granted upon application (as when attempting to create a will) is actually assumed by default, perhaps on entry to society, and its revocation is a real and potent sanction. It could be argued in fact, that the rights to property, to one’s body, and to one’s life are all assumed upon entering society (in theories such as Hobbes’), in which case seizure of property, imprisonment and the death penalty are simply nullifications of existing rights. This is highlighted by the use of “outlawry” as the most serious sanction in old common law – the revocation of all presumed social rights, including life.
In conclusion, we have shown it possible to paint a straightforward conception of law compatible with traditional punishments as well as nullity and positive creation of liberty, that can accommodate both Austinian addressed commands backed by threats and Hartian “keep off the grass” rules, that does not posit any odd metaphysical entities nor contradicts Hohfeld’s well-established corelative definitions, sits comfortably within the bounds of statement theories of language, does not claim or imply metaphysical truth-aptness or truth-tracking, and avoids treading on the toes of moral philosophy’s definitions of rights. Law is defined thus: A universally addressed externally truth-trackable description of rights, necessarily attended by sanctions which consist only in the prescription of creation and nullity of further rights.
- Austin, Lectures on Jurisprudence, p.96
- Kelsen, General Theory of Law and State, p.23
- Hart, The Concept of Law, p.33
- Austin, Op cit, p.60
- Bayles, Hart’s Legal Philosophy p.165
- MacCormick, H. L. A. Hart, p.17
- Hart, The Ascription of Responsibility and Rights, p.151
- Kelsen, Op cit, p.63
- Austin, Op cit, p. 76
- Hohfeld, Fundamental Legal Conceptions as Applied in Judicial Reasoning, p.89
- Kelsen, Op cit, p.75
- Austin, Op cit, p.96
- Hart, Op cit, p.27
- Bayles, Op cit, p.24
- Kelsen, Op cit, p.16
- Bayles, Op cit, p.153
- Himma – Inclusive Legal Positivism in Oxford Handbook of Jurisprudence, p.126
- Kelsen, Op cit, p.63
- Austin, Op cit, p.160
- Kamm, Rights, in Oxford Handbook, p.477
- Kamm, Idem.
- Wellman, Real Rights, p.8
- Hohfeld, Op cit, p.71
- Austin, The Province of Jurisprudence Determined, p.19
- Hart, Op cit, p.22
- Bayles, Op cit, p.22
- Hart, Op cit, p.23
- Hart, Idem.
- Austin, Lectures on Jurisprudence, p.160
- Hohfeld, Wesley Newcomb. Fundamental Legal Conceptions, as Applied in Judicial Reasoning. New ed. Aldershot: Ashgate/Dartmouth, 2001.
- Wellman, Carl. Real Rights. New York ; Oxford : Oxford University Press, 1995.
- Hart, Herbert Lionel Adolphus. The Concept of Law. 2nd ed. with a postscript edited by Penelope A. Bulloch and Joseph Raz. Oxford: Clarendon Press, 1994.
- Martin, Rex. A System of Rights. Oxford: Clarendon Press, 1993.
- Bayles, Michael D. Hart’s Legal Philosophy: An Examination. Dordrecht; London: Kluwer Academic, c1992
- MacCormick, Neil. H.L.A. Hart. London: Arnold, 1981.
Paton, George Whitecross. A Text-Book of Jurisprudence. 3d ed. Edited by David P. Derham. Oxford: Clarendon Press, c1964.
- Hart, Herbert Lionel Adolphus. Law, Liberty and Morality. London: Oxford University Press, 1963.
- Kelsen, Hans. General Theory of Law and State. New York : Russell & Russell, 1961, c1945.
- Austin, John. Lectures on Jurisprudence, or, The Philosophy of Positive Law. 3rd ed. / revised and edited by Robert Campbell. London: J. Murray, 1869.
- Austin, John. The Province of Jurisprudence Determined. London: J. Murray, 1832.
- Greenawalt, Kent. “Legitimate Authority and the Duty to Obey.” In The Duty to Obey the Law: Selected Philosophical Readings. Edited by William Edmundson. Lanham; Oxford: Rowman & Littlefield, c1999.
- Soper, Philip. “Legal Theory and the Claim of Authority.” In The Duty to Obey the Law: Selected Philosophical Readings. Edited by William Edmundson. Lanham; Oxford: Rowman & Littlefield, c1999.
- Green, Leslie. “Who Believes in Political Obligation?” In The Duty to Obey the Law: Selected Philosophical Readings. Edited by William Edmundson. Lanham; Oxford: Rowman & Littlefield, c1999.
- Marmor, Andrei. “Exclusive Legal Positivism.” In The Oxford Handbook of Jurisprudence and Philosophy of Law. Oxford: Oxford University Press, 2002.
- Himma, Kenneth Einar. “Inclusive Legal Positivism.” In The Oxford Handbook of Jurisprudence and Philosophy of Law. Oxford: Oxford University Press, 2002.
- Kamm, F. M. “Rights.” In The Oxford Handbook of Jurisprudence and Philosophy of Law. Oxford: Oxford University Press, 2002.
- Green, Leslie. “Law and Obligations.” In The Oxford Handbook of Jurisprudence and Philosophy of Law. Oxford: Oxford University Press, 2002.
- Hart, Herbert Lionel Adolphus. “The Ascription of Responsibility and Rights” in Logic and Language. 1st series, Garden City, NY, Doubleday & Co, Anchor Books, 1965, c1948.