Logical Techniques for International Law
Jaap Hage and Antonia Waltermann*
University of Maastricht
Conflicts of rules occur when it is possible that two or more rules attach incompatible legal consequences to a case. Such conflicts can be conflicts of imposition, where two rules impose incompatible facts upon the world, or they can be conflicts of compliance, where two rules demand incompatible behavior from one and the same agent. These conflicts can occur between rules of the same system, but equally between rules of different (legal) systems. International law is not excluded from this development, and is in fact in many ways particularly prone to such rule conflicts.
To deal with these conflicts, it is often necessary to make exceptions to one or more of the rules that are involved in the conflicts. One major purpose of this contribution is to investigate which techniques logic has on offer to deal with conflicts between rules and the exceptions that are needed in that connection.
A second major purpose of the present contribution is to see what logic can offer us as techniques to avoid rule conflicts. If only one of two conflicting rules is applicable to a case, there is no conflict.
Although logic can provide us with techniques to avoid and to handle conflicts of rules, it cannot make the decisions for us. The knowledge which techniques are available only gives us an indication of what is possible. Next to that we need a view on what is desirable, and this view requires the study of law, politics, international relations and morality.
Keywords: applicability of rules, application of rules, burden of proof, conflict of compliance, conflict of imposition, derogation, exceptions to rules, incorporation, interpretation, legal systems, logic of rule application, rule conflict, scope of rules
Legal rules often conflict with one another. When a conflict occurs, it may be necessary not to apply a rule even though it is applicable to a certain case. Running ahead of a more precise characterization, we can call this making an exception to a legal rule. One of the reasons for making an exception to a rule is that this rule conflicts with some other rule and that it is not possible or desirable to apply both rules. Then at least one of the rules must give in, leading to an exception to that rule. This kind of rule conflict can occur within a single legal system, as when a municipality locally prohibits pubs to be open on Sundays, while that is allowed nationally. It can equally occur across different legal systems, for example when a Member State of the European Union taxes that which is not taxable according to EU law. We can also find rule conflicts between state law and religious laws, or between rules of law and rules of morality.
International law is not exempt from this problematic of rule conflicts. In fact, it has many features that make conflicts of rules particularly frequent and recalcitrant. First, it is not clear whether international law constitutes a single legal system, or an archipelago of mutually interacting international legal regimes. This so-called fragmentation of international law conjoined with the growing interaction between the international, domestic and regional legal orders implies that any discussion of rule conflicts in international law will have to account for conflicts across different legal orders (Koskenniemi & Leino 2002, Koskenniemi 2006, 8). Second, the importance of the formal sources of international law has become smaller: the list of formal sources has little power to control what counts as law and what does not. In fact, many academics suggest that having a clear view of international law requires us to bypass the doctrine of sources and identify as law those rules that have real world effectiveness and/or political legitimacy (cf. Franck 1990 and Arend 1999). The resulting obscurity of what counts as international law is also a potential cause of rule conflicts. This second point is loosely related to a third: international law operates without a central legislator. Treaties and customs are generated through the agency of more or less uncoordinated states, which means that no central authority can take efforts to minimize rule conflicts or prevent them from arising. In short, international lawyers as well as academics will inevitably be confronted with rule conflicts in international law. These rule conflicts may in turn necessitate exceptions to legal rules, whether these rules have their origin in international or in national or religious sources.
In this contribution we attempt to give an overview of the ways in which law deals with conflicts between rules within and across normative systems and with exceptions to rules. The emphasis will be on rule conflicts in international law, but some other rule conflicts and causes of exceptions will also pass in review. Our approach is inspired by the tools that have been developed during the last three decades in logical research of legal reasoning (Schauer 1991; Gordon 1995; Prakken and Sartor 1996; Hage 1997 and 2005; Prakken 1997; Sartor 2005). We try to use these tools, but although our argument is inspired by research in formal logic, our approach will be completely informal.
The argument of this contribution is structured as follows. In section 2 we will sketch the general background against which rule conflicts and exceptions to rules occur, and discuss several ways in which rule conflicts and their ensuing exceptions to rules can be avoided. If rule conflicts nevertheless occur, there is need to make an exception to one of the conflicting rules and this technique is discussed in section 3. Section 4 concludes this contribution.
A discussion of techniques to avoid rule conflicts presupposes a common understanding of what rules are. Such a common understanding is not obvious, if only because some assume that legislation and treaties contain legal rules, while others assume that although legislation and treaties are sources by means of which legal rules are created, they do not contain the rules themselves. Here we adopt the latter view: legislation in a broad sense, including the creation of treaties, is a means to create law, but means should not be confused with what they produce. This seemingly subtle distinction has immediate practical implications, because it is obvious that legislation needs to be interpreted, while it is far from obvious that rules are also in need of interpretation. Here we assume that a rule consists of conditions and a conclusion part, and that a rule is applicable to a case if the case satisfies the conditions of the rule. We will generally use ‘rule’ in a sense that encompasses principles throughout the rest of the paper.
We will use the term ‘interpretation’ for the step from a rule-creating text to the formulation of the rules in terms of its conditions and its conclusion. The decision whether a set of concrete facts satisfies the conditions of a rule is not called ‘interpretation’ of the rule, as is often done, but rather ‘classification’ of the facts. Classification is always case-related; interpretation is, at least in theory, independent of cases.
Under international law, several distinctions can be made with regard to the existence of conflicts between rules. Firstly, there are two possible perspectives: that of subject-matter of the rules, and that of the legal subjects (Koskenniemi 2006, 21). We will here focus mainly on the perspective of the legal subject. Secondly, the term “conflict” can be interpreted broadly, meaning a conflict can arise, for example, between the aims of two treaties, or narrowly (Koskenniemi 2006, 24). We take a narrow interpretation here, and argue that there are at least two ways in which rules can conflict with each other. The one is that one rule makes that something is the case, while another rule makes that it is not. For instance, according to the principle of effective possession, a certain territory which has been in factual possession of a State for some considerable time, is deemed to be part of that state. According to the principle of stability of borders, however, that same territory is deemed to belong to a different State, because it reflects legally agreed upon borders. Given how territory works, the land cannot belong to both States at the same time, and the effects of the rules are incompatible. Another example is that the UN Charter prohibits the use of force, but that military action under the auspices of the Security Council is permitted. One rule prohibits an armed intervention into another sovereign state, while the other permits precisely this kind of behavior. The states of affairs that armed intervention is both prohibited and permitted are incompatible, and therefore these rules are in conflict. The conflicting rules impose incompatible facts upon the world, which cannot co-exist. Because rules impose facts upon the world – they make it the case that these facts obtain as legal consequences – we will call conflicts such as these ‘conflicts of imposition’.
The second way in which rules can conflict with each other has to do with rules which prescribe incompatible forms of behavior. Such a conflict occurs if a State, because of the human rights to which it is committed, is obligated to invest more money in both education and health care, while the state has only money for one of them. Another example would be that a journalist is obligated to reveal the sources on which she based a controversial publication, while she promised her informant not to reveal his identity. Here, the conflict arises because the obliged agent cannot comply with both obligations, be it because the obligations themselves are in conflict (a prescription and a prohibition of the same behavior), or because factual circumstances make compliance impossible. Consequently, conflicts such as these may be called ‘conflicts of compliance’.
Conflicts of imposition are to be avoided if one wants the set of institutional facts created by law to be consistent. Conflicts of compliance should be avoided to safeguard agents against the choice which obligation to violate. Conflicts of compliance can only exist between mandatory rules (prescriptions and prohibitions), while conflicts of imposition can exist between all rules, including rules that assign status (e.g. the status of a piece of land of belonging to a particular state).
Conflicts of imposition and conflicts of compliance can arise within one normative system, as was the case with the aforementioned example of the UN Charter which prohibits and permits the use of force. They can also arise between different normative systems, both as conflicts between law and other types of normative systems such as morality, or between different legal systems. The above example about the journalist who promised not to reveal her source illustrates a conflict between a legal requirement to reveal and a moral obligation not to do so.
When law conflicts with some other normative system, Raz’s view on the authority of law raises relevant concerns. According to Raz, law claims authority (Raz 1979, 28-33) and this claim involves two things. First, law would provide us with reasons for action, and second, these reasons would exclude other reasons. With this last point, Raz means that according to law, reasons for deviating from what the law prescribes should in principle be ignored. It is true that law makes exceptions to this principle, for instance by sometimes allowing conscientious objections, but Raz’s starting point is that the law claims not only to provide reasons for action itself, but also to take away the reason-giving force of norms from other systems. From the legal point of view it would in principle be irrelevant if what the law prescribes conflicts with morality, or with some other legal system. In this way, a legal system distinguishes itself from those other systems.
The distinction becomes visible in the necessity to add subscripts to legal judgments. For instance, it is not the case anymore that state A may not use military force to intervene in state B to prevent gross human rights violations. If normative systems are distinct, the judgement must be that legally, state A is prohibited from intervening (in the absence of Security Council authorization), but morally, it should intervene. In this example the subscripts distinguish between the legal and the moral point of view. However, it is also possible to distinguish between different legal points of view. For instance, according to European Union law, certain forms of positive discrimination are not permissible, whereas according to the Committee on the Elimination of Discrimination against Women, they are required (cf. Waddington and Visser 2012).
If the legal and other judgments are subscripted, seemingly conflicting judgments are rendered logically consistent. This means that the rules that lead to these conflicts cannot create a conflict of imposition. However, an agent who is confronted with mandatory rules from different legal systems that both claim obedience but which rules cannot both be complied with, is still burdened with a conflict of compliance. Despite the fact that conflicts between rules of different legal systems cannot be conflicts of imposition, they may still be conflicts of compliance and therefore there is still pragmatic reason to avoid or to deal with them. Some of the logical techniques discussed in this paper also point toward solutions for compliance conflicts across legal systems.
Most legal rules identify by means of their conditions the kind of cases or the persons to which they are applicable. This can be everybody, as in Art. 2 of the European Convention on Human Rights, or classes of agents, such as the judges in the European Court of Human Rights, as in Art. 21 of the same Convention. However, there are also limitations on the cases and persons to which a rule applies that are not mentioned in the conditions of the rule. These are scope conditions, which combine with the rule conditions in the narrow sense to determine to which cases or persons rules should be applied.
There can be personal, spatial (territorial), and temporal scope conditions. Personal scope limitations occur when one rule applies to a certain class of persons and the other rule applies to a different class of persons. So, for instance, Talmudic law applies to members of the Jewish people, but not to non-Jews. Spatial or territorial scope limitations, meanwhile, refer to a distinction depending on the place where certain events takes place. So for instance, the penal laws of States typically apply to events that took place in their own territory, but not in the territory of another state. Temporal scope limitations solve conflicts between rules by postulating that the rules apply in different time periods. For example the customary rules of treaty interpretation and the rules found in the Vienna Convention on the Law of Treaties 1969 are different and thus they seem bound to conflict, but the conflict can be avoided by postulating that the rules of the Vienna Convention apply only to treaties that entered into force after 1969, as in fact stated in the convention itself. In this way, the scope of a rule can prevent conflicts between rules from ever arising, at least where the potentially conflicting rules have different scopes.
A conflict of rules is typically case-related: two rules impose incompatible consequences on a case, or demand incompatible lines of action in a particular situation. This means that only rules that are applicable to a case can lead to a conflict, and also that a potential conflict can be avoided by interpreting a legal source in such a way that the resulting rule is not applicable to the case in question. The following example, inspired by an ICJ case, illustrates this technique.
Assume that the Constitution of a federal state prescribes that the State respect the constraints of federalism, while international law prescribes that the federal State stop one of its constituent States from engaging in internationally wrongful behavior. These actions cannot be performed both. The rule conflict is clear, and a breach of either an international or domestic duty seems unavoidable. Given this impasse, it may be possible to interpret a text that seemingly creates an obligation of a specific type - e.g. stop damage causing behavior - as a text creating an obligation of a more general type - e.g. avoid causing lasting damage. It is possible to comply with this latter obligation without violating the former obligation, for instance by compensating the damage that results from the behavior. If this is done, the demands of both legal systems are deemed to be satisfied, and the conflict is avoided.
A related technique is that a new type of action is created that can solve the impasse between two colliding norms for an important range of cases. On a straightforward interpretation, the UN Charter rules out all acts of aggression that are non-defensive and not authorized by the Security Council. Arguably there is a developing rule of customary international law that suggests that states have a duty to intervene in cases of genocide, war crimes and crimes against humanity to defend the civilian population, and this duty exists irrespective of whether Security Council authorization has been given or not. If a case of genocide breaks out, and the Security Council does not authorize action, the two rules seem to conflict. One way to avoid the conflict is to devise new action types such as ‘peacekeeping’ that do not fall within the concept of aggression that is prohibited by the UN Charter and that do not violate the principle of non-intervention in the same (cf. Thompson 2008).
Conflicts of compliance exist when it is impossible for an agent to comply with two (or more) mandatory rules at the same time. This impossibility can be removed by changing the background facts that make it impossible to comply with both rules. Take again the obligations of a State to invest in both education and health care. The State cannot comply with both obligations because of a lack of financial means. By generating more money, for example by raising taxes, the State can take this cause of impossibility away. As soon as the State can invest in both education and health care, it can comply with both of its obligations, and there is no conflict of compliance anymore.
At times, rules conflict with factual necessity, or with rules made to cope with factual circumstances such as in states of emergency. Human rights on privacy might for instance conflict with measures taken to prevent terrorist attacks or investigate them. One technique to prevent such conflicts from arising is derogation. Derogation allows a state to take measures derogating from its obligations under a treaty, to the extent necessary to handle certain situations (see e.g. Article 15 of the European Convention on Human Rights). Logically speaking, there is no conflict in cases of derogation between the rule of the treaty and the rule on the basis of which the measures are taken, because derogation means that the treaty rule does not apply for the duration of the derogation. In short, derogation involves the temporary suspension of applicability of a potentially conflicting rule, thereby avoiding the conflict.
The easiest way to avoid the dilemma of inter-systemic rule conflicts is to ensure that such conflicts do not occur. We have seen that interpretation and derogation, but also scope limitations, can fulfil this function, in that e.g. the national law of one state is limited in its application to the territory of that state only. With regard to international law in particular, however, scope limitations do not manage to avoid all conflicts. Methods such as incorporation or reference can prevent inter-systemic conflicts from arising as well.
Rules of a ‘foreign’ system can be used in a legal system through a technique which may be called ‘reference’. The foreign rules are not incorporated in the legal system, but their existence and content is considered by the system as facts that are legally relevant from the point of view of the legal system. Reference avoids conflicts between the rules of the referring system and the rules of the system to which reference is made, because the content of the referring system is adapted to the content of the referred system. Private International Law provides many examples of this, because it contains meta-rules that determine which national legal system provides the applicable rules. For example, the judgment whether a couple has divorced is given in country A on the basis of the rules of country B, the validity of which is from the perspective of country A a matter of fact. These rules are not incorporated in some international set of object-level rules, and neither is there an independent international system. As a consequence, the rules of the national systems determine the outcomes of cases, without a potential conflict with rules of an international system. In this way, rule conflicts can be avoided.
In case of reference, the content of a foreign system is treated by the own legal system as a matter of fact that co-determines the application of the domestic law. In case of incorporation, meanwhile, foreign law becomes part of domestic law. The typical example of this phenomenon is the incorporation of international law in a national legal system in so-called ‘monist’ legal systems. The Dutch legal system nicely illustrates incorporation. Provisions from international treaties ratified by the Netherlands and rules created by international organizations in which the Netherlands participate (in particular the European Union) automatically become part of the Dutch legal system (Art. 93 Grondwet). The foreign rules are not foreign anymore, except in the sense that they were not created by native Dutch legislative bodies. However, they are part of the Dutch legal system to the same extent as home-made rules.
Strictly speaking, incorporation is not a technique to deal with conflicts between rules of different systems, but a way to ensure that only one legal system is relevant. If EU regulations become automatically part of the Dutch national law, there is no need any more to pay attention to EU law, because the relevant rules are already part of Dutch national law. In the case of the EU one may even ask whether there exists such a thing as the EU legal system, because arguably the EU only provides organs which can create (uniform) law that becomes part of the national legal system of the Member States. If all counties would have monist systems with regard to the relation between laws of domestic and laws of non-domestic origin, the same might be said about the provisions of human rights treaties. These treaties would then create uniform human rights in different legal systems and it might be argued then that there is no separate international human rights system. However, theoretically it is imaginable that some legal system incorporates part of a foreign legal system, while that foreign system has independent existence. The situation is then comparable to one country that uses the national currency of some other country.
If ‘foreign’ rules are incorporated in a national legal system they are not foreign rules anymore but merely rules with a foreign origin. Such rules may still conflict with rules of a national origin, or with other rules of foreign origin. However, because of the incorporation, such conflicts are not conflicts between legal systems anymore. What is avoided by incorporation is not a conflict of rules, but a conflict between legal systems. If there is still a conflict of rules, the techniques used within a single legal system to deal with conflicts, such as making exceptions, can be used to deal with possible conflicts between rules from national and international sources.
A common way to avoid inconsistencies within a single legal system is to avoid rule conflicts by preventing conflicting rules from entering into existence at all. A national legislator for instance might make an exception to the general right of free speech for cases of hate speech. This will not allow a local legislator to make an exception to the exception for hate speech against people of a particular origin, such as French speaking people from Walloon. If a local legislator nevertheless attempted to do so, its rules would simply not be recognized as valid law: the local legislator does not have the power to make rules that conflict with the ‘higher’ rules of the national legislator. This limitation of power avoids conflicting rules.
With regard to international law, however, limitations of power occur less frequently. This is because such limitations suggest an overarching organization of the distribution of rule creating power that can divest certain actors of their ability to create rules. Such organization is absent in the international plane. As mentioned in the introduction, rule creation in international law is decentralized and tends towards anarchy as even the importance of the doctrine of the sources of law seems to be decreasing. Instead of limitation of power, international law usually creates prohibitions, which can be violated. The typical sanction is not invalidity, but the need to make some sort of compensating behavior, which may range from monetary compensation to restitution. This is different when it comes to European Union law, which ‘has as a corollary the impossibility, for the member-State, to give preference to a unilateral and subsequent measure against a legal order accepted by them’ (Case 6/64, Costa v ENEL, ECR 585).
In the previous section we have seen that selecting which rules are applicable to a case can avoid rule conflicts, and that there are a number of logical techniques available to aid in the selection of rules, from scope limitations to derogation, interpretation, incorporation, reference and limitations of rule-creating powers. However, it is clear that even with these techniques to avoid conflict, rule conflicts nevertheless occur. Then it is desirable to make an exception to one of the conflicting rules. In this section we will take a closer look to rule exceptions.
An exception occurs if a rule that is applicable to a case is nevertheless not applied to that case, and does not generate its normal legal consequences. Applicability and application are in this connection somewhat technical notions with a precise meaning. Both applicability and application concern a relation between a rule and a particular case. If a rule is applied to a case, this means that the rule attaches its consequences to this case. Such a consequence may be that an agent, for instance a State, has the competence to perform some juridical act such as concluding a treaty. It may also be that an agent such as an IGO becomes liable for damages, for instance because it violated an international obligation.
Normally a rule will be applied to a case if and only if it is applicable to that case. Applicability is determined by three factors:
1. a rule must exist, or – which boils down to the same thing - it must be valid;
2. the case must fall within the – territorial, temporal and personal - scope of the rule; and
3. the case must satisfy the conditions of the rule.
The conditions of a rule are given with the rule formulation, which expresses the conditions and conclusion of the rule. For example, Article 36 of the UN General Assembly Resolution 56/83 creates a rule that defines a State’s liability for particular damage, and gives as conditions for this liability that there was an internationally wrongful act, that a State was responsible for this act, that the damage was caused by this act, and that the damage was not (yet) made good by restauration.
If a rule is not applicable to a case, but nevertheless applied, this is most often a case of rule application by analogy. We will not pay attention to this possibility here.
If a rule is applicable to a case, but nevertheless not applied, an exception was made to the rule.
The idea that legal rules are open to exceptions is somewhat controversial. That has to do with a certain ambiguity in the notion of a legal rule. A legal rule may be seen as something that flows directly from a legal source, such as a treaty, legislation, international custom, or case law. It may also be seen as a general connection between operative legal facts and legal consequences. These two are not identical, as may be illustrated by a simple example.
Under the UN Charter, the use of force is prohibited. However, the Security Council may authorize military intervention in cases of threats to international peace and security. If we have a case in which there is a threat to international peace and security and the Security Council authorizes a military intervention, the two rules seem to conflict. Since the second rule is a lex specialis with regard to the main rule it would normally prevail over it, and application of the second rule would make an exception to the first rule.
However, it is also possible to combine the two original rules into a ‘derived rule’ without exceptions, and this derived rule would read that the use of force between States is prohibited unless the Security Council has authorized such use because there is a threat to international peace and security. This derived rule has the absence of the exception as one of its conditions, and it would not even be applicable if the Security Council has authorized the use of force. Since exceptions are only possible to rules that are applicable, this new derived rule would not suffer from an exception in case of authorized use of force.
Let us have a more systematic look at this phenomenon of ‘derived rules’. If a rule is applicable to a case this normally means that the rule is applied to that case and attaches its legal consequences to it. This is so normal that the logic of rule application seems to be nothing else than an ordinary syllogistic argument (Alexy 1983, 273-283; MacCormick 1978, 19-53). The facts of a case are subsumed under a general rule, and the conclusion that describes the legal consequences of the case follows deductively. This deductive application of rules seems so natural that it requires explanation that exceptions to rules are possible. If a rule seems to have an exception, why not say that the rule was not formulated properly, and that it actually has an additional condition namely that the exceptional circumstances are absent?
The insight that rules can have exceptions can be reconciled with the impression that rules can be applied in deductive arguments by means of so-called ‘case-legal consequence pairs’ (CLCPs) (Hage 2005, 27). CLCPs describe the effects of rules such as the prohibition of the use of force and the permission of military action when authorized by the UN Security Council. The two inconsistent rules are combined into a single ‘rule’ that leads to a single consistent result. We use quotation marks here to indicate that this ‘rule’ differs from the two rules that were used to construct it. Both the rule prohibiting the use of force and the rule permitting it in cases of Security Council authorization are directly based on an official legal source, in this case the UN Charter. The derived ‘rule’, however, cannot be traced back directly to such a source, but is the result of combining the two original rules in light of their apparent purposes, thereby creating a CLCP.
It is possible to characterize a legal system as defined by an exhaustive set of such CLCPs: for every kind of case (abstract case) that has legal consequences, there exists a CLCP that gives the characteristics of the kind of case and the legal consequences attached to it. These CLCPs are the outcome of the original rules (including rights and legal principles or incorporated or referred rules) of the system, interpretation and solutions of potential rule conflicts by means of prevalence (such as les superior) or any other technique the system in question employs to resolve rule conflicts.
The CLCPs are constructed in such a way that no particular case can fall under two different abstract cases to which incompatible consequences are attached. For example, there will be a case for ‘military action without Security Council authorization’ and one for ‘military action with Security Council authorization’, but not one for military action in general, because the latter might give a different legal consequence in a concrete case of military action with Security Council authorization. Understanding a legal system as an exhaustive set of CLCPs, it is not possible that a case has inconsistent legal consequences. Imminent inconsistencies are filtered out in the step from the original (conflicting!) rules to the CLCP. Moreover, there are no exceptions to CLCPs. If there seems to be an exception, this means that the CLCP was formulated too broadly: there should be two different CLCPs, one for the normal cases and one for the exceptional cases.
If a legal system is seen as an exhaustive set of CLCPs, exceptions to rules only play a role in this step from the original rules based on legal sources to the derived ‘rules’ – the CLCPs - that define the outcome of all the interacting original rules. It is this step that requires a non-deductive logic. The derivation of the legal consequences of a case by applying the relevant CLCP to that case can be purely deductive, because all exceptions have already been filtered out in constructing the CLCP. When we discuss exceptions to rules, we are not necessarily talking about the immediate application of rules to cases; we may also be talking about the construction of CLCPs which can in turn be used for legal justification is a deductively valid manner.
It is possible to characterize a legal systems by means of an exhaustive set of CLCPs, but is this also desirable? One factor that should play a role in answering this question is whether exceptions to rules have the same function as negative rule conditions. If their function is the same, the existence of exceptions should not impede the reformulation of a legal system as a set of CLCPs. However, if exceptions have a function that exceeds that of negative rule conditions, things become different.
As a matter of fact, exceptions do have a role that exceeds that of negative rule conditions. However, this function can only be recognized if we stop treating rules as a kind of premises in deductive arguments, and start to treat them as tools in the production of arguments in legal dialogues. In such a dialogue a party who wants to justify a claim can adduce a rule supporting that claim and facts that satisfy the conditions of this rule. If this happens and the opponent in the dialogue does not react, the claim counts as justified. If an exception should be made to the rule, the opponent in the dialogue must claim this, and justify the claim by providing reasons for making the exception. This means that the opponent who wants to invoke the exception, has the burden of proof for this exception. The burden of proof for the applicability of the main rule rests on the person who uses the rule in justifying the main claim. In other words, the invocation of an exception goes hand in hand with a shift in the burden of proof. Such a shift would not take place if the exception would merely be a negative rule condition, since the burden of proof for the conditions of a rule, including the negative ones, rests with the dialogue party who invokes this rule.
Because exceptions fulfill a different function than negative rule conditions, it is not desirable to represent a legal systems as an exhaustive set of CLCPs, at least not if the function of exceptions as distributor of the burden of proof is relevant.
There must be a reason for making an exception to a rule; the exception itself is the outcome of an argument on whether the rule should be applied, and not an independent reason against application. It is, for instance, not possible to say that state A violated WTO law, but should not be sanctioned because there is an exception to the violated rule, even though there are no reasons for making such an exception.
Precisely what counts as a reason to make an exception to a rule is not a matter of logic or even legal theory; it is a matter of substantive law. To find out what these reasons are in a particular legal domain such as international law, one should study the domain in question to discover which reasons are recognized as reasons to make exceptions to rules. However, it is possible to say a little bit more in general: reasons for exceptions that are generally recognized are that application of a rule would be against the purpose of the rule, and that application of a rule would lead to a conflict with another rule which is also applicable. Here we will pay no additional attention to conflict with the rule’s purpose as a reason for making an exception to a rule. If the rule conflicts with another applicable legal rule, the rule conflict itself will normally be treated as a reason to make an exception to one of the conflicting rules.
The ‘logic’ of exceptions to rules works as follows. If a rule is applicable to a case, this is a contributory reason to apply the rule, and to attach the rule’s conclusion as a legal consequence to the case. Normally, there are no reasons against applying an applicable rule, and then the applicability of the rule suffices as reason to actually apply the rule, and to make it attach its consequences to the case. If, for example, the United States violated an international obligation and are responsible for this violation, and if damage was caused by this act and the United States did not yet compensate that damage, these facts together constitute a reason to apply the rule of Article 36 of the UN General Assembly Resolution 56/83 to the effect that the United States must compensate this damage.
Sometimes, there are contributory reasons against applying an applicable rule. If there are both contributory reasons for and against applying a rule to a case, these reasons must be ‘balanced’. This ‘balancing’ is little more than taking a decision which reasons outweigh the other reasons. If the ‘balancing’ of reasons leads to the conclusion that an applicable rule should not be applied, we say that there is an exception to the rule. This exception is in this case nothing else than the outcome of ‘balancing’ the reasons for and against application of the rule. If in a particular case there is an exception to a rule that is applicable to that case, this rule should not be applied, and its consequences are not attached to the case. Furthermore, if a similar case arises later, the exception should hold in that case as well, or an argument needs to be made why this is not the case.
Since it is not desirable that the rules of a legal system attach incompatible legal consequences to a case, the possibility that this might occur is a reason not to apply one of two rules that are in conflict. This immediately raises the question which rule to apply and which rule to discard. In other words, it becomes necessary to make a choice as to which rule prevails over the other. Again, this is not a choice that logic is suited to make by itself, but logic can provide the techniques to deal with prevalence between rules and the implications this has for avoiding actual rule conflicts. The relevant argument in this connection goes as follows:
If two rules are applicable to a case, and their application would result in incompatible legal consequences, the rules are in conflict in this case. Such a conflict is a contributory reason against the application of one of the two rules. The fact that application of both rules would lead to an inconsistency is a stronger reason against application than the applicability of both rules is as reason for their application. The intermediate conclusion is that one of the two conflicting rules should not be applied. Making a choice as to which one is a matter of prevalence: the rule that prevails over the other should be applied.
Several contributory reasons can be - and in fact are - recognized in this connection. One option is that the rule that better fits in the overall legal system prevails over the less fitting rule (coherence). Another option is that the rule that was made by the ‘higher’ authority prevails over the rule made by the lower authority (lex superior). Equally, the more specific rule could prevail over the more general rule (lex specialis), or the more recent rule over the older one (lex posterior).
Taking the lex posterior rule even one step further is the technique of implied repeal, whereby it is presumed that if the later rule conflicts with the earlier rule, the later rule not only prevails, but the earlier rule is in fact repealed. However, if one of the ‘conflicting’ rules counts as repealed, there is no real conflict, since the repealed rules does not exist anymore and can for that reason not be applicable.
The development of the internet, the rise of transnational law, the co-existence of different legal traditions and sub-traditions, and globalization all mean that the opportunities for conflicts between rules increase. International law is not excluded from this development, and is in fact in many ways prone to such rule conflicts. A conflict of rules occurs when it is possible that two or more rules attach incompatible legal consequences to a case. Such conflicts can be conflicts of imposition, where two rules impose incompatible facts upon the world, or they can be conflicts of compliance, where two rules demand incompatible behavior from one and the same agent. These conflicts can occur between rules of the same system, but equally between rules of different (legal) systems.
To deal with these conflicts, it is often necessary to make exceptions to one or more of the rules that are involved in the conflicts. One major purpose of this contribution is to investigate which techniques logic has on offer to deal with conflicts between rules and the exceptions that are needed in that connection. Logic cannot and should not dictate a particular way of dealing with rule conflicts, but it can be of help by providing a conceptual framework that clearly defines when a rule conflict occurs, and which techniques are available to deal with it. Moreover, it can help us understand why the view that rules are without exceptions can be very attractive.
A second major purpose of the present contribution is to see what logic can offer us as techniques to avoid rule conflicts. If only one of two conflicting rules is applicable to a case, there is no conflict. Subscripting rules from different normative systems, the use of scope conditions, interpretation, changing background conditions, derogation, incorporation, reference and limitations of the power to create rules are all discussed as techniques to prevent conflicts.
It is important to emphasize, however, that although logic can offer us techniques to avoid and to handle conflicts of rules, it cannot make the decisions for us. The knowledge which techniques are available only gives us an indication of what is possible. Next to that we need a view on what is desirable, and this view requires the study of law, politics, international relations and morality.
R. Alexy, Theorie der juristischen Argumentation, 7e Auflage, Frankfurt a/M: Suhrkamp 1983.
Anthony Clark Arend, Legal rules and international society. Oxford: Oxford University Press 1999.
Bath and Krabbe 1982
E.M. Barth and E.C.W. Krabbe, From Axiom to Dialogue, Berlin: Walter de Gruyter 1982.
Jonathan Dancy, Ethics Without Principles, Oxford: Clarendon Press 2004.
Thomas M. Franck, The power of legitimacy among nations. Oxford: Oxford University Press 1990.
Th. F. Gordon, The Pleadings Game. An Artificial Intelligence Model of Procedural Justice, Dordrecht: Kluwer 1995.
Jaap C. Hage, Reasoning with Rules, Dordrecht: Kluwer 1997.
Jaap Hage, ‘Rule Consistency’, Law and Philosophy 19 (2000), p. 369-390.
Jaap C. Hage, Studies in Legal Logic, Dordrecht: Springer 2005.
Jaap Hage, ‘The (Onto)logical Structure of Law: A Conceptual Toolkit for legislators’, in M. Araszkiewicz and K. Pleszka (eds.), Logic in the Theory and Prqactice of Legislation, Cham: Springer 2015.
Koskenniemi & Leino 2002
Koskenniemi, M., & Leino, P. ‘Fragmentation of international law? Postmodern anxieties.’ Leiden Journal of International Law, 15 (2002), 553-579.
General Assembly A/CN.4/L.682, ‘Fragmentation of international law: difficulties arising from the diversification and expansion of international law, Report of the Study Group of the International Law Commission, finalized by M. Koskenniemi (2006)
Lorenzen and Lorenz 1978
P. Lorenzen and K. Lorenz, Dialogische Logik, Darmstadt: Wissenschafliche Buchgesellschaft 1978.
Neil MacCormick, Legal Reasoning and Legal Theory, Oxford: University Press 1978.
Navarro and Rodríguez 2014
Pablo E. Navarro and Jorge L. Rodríguez, Deontic Logic and Legal Systems, Cambridge: Cambridge University Press 2014.
Henry Prakken, Logical Tools for Modelling Legal Argument. A Study of Defeasible Reasoning in Law, Dordrecht: Kluwer 1997.
Prakken and Sartor 1997
Henry Prakken and Giovanni Sartor (eds.), Logical Models of Legal Argumentation, Dordrecht: Kluwer 1997.
Joseph Raz, The Authority of Law, Oxford: Clarendon Press 1979.
G. Sartor, Legal Reasoning, A Cognitive approach to the Law, Dordrecht: Springer 2005.
F. Schauer, Playing by the Rules, Oxford: Clarendon Press 1991.
M. Thompson,. Life and action, Cambridge: Harvard University Press 2008.
Waddington and Visser 2012
L. Waddington and Laura Visser, ‘Temporary Special Measures under the Women’s Convention and Positive Action under EU Law: Mutually Compatible or Irreconcilable?’ in I. Westendorp (ed.), The Women’s Convention Turned 30: Achievements, Setbacks, and Prospects, Cambridge: Intersentia 2012.
* The authors thank Gustavo Arosemena and Gabriel Lentner for many valuable suggestions that improved this article. Remaining errors are the sole responsibility of the authors, though.
 Cf. ECJ 26/63 (Van Gend & Loos).
 This does not mean to say that the interpretative step from a legislative text to a rule formulation is never inspired by a particular case.
 Koskenniemi, in the International Law Commission’s study group’s report on the fragmentization of international law, argues that “Focusing on a mere logical incompatibility mischaracterizes legal reasoning as logical subsumption. In fact, any decision will involve interpretation and choice between alternative rule-formulations and meanings that cannot be pressed within the model of logical reasoning.”We will attempt to show, in this contribution, that this depends on the model of logical reasoning chosen.
 This clause ‘given how territory works’ indicates that incompatibility of rule consequences, and therefore also rule conflicts, may depend on the existence of other rules, which exclude the co-existence of certain facts. If world history had been different, we might have ended up with a system where states could be co-owners of land and the conflict described above would not exist. This is not the case in the current political organization of the world.
We will encounter a similar phenomenon, that conflicts result from the presence of certain facts, in section 2.6. A theoretical discussion of rule consistency which assigns a central role to rules in the determination of which other rules conflict, can be found in Hage 2000 and 2005 (p. 135-158).
 For a discussion of this type of conflict, see Frontier Dispute, Judgment, I.C.J. Reports 1986, p. 554.
 Notice that this characterization of rule conflicts is ontological. It may be useful to employ different categorizations of conflicts if the intended use is not ontological.
 This characterization of rule conflicts is pragmatic. Again, it may be useful to employ different categorizations of conflicts if the intended use is not pragmatic.
 Distinctions between different kinds of rules are discussed more elaborately in Hage 2015, section 1.5.
 The following example presupposes that different legal regimes created by different legal instruments, such as the human rights regimes of the European Union and of the Convention on the Elimination of All Forms of Discrimination against Women, constitute different legal systems. Whether and to what extent this is the case is a difficult question, which cannot be dealt with in the present contribution.
 There are several exceptions to this principle of territorial scope, but they are exceptional.
 More on the notion of applicability in section 3.1.
 Avena and Other Mexican Nationals (Mexico v. United States of America ; ICJ 09/01/2003 General List No. 128)
 The discussion in this section will be confined to conflicts of compliance, and will only discuss facts as causes of rule conflicts. A more general discussion that also deals with conflicts of imposition and with rules as causes of conflicts can be found in Hage 2000 and Hage 2005, p. 135-157.
 Cf. Article 24 and 28 Convention of the Rights of the Child and Article 12 and 13 of the International Covenant of Economic, Social and Cultural Rights. That both articles contain only obligations of progressive achievement does not take away from the argument of this section that changing the background can be a tool to deal with conflicts.
 Technically speaking, derogation of a rule can be construed in two ways. The one is that an exception is made to the rule. Exceptions are not limited to rule conflicts, but can be made in any case when decisive reasons are available against applying an applicable rule. The other way to construe derogation is to assume that the derogated rule is temporarily not valid. Because only valid rules can be applicable, derogation thus construed makes the rule inapplicable.
 This holds at least from the perspective of the incorporating system. However, the mere incorporation of rules of international law into a national legal system does not make any statement about the place of the incorporated rules in the hierarchy of norms of that legal system. International law will hold itself to be above the constitution, while national law might give the incorporated rules a different status. This brings us back to the issue of subscripting, whereby the national legal system holds that it is the only relevant system because it has incorporated rules of international law, while international law might nevertheless claim relevance.
 Since not all countries use a monist system, this exercise is quite theoretical. However, it is useful to see what the effects of incorporation might be.
 As a matter of fact, incorporation makes more rule conflicts possible, because rules from different systems can only lead to conflicts of compliance, while rules that belong to the same system can also lead to conflicts of imposition.
 This emphasis on prohibition, rather than lack of competence, partially explains why the Lotus-principle – that States are allowed to do what was not explicitly prohibited – belongs to the foundations of public international law.
 Cf. Chapter II of the Draft Articles on the Responsibility of States for Internationally Wrongful Acts
 Another exception may be the invalidity of soft law, when it does not complement, but contradicts hard law. The precise nature of soft law and its relation to hard law are too complicated, however, to deal with here.
 Theoretically it is possible to live with institutional legal facts that seem incompatible, and with rules that prescribe actions that cannot be combined. But that would not be desirable.
 The interested reader may consult Hage 1997, p. 118-121.
 Even if a rule ‘flows directly’ from a source, there is a need for an interpretive step that translates the words of the source into the rule formulation, that is the conditions and the conclusion of the rule. The possibility to combine rules into a derived rule goes a step further than this relatively straightforward interpretation.
 Actually, we used this ‘derived rule’ in an example in section 2.4.
 The question whether it is possible to derive rules (the more often used tem is ‘norms’) from other rules is highly debated. For an overview, see Navarro and Rodriguez 2014, chapter 2.
 This step towards seeing legal rules as tools in the construction of legal dialogues was originally inspired by the work of the so-called ‘Erlanger Schule’ (Lorenzen and Lorenz 1978). It was introduced in legal theory through the work of Alexy (1983) and in legal logic through the work of Gordon (1995). For an overview, see Hage 2005, p. 227-264.
 This is almost true by definition, since a negative rule condition that must be proven by the opponent of the rule’s application will count as an exception to the rule. We encounter here a second definition of an exception to a rule, namely as a negative condition for the application of the rule for which the burden of proof rests on the party which does not want the rule to be applied.
 A contributory reason is a reason that may have to be balanced against other reasons. For an elaborate discussion, see Dancy 2004, chapter 2. The logic of contributory rules is described in Hage 1997, p. 130-158.
 Such a decision may itself be the result of an argument or the balancing of reasons. The reasoning involved in balancing reasons is discussed in Hage 2005, p. 122-129.
 There may, for example, be an exception to the rule that supported the exception in first instance.
 The following is a non-exhaustive list.