A. Belligerent Reprisals

Belligerent reprisals are ‘intentional violations of a given rule of the law of armed conflict, committed by a Party to the conflict with the aim of inducing the authorities of the adverse party to discontinue a policy of violation of the same or another rule of that body of law’.Footnote ² Specifically designed for the purposes of IHL, belligerent reprisals remain the only form of reprisalsFootnote ³ which may involve the use of force lawfully. Consequently, they are also the only form which still uses the term ‘reprisal’, the other lawful forms of reprisals nowadays being called countermeasures.Footnote ⁴ Belligerent reprisals have several characteristics.

First, they are a tool of law enforcement. Taken in response to a previous violation of IHL by one of the parties to a conflict, their aim is to stop the violation and ensure its non-repetition and/or reparation. They protect pre-existing legal relations (status quo ante delictum) and seek their restoration.

Second, belligerent reprisals are an instrument of jus in bello. They constitute ‘a species of the genus reprisals’Footnote ⁵ applicable only in armed conflict. They are subject to special rules which are not necessarily identical with those applicable to countermeasures in general. As the ILC pointed out, a special rule can either elaborate upon or derogate from a general rule. In the case of belligerent reprisals, both aspects are present: the regulation elaborates upon the general rule by setting particular conditions for the legality of reprisals; it derogates from the general rule in that reprisals may imply the resort to coercive force. This obviously has no implications at the jus ad bellum level, as the two areas of law are separate.Footnote ⁶

Third, belligerent reprisals constitute an instrument of self-help. As such, they belong to the enforcement tool kit of traditional international law, characterized by its horizontal nature and the lack of a central law-maker and law-enforcer. Belligerent reprisals apply among peer-subjects and are dependent on a unilateral assessment of the situation, with all the risks of misapplication based on mistake or abuse that this implies. Moreover, they rest upon the logic of collective action: a violation of IHL committed by one member of an entity may be legitimately redressed by means of repressive action taken against other members, or property, of that entity. These features can make reprisals appear outdated at a time when international law is progressing towards ‘verticalisation’Footnote ⁷ and ‘humanization’.Footnote ⁸

Fourth, belligerent reprisals are a form of circumstances precluding wrongfulness, as understood by the ILC.Footnote ⁹ This is not uniform but encompasses different types of circumstances. Lowe distinguishes between ‘behaviour that is right; and … behaviour that, though wrong, is understandable and excusable’,Footnote ¹⁰ using ‘exculpation’ and ‘excuse’ for the two cases respectively. ‘Exculpation’ covers situations in which a prima facie wrongful act turns out not to be wrongful, and hence no responsibility is at stake. ‘Excuse’ applies to cases in which a prima facie wrongful act turns out to be wrongful, yet some circumstances preclude it from generating responsibility. Belligerent reprisals belong to the former category of circumstances precluding wrongfulness stricto sensu. Thus, a state resorting to belligerent reprisals does not truly violate IHL, since an act taken in lawful reprisals is placed outside the area covered by IHL prohibitions. The primary rules of IHL do not cease to apply but are rendered temporarily inoperative. Once the reprisals have achieved their goals, these primary rules resume their force.Footnote ¹¹

Under the rules of IHL applicable to an international armed conflict (‘IAC’), belligerent reprisals are subject to relatively strict regulation. This regulation makes use of two regulatory techniques. The first consists of explicit and absolute prohibitions completely outlawing certain forms of reprisals. Such prohibitions apply—in IAC—to reprisals against the wounded, sick and shipwrecked, medical personnel, units and installations, prisoners of war, enemy civilians, civilian objects, cultural property, objects indispensable to the survival of civilians, natural environment and installations containing dangerous forces.Footnote ¹² The second technique is that of subjecting the conduct of those reprisals which are lawful to certain requirements. These requirements, never codified in a treaty, are a part of customary IHL.Footnote ¹³ Although there is not a complete consensus on the list, most sources include five requirements:Footnote ¹⁴ (a) reprisals may only be carried out in reaction to a previous violation of IHL; (b) they need to be measures of last resort; (c) they must be proportional to the original wrong and/or to the aim they pursue; (d) the decision to use them has to be taken at a high level of the military or civil leadership of the State; and (e) they need to conform to the rules of humanity and morality.

B. Non-International Armed Conflicts

The concept of non-international armed conflict or, more exactly, of ‘armed conflict not of an international character’ (Common Article 3), was introduced into IHL in the mid-twentieth century.Footnote ¹⁵ Yet although both are used in several IHL treaties it has never been defined in any binding instrument. Both the ICRCFootnote ¹⁶ and legal doctrineFootnote ¹⁷ have failed to devise a universally acceptable definition. The ICTY, however, has put forward a definition under which a NIAC is ‘protracted armed violence between governmental authorities and organised armed groups or between such groups within a State’.Footnote ¹⁸ The definition indicates that NIAC may be either vertical (ie the government against armed opposition groups) or horizontal (ie between various armed opposition groups). In all cases, there needs to be a sufficient degree of organization of parties and the conflict needs to be of a protracted character, normally assessed in terms of the intensity and duration of violent clashes.

The IHL regulation of NIAC has traditionally developed separately from that of IAC.Footnote ¹⁹ Over the past decades, however, the two bodies of regulation have been converging. Yet differences persist. NIAC are still subject to more rudimentary regulation than IAC and some institutions which are relevant for the purposes of IAC (eg combatants, prisoners of war) are not relevant for the purposes of NIAC. Belligerent reprisals are usually considered one of such institutions. NIAC themselves are not regulated by a uniform set of rules. Since the adoption of Additional Protocol II to the Geneva Conventions (‘APII’) in 1977 two different regimes have been applicable to them. Vertical armed conflicts of higher intensityFootnote ²⁰ are regulated by Common Article 3, APII, relevant ‘Hague Law’ treatiesFootnote ²¹ and customary rules. All other types of NIAC are subject to the same regulation, but with the exception of APII. While this difference was significant in the past, recent developments have reduced it. Moreover, practice shows that the distinction between the two types of NIAC has never become operative in practice, as most States prefer to have just one set of rules applicable to any civil strife.

C. Three Approaches to Belligerent Reprisals in Non-International Armed Conflicts

There are three ways of conceptualizing the relationship between belligerent reprisals and non-international armed conflicts from a legal perspective. The ‘extralegal’ approach postulates that belligerent reprisals do not form part of the legal regime of NIAC at all. As an inter-State tool, their use is confined to IAC. Consequently, there is no need or possibility of regulating belligerent reprisals in the course of NIAC, and all actions undertaken are to be assessed against primary norms of IHL. The ‘permissive’ approach asserts that belligerent reprisals do form part of the legal regime of NIAC and there are no legal limits imposed on them. On this basis, parties to NIAC (or at least some of them) are free to use reprisals without any legal impediments. The ‘restrictive’ approach also assumes that belligerent reprisals form part of the legal regime of NIAC; yet it claims that they are subject to legal restraints. Each of these three approaches received endorsement during the largest ever debate on the topic, held at the 1974–77 Diplomatic Conference on Humanitarian Law in Geneva. Each of them also enjoys some support from IHL scholars and, occasionally, other actors.

A. The ICTY

The two leading ICTY cases relating to belligerent reprisals in NIAC are the Martić Footnote ²² and the Kupreškić cases.Footnote ²³ The Martić case dealt with the shelling of Zagreb by the armed forces of the Republika Srpska Krajina (‘RSK’) in Croatia, which resulted in a number of civilian deaths. The former president of the RSK, Milan Martić was accused of having planned the shelling, and thereby committing the crime of attacks on civilians under Article 3 of the ICTY Statute. The Kupreškić case involved the accusation that several Bosnian Croats, including Zoran Kupreškić, were responsible for the ethnic cleansing of Bosnian Muslims in the Lašva Valley, including the massacre of some 100 civilians in the village of Ahmiči. The massacre was, again, qualified as the crime of attacks on civilians.

In the Martić case, the defendant expressly raised the issue of reprisals. He admitted ordering the shelling on Zagreb but claimed it was a ‘lawful reprisal, carried out with the aim of putting an end to violations of international humanitarian law committed by the Croatian … forces’.Footnote ²⁴ In the Kupreškić case, the defendants invoked the tu quoque principle, but the Court found it useful to discuss reprisals as well. In both cases, the ICTY came to the conclusion that the acts could not be justified as lawful reprisals because they had not met the legal requirements. The attack on Zagreb ‘was not carried out as a last resort’ and ‘no formal warning was given prior to the shelling’.Footnote ²⁵ The massacre of Ahmiči failed to respect the prohibition of reprisals against the civilian population.Footnote ²⁶

This reasoning shows that the ICTY supports the ‘restrictive’ approach. The Tribunal believes that certain reprisals, namely those directed against the civilian population, are prohibited ‘in all armed conflicts’.Footnote ²⁷ It also claims that ‘it is not necessary … to determine whether the armed conflict was international or internal’.Footnote ²⁸ If the nature of the conflict does not matter, then reprisals are clearly—albeit implicitly—held applicable in both IAC and NIAC. Moreover, since the two cases relate to actions by armed opposition groups, the ICTY does not believe that reprisals can only be undertaken by States. The cases also show that the ICTY does not consider the right to reprisals to be unlimited. Although the Tribunal only discusses specific instances of alleged reprisals, it is clear that it considers the traditional conditions (last resort, proportionality) to be applicable and certain forms of reprisals (eg against civilians) to be as outlawed.

The ICTY also seeks to identify the legal basis for the regulation of reprisals in NIAC. Whilst it looks at both treaties and customary IHL, it has a clear preference for the latter. Treaty sources are mentioned only cursorily in Kupreškić. The ICTY notes that ‘whether or not the armed conflict of which the attack on Ahmiči formed part is regarded as internal, indisputably the parties to the conflict were bound by the relevant treaty provisions prohibiting reprisals’.Footnote ²⁹ While the decision is not explicit on this point, it seems that the ICTY had in mind Common Article 3 and Article 4 of APII when speaking of ‘the relevant treaty provisions’. Yet it uses treaty sources in a subsidiary way and customary IHL is the main focus of attention. Although the ICTY is persuaded that the limiting conditions are ‘well established in customary law’Footnote ³⁰ the evidence it gives for this relates to IAC, not NIAC. Indeed, the ICTY explicitly draws from the practice relating to what became Rule 145 of the 2005 ICRC Study; a rule, which, however, is only established ‘as a norm of customary international law … in international armed conflicts’.Footnote ³¹ It is not clear whether the ICTY thinks that the same rule is also applicable for NIAC or whether it believes that rules of IAC automatically apply in NIAC.

The analysis of the prohibition of reprisals against civilians is more detailed but no less confusing. The ICTY simply declares that a customary rule has developed prohibiting reprisals both against civilians in the hands of an adversary and against those in the combat zones.Footnote ³² While admitting that practice is scarce, the ICTY relies heavily on the opinio juris which has allegedly developed over the past decades. To corroborate its view, the Tribunal quotes a variety of sources including: provisions of the GCs and API prohibiting reprisals; primary rules of IHL prohibiting attacks against civilians (CA3, Article 4 of APII); Common Article 1 of the GCs; UN General Assembly resolution 2675; the Martens clause; human rights law. The ICTY has no doubt that the prohibition of reprisals against civilians applies in all armed conflicts. Indeed, its assertion that ‘it would be absurd to hold that while reprisals against civilians entailing a threat to life and physical safety are prohibited in civil wars, they are allowed in international armed conflicts’Footnote ³³ suggests that the ICTY considers the rule to be less controversial in civil wars than in inter-State wars.

The ICTY decisions have given rise to criticism, largely focusing on the controversial status of the prohibition of reprisals against civilians in IAC. In contrast, its conclusions regarding NIAC have been largely undisputed, with Greenwood expressly stating that ‘if the approach … to the question of reprisals in an international conflict is flawed, there nevertheless remains their comments on reprisals in non-international conflicts’.Footnote ³⁴Sections III–V below question this opinion.

B. The ICRC

The ICRC is an internationally recognized guardian of IHL. As such, it is viewed as an important source of information on IHL. Yet the position of the ICRC regarding belligerent reprisals in NIAC is ambivalent. It oscillates between the belief that reprisals are not applicable in NIAC (‘extralegal’ approach) and the fear that inapplicability could be wrongly interpreted as an absolute freedom to use reprisals (along the lines of the ‘permissive’ approach). This fear has resulted in the ICRC trying to complement the ‘extralegal’ approach with elements of the ‘restrictive’ approach, claiming that if reprisals were applicable in NIAC, they would need to be subject to legal limitations. This line of reasoning has been followed since the 1950s, through the 1974–77 Conference and up to the 2005 ICRC Study.Footnote ³⁵

The ICRC Study clearly shows the ICRC's ambiguous stance on the issue. Out of the 161 rules listed in the Study, the vast majority are considered to apply in both IAC and NIAC. Among the few rules for which this is not true are those in Chapter 41 (Rules 144–148) which focus on the enforcement of IHL and, primarily, on belligerent reprisals. Rules 145–147 recall the limitations imposed on lawful reprisals and the prohibition of reprisals directed against protected persons, and protected objects. These rules are said to apply in IAC only. Reprisals in NIAC are the object of a special rule, Rule 148. This Rule consists of two parts.

The first part states that ‘parties to non-international armed conflicts do not have the right to resort to belligerent reprisals’. Here, the Study adheres to the ‘extralegal’ approach denying the applicability of reprisals in NIAC. Explaining its approach, the ICRC claims that ‘there is insufficient evidence that the very concept of lawful reprisal in non-international armed conflict has ever materialized in international law’.Footnote ³⁶ Yet, having discarded the applicability of reprisals in NIAC, the ICRC goes on to claim that ‘reprisal which entails one of these acts [prohibited by CA3] is prohibited’; ‘any reprisal … incompatible with this requirement of humane treatment is … prohibited’; and ‘Article 4 of Additional Protocol II similarly allows no room for reprisals’.Footnote ³⁷ This represents a shift to the ‘restrictive’ approach, under which reprisals in NIAC are applicable but restricted by primary norms of IHL (CA3 and Article 4 of APII). At this point, the ICRC contradicts itself: belligerent reprisals in NIAC cannot be both inapplicable and prohibited at the same time. There are two ways in which this contradiction can be overcome.

First, it is possible that the ICRC wants to stress that since reprisals are not applicable in NIAC, acts labelled as such remain within the domain of primary norms of IHL. This assertion would be correct; yet, it is hardly compatible with the fact that the ICRC uses the term ‘reprisal’ here. Second, the ICRC may wish to show that though the institution of reprisals has not materialized in NIAC, practice indicates that were it to materialize, some acts—those in breach of CA3 and Article 4 of APII—would be impermissible. Thus the ‘restrictive’ approach would serve as a subsidiary safety net, ensuring that the ‘permissive’ approach could never prevail: reprisals in NIAC are either inapplicable or, if applicable, are subject to limitations derived from primary IHL rules. The second approach to understanding the first part of Rule 148 is more persuasive.

The second part of Rule 148 states that ‘other countermeasures against persons who do not or who have ceased to take a direct part in hostilities are prohibited’. The term ‘other countermeasures’ is not explained. It seems reminiscent of the discussions at the 1974–77 Conference, where some States suggested that ‘reprisal’ should be replaced by a more neutral term. It is not entirely clear what would be gained by this linguistic turnabout. The institution at issue, whatever called, would be subject to the same paradox as the first part of the Rule, since, as Turns highlights, it is ‘hard to see how something that does not even exist as a concept can be the subject of a specific customary rule prohibiting it’.Footnote ³⁸ One might argue that the term ‘countermeasures’ does not refer to reprisals but to non-forcible measures under Article 22 of the ILC Articles on the Responsibility of States for Internationally Wrongful Acts (‘ARSIWA’).Footnote ³⁹ Yet such measures could hardly be directed ‘against persons who do not or who have ceased to take a direct part in hostilities’. The most probable explanation for the presence of this part of the Rule lies, again, in the attempts of the ICRC to prevent any possible deviations from fundamental rules applicable in NIAC.

The same rationale most likely accounts for the very inclusion of Rule 148 in the Study. Since Rule 148, which declares the non-applicability of belligerent reprisals in one type of conflict, is more of a non-rule than a rule, the Study could well do without it, given that the purpose of the Study was to ‘establish what rules of customary international law can [rather than cannot] be found inductively’.Footnote ⁴⁰ Yet the rule and its commentary are useful in that they reveal the ICRC's view on reprisals in NIAC.

C. Legal Doctrine

So far, legal doctrine has paid limited attention to belligerent reprisals in NIAC. The two main authors writing on the topic, Kalshoven and Darcy, both admit to not having a firm opinion on the subject. Both have doubts whether reprisals are more than ‘an interstate mechanism … of no relevance in noninternational armed conflicts’Footnote ⁴¹—thus expressing some sympathy for the ‘extralegal’ approach. Conceding, however, that ‘it cannot be stated categorically that the doctrine of belligerent reprisals is not of relevance in noninternational armed conflicts’,Footnote ⁴² they are, at the same time, afraid that the ‘permissive’ approach could prevail.Footnote ⁴³ This makes them, albeit reluctantly, speak in favour of the ‘restrictive’ approach, and to suggest, even more reluctantly, that reprisals in NIAC could be subject to limitations imposed by Common Article 3 of GCs and APII,Footnote ⁴⁴ human rights,Footnote ⁴⁵ rules on reprisals applicable in IAC and used per analogia,Footnote ⁴⁶ or practical and moral arguments.Footnote ⁴⁷ This position is similar to that of the ICRC.

Authors who have considered the issues less frequently tend to be slightly more categorical about it. Most favour the ‘restrictive’ approach, claiming that belligerent reprisals are applicable in NIAC and that they can be resorted to by both States and armed opposition groups. The applicability of reprisals is seen as the ‘necessary consequence’Footnote ⁴⁸ of the fact that parties to NIAC are bound by primary norms of IHL. Those scholars generally agree that the right to resort to reprisals in NIAC is not unlimited. They disagree, however, over the legal basis of the limits imposed. Greenwood sides with the view expressed by the ICRCFootnote ⁴⁹ whereas Cassese refers to the very nature of some of the primary obligations which are, in his view, absolute and allow no derogations.Footnote ⁵⁰

Quénivet admits that ‘there is no ready-made answer’Footnote ⁵¹ as far as the source of the regulation is concerned and, ultimately, suggests that the rules applicable to reprisals in IAC apply by analogy, though she warns that ‘this always needs to be done with great caution’. MoirFootnote ⁵² and ZegveldFootnote ⁵³ invoke primary norms of IHL, especially Common Article 3 and Article 4 of APII. Jones and Newton offer dissident voices: Jones favours the ‘extralegal’ approach, believing that ‘reprisals are exacted against other nations or international actors’Footnote ⁵⁴ and so are inapplicable to NIAC. Newton seems to have some sympathy for the ‘permissive’ approach, with the right to use reprisals reserved to States.Footnote ⁵⁵

A. Convention on the Protection of Cultural Property in the Event of Armed Conflict (1954)

None of the IHL treaties adopted at the turn of the nineteenth and twentieth centuries (1899 and 1907 Hague Conventions) and after the World War II (1949 Geneva Conventions) contained provisions on belligerent reprisals in NIAC. Most of these conventions did not deal with belligerent reprisals at all and those which did, such as the 1949 GCs, confined their regulation to IAC. While provisions outlawing reprisals against protected persons appear in all four Geneva Conventions,Footnote ⁵⁶ no similar provision is included in Common Article 3 (‘CA3’), the only provision of the GCs applicable in NIAC. The travaux préparatoires show that this omission was not the outcome of a conscious decision concerning the non-applicability of belligerent reprisals in NIAC. Rather, it was the result of the question having been neglected and not discussed.Footnote ⁵⁷

The first treaty which is sometimes seen as dealing explicitly with the issue is the 1954 Cultural Property Convention.Footnote ⁵⁸ Although most provisions of the Convention apply only in IAC, Article 19 provides that ‘in the event of an armed conflict not of an international character occurring within the territory of one of the High Contracting Parties, each party to the conflict shall be bound to apply, as a minimum, the provisions of the present Convention which relate to respect for cultural property’. The main provision concerning respect for cultural property, and which therefore should apply in all armed conflicts, is Article 4 (Respect for Cultural Property). Article 4 refers to belligerent reprisals, stating that ‘they (Parties/parties to the conflict) shall refrain from any act directed by way of reprisals against cultural property’ (para 4). Prima facie, the 1954 Convention seems to indicate that belligerent reprisals form part of the legal regulation of NIAC under IHL.

This opinion is shared by several authors. Toman claims that ‘State Parties to the Hague Convention are bound by the prohibition [of reprisals] in both international and non-international conflicts’.Footnote ⁵⁹ O'Keefe believes that Article 4(4) is ‘applicable via article 19 to both international and non-international armed conflicts’.Footnote ⁶⁰ The ICRC takes a less categorical stance. It places Article 4(4) of the 1954 Convention under the rubric of ‘Practice’ relating to Rule 148 (Reprisals in Non-International Armed Conflicts), but does not comment on it or draw any specific conclusions from it.Footnote ⁶¹

The travaux préparatoires of the Convention confirm the ICRC's cautious approach. Article 19 was inserted into the draft at the stage when Article 4, containing a prohibition of reprisals, bore a general title (Obligations in respect of cultural property situated within the territory of another Contracting Party) and the reference to respect for cultural property figured only in its first paragraph. It can be argued that at that time obligations relating to ‘respect’ were primarily assembled in Draft Article 2, which did not contain any reference to reprisals. Thus, Kalshoven is right to claim that ‘it was not the intention of the authors of the … draft that the prohibition of reprisals … would be counted among the provisions relating to respect for cultural property’ and that ‘this original intention was not modified in the course of the subsequent proceedings’.Footnote ⁶²

Consequently, the 1954 Convention neither confirms nor discards the applicability of reprisals in NIAC. Those drafting the text did not intend to take any position on this issue. They most probably did not consider the issue at all, as is suggested by the lack of any debate on it. This is corroborated by the fact that Article 19 of the 1954 Convention was inspired by Common Article 3 of the 1949 GCs, which does not contain any reference to reprisals. When the issue of reprisals in NIAC finally came up in the 1974–77 Conference, it gave rise to strong disagreements. It is unlikely that those disagreements would have arisen for the first time in the 1970s, if there had been general agreement 20 years before. It is equally unlikely that States would disagree over the applicability and legality of reprisals against the civilian population and civilian objects (matters which were discussed in 1974–77) but would agree on the applicability and illegality of reprisals against cultural property. Thus, Article 4(4) of the 1954 Convention was not intended to apply in NIAC and cannot give any guidance as to whether belligerent reprisals are applicable in these conflicts.

B. Additional Protocol II to the Geneva Conventions (1977)

The most active discussion of incorporating provisions on belligerent reprisals into an IHL treaty regulating NIAC occurred during the 1974–77 Diplomatic Conference on Humanitarian Law held in Geneva. The conference drafted and adopted the texts of the two 1977 Additional Protocols to the Geneva Conventions. The initial drafts of these Protocols, relating respectively to the protection of victims of IAC (DPI) and NIAC (DPII), had a similar structure and both contained provisions on belligerent reprisals.Footnote ⁶³ DPI aimed at filling the gaps left in the regulation of reprisals by the previous instruments.Footnote ⁶⁴ DPII was, in contrast, intended to be the first instrument regulating reprisals in NIAC. DPII contains three relevant provisions, focussing on three main categories of persons to be protected against reprisals, namely detained persons;Footnote ⁶⁵ the wounded, the sick, the shipwrecked and the medical personal;Footnote ⁶⁶ and the civilian population.Footnote ⁶⁷

In 1976, the ICRC suggested that the three articles be complemented with or replaced by a general provision (Draft Article 10bis) containing an en bloc ban on reprisals against persons not taking a direct part in hostilities or hors de combat. Footnote ⁶⁸ It was decided that until a decision was taken on this proposal, the work on the three original provisions would be suspended. This temporary suspension turned into a definitive one, once Draft Article 10bis was adopted by one of the Committees of the Conference in May 1977Footnote ⁶⁹ and passed to the Plenary Meeting. In its fifty-first session held in June 1977, the Plenary rejected Draft Article 10bis,Footnote ⁷⁰ doing away with the only remaining provision on reprisals in DPII. At the same time, negotiations on this instrument were coming to a dead end, as more and more States started having doubts about the usefulness of a treaty on NIAC. At the last moment, the situation was saved by Pakistan, which put forward a radically abridged text of APII.Footnote ⁷¹ The text reflected the bare humanitarian minimum, omitting all controversial issues, including reprisals.Footnote ⁷² When in June 1977 Pakistan's proposal secured consensus, there was, then, no reference to reprisals.

The silence of APII on reprisals, together with the rejection of Article 10bis by the Plenary Meeting, should not be automatically read as confirming that belligerent reprisals were considered to have no place in the legal regulation of NIAC. The debates during the 1974–77 Conference indicate that the positions were more nuanced, with, in simple terms, States being divided into two groups, though the line between them was anything but firm, and with a high degree of heterogeneity within each group.

The first group consisted of States which were in favour of including provisions on belligerent reprisals in APII and so could be considered supporters of the ‘restrictive’ approach. Some of these States supported the original 1973 ICRC Draft (three specific provisions on reprisals); others preferred the 1976 ICRC proposal of Article 10bis (one general provision); and still others put forward their own suggestions. Out of all these States, only two (FinlandFootnote ⁷³ and SwedenFootnote ⁷⁴) made it clear that they considered reprisals applicable in NIAC. Other countries supporting inclusion (the German Democratic Republic,Footnote ⁷⁵ PolandFootnote ⁷⁶ and RomaniaFootnote ⁷⁷) simply referred to humanitarian considerations or gave no explanation of their position at all. Even the ICRC, though it drafted most of the provisions on reprisals, failed to offer a clear view on the legal status of reprisals in NIAC and preferred to emphasize the practical, humanitarian side.Footnote ⁷⁸

States belonging to the second group believed that APII should not contain provisions on belligerent reprisals or, at least, not the provisions put forward during the Conference. The rationale behind this position differed. Two countries (CameroonFootnote ⁷⁹ and NigeriaFootnote ⁸⁰) implied that they opposed the express prohibition because of their belief that reprisals could be resorted to by governments fighting internal rebellion. Thus, they agreed that reprisals were applicable in NIAC, yet they reserved their use for States. This was close to the ‘permissive’ approach for States, combined with an ‘extralegal’ approach for armed opposition groups. All the other countries in the second group abhorred the practice of reprisals, their objection to the presence of provisions on reprisals in APII being motivated by arguments other than sympathy for the institution.

Some States (AustraliaFootnote ⁸¹ and MexicoFootnote ⁸²) argued that the incorporation of provisions on reprisals into APII would be legally erroneous, because the institution was only applicable in IAC—in line with the main premise of the ‘extralegal’ approach. Other countries (Belgium,Footnote ⁸³ UKFootnote ⁸⁴ and the USAFootnote ⁸⁵) considered such provisions unnecessary and superfluous. They believed that reprisals were adequately addressed by the primary norms of IHL which applied ’at any time’ (Common Article 3)—thus favouring the ‘restrictive’ approach. Finally, several States (IndonesiaFootnote ⁸⁶ and IraqFootnote ⁸⁷) thought that reprisals should not be regulated in APII because they were an internal matter, to be legislated upon by individual countries. This position was close to that of Cameroon and Nigeria and its supporters could be viewed as mainly in favour of the ‘permissive’ approach.

Other States belonging to the second group opposed the provisions on reprisals for political reasons. For instance, the Federal Republic of Germany claimed that ‘there were no objections from the legal point of view to the use of the word “reprisal”, but from the political point of view it could be inferred that its use gave the Parties to a conflict a status under international law which they had no right to claim’.Footnote ⁸⁸ This position seems to go in the direction of the ‘restrictive’ approach. Finally, some States (SyriaFootnote ⁸⁹) opposed the inclusion of prohibitions of specific categories of reprisals in APII out of the fear that it could be interpreted as an a contrario approval of other categories of reprisals. This view is in line with the ‘restrictive’ approach, since it considers reprisals applicable to NIAC but subject to legal regulation.

The final text of APII does not contain any reference to belligerent reprisals. From that perspective, the situation after 1977 was prima facie similar to that after 1949 or 1954. However, in this case the silence did not result from the neglect of the question but from the inability of States to reach consensus on whether reprisals had any place in NIAC and, if so, what legal regulation was to be imposed on them. All three theoretical approaches—‘extralegal’, ‘permissive’ and ‘restrictive’—had their supporters among States present at the conference. Following the opinion expressed by Iraq that ‘the principle of prohibiting reprisals in non-international conflicts was controversial, and no controversial concepts should be introduced into draft Protocol II’,Footnote ⁹⁰ the drafters of APII decided to leave the questions of reprisals open.

C. Protocol II to CCW on Prohibitions or Restrictions on the Use of Mines, Booby-Traps and Other Devices (1980/1996)

Until the late 1990s, the ‘Hague law’ treaties, with the exception of the 1954 Convention, were inapplicable in NIAC. None of them, moreover, except the 1954 Convention, contained provisions on reprisals. This changed in 1996 when an amended version of 1980 Protocol on Prohibitions or Restrictions on the Use of Mines, Booby-Traps and Other Devices (Protocol II to CCW) was adopted. Article 3(2) of the 1980 Protocol made it unlawful ‘in all circumstances to direct weapons to which this Article applies, either in offence, defence or by way of reprisals, against the civilian population as such or against individual civilians’. The original 1980 version of the Protocol applied only to IAC only but the 1996 amended version applies to both IAC and NIAC.Footnote ⁹¹

The amended text contains a provision on belligerent reprisals, which is even more extensive than the original Article 3(2). The new Article 3(7) states that ‘it is prohibited in all circumstances to direct weapons to which this Article applies, either in offence, defence or by way of reprisals, against the civilian population as such or against individual civilians or civilian objects’. In 2001, the 1980 Convention (CCW) itself was amended and became, together with its Protocols I–IV, applicable in NIAC. Thus both the 1996 amended Protocol II to CCW and the original 1980 version now applies to all armed conflicts. Interestingly, this makes Articles 3(2) of the 1980 Protocol II to CCW and 3(7) of the 1996 Amended Protocol II to CCW the first ever rules of IHL to confirm (or extend) the applicability of belligerent reprisals to NIAC.

One may, however, have doubts as to whether the drafters of the 1996 and 2001 amendments truly intended or, indeed, noticed this effect. One may also argue that while the scope of the application of Protocol II to CCW has been extended to NIAC, some of the provisions may still apply to States only. The 1996 amended Protocol II itself distinguished between provisions which apply to ‘High Contracting Parties’, and other provisions which, arguably, apply to all parties to a conflict. The provision on belligerent reprisals makes no reference to ‘High Contracting Parties’ and so falls into the latter category. The unavailability of the travaux préparatoires and the lack of any substantive practice linked to the application and interpretation of Articles 3(2) of the 1980 Protocol II to CCW and 3(7) of the 1996 Amended Protocol II makes it impossible to confirm or reject this view in a definitive way. Thus, caution should be used before any far-reaching conclusions are drawn from the developments linked to Protocol II to CCW.

A. State Practice and Opinio Juris

Drawing partly on the materials collected in the 2005 ICRC StudyFootnote ⁹⁶ and partly on the materials found elsewhere, this subsection offers an independent analysis of the ‘the actual practice and opinio juris of States’.Footnote ⁹⁷ The data used encompassed actual battlefield behaviour by States, special agreements, official statements, national legislation, military manuals, and resolutions by UN bodies. So far, there are no recorded instances of the resort to reprisals by States in NIAC. While the media and NGOs occasionally use the term ‘reprisals’ to refer to State actions against armed opposition groups and their leaders,Footnote ⁹⁸ States themselves refrain from employing this terminology. Usually there is no reason why they should; the term is used in many contexts, often referring to acts of retaliation or revenge that have nothing to do with legal reprisals.

The fact that States do not resort to reprisals in NIAC, or at least do not publicly claim to do so, does not necessarily mean that they are persuaded that they lack the right to do so. Other interpretations are equally plausible. States might, for instance, believe that they have this right, but that in the given circumstances it would be legally or factually inappropriate to exercise it. They may also believe (or pretend to believe) that the situation in their territory does not reach the threshold of an armed conflict and, therefore, IHL is not applicable. Finally, from the practical point of view, there is also the possibility that States believe they have the right to undertake reprisals, and even occasionally exercise this right, but that they call the institution by a different term, making its use or invocation difficult to detect, and/or they do not publicize their position. The simple absence of any recorded battlefield practice does not in itself allow us to decide between these alternative explanations, and it is therefore necessary to look into other manifestations of State practice.

One of these manifestations consists in special agreements, concluded by parties to a concrete NIAC. The main purpose of these agreements is to extend the circle of applicable norms and to bring into force those parts of IHL which are otherwise inapplicable in this type of conflict. The conclusion of special agreements is explicitly foreseen by Common Article 3 and Article 19 of the 1954 Convention.Footnote ⁹⁹ During the Cold War, such agreements were concluded in Yemen in 1967 and in Afghanistan in the mid-1980s,Footnote ¹⁰⁰ though no details are available. In the post-Cold War period, they were used in the conflicts in Croatia,Footnote ¹⁰¹ Bosnia and HerzegovinaFootnote ¹⁰² and the Philippines.Footnote ¹⁰³ Certain ceasefire and peace agreements also contain provisions similar to those of special agreements.Footnote ¹⁰⁴ Belligerent reprisals are mentioned in several agreements. The Philippine agreement lists ‘all acts of violence and reprisals’ among acts which ‘are and shall remain prohibited at any time and in any place’.Footnote ¹⁰⁵ The Croatian and Bosnian agreements require that hostilities be conducted in accordance with Articles 48–58 of the 1977 API, containing several prohibitions of reprisals against protected property.Footnote ¹⁰⁶ These provisions do not tell us whether States (and armed opposition groups) concluding the agreements believed reprisals were or were not part of the legal regulation of NIAC. However, they do indicate that they considered reprisals as potentially applicable in these conflicts, and what limits they would place upon them.

Official statements count as a form of State practice, whether they are made at international conferences or on other occasions.Footnote ¹⁰⁷ As was shown in Section III above, the 1974–77 Conference, which witnessed the most intense debate about the topic, gave rise to divisions both as to the applicability and as to the desirable regulation of belligerent reprisals in NIAC. No similar conference has been held since the 1970s to tell us whether, and potentially how, the positions of States have changed over the past 40 years. The absence of battlefield practice by States means that there are no reactions by other States. What is more interesting is that States do not feel the need to react to armed opposition groups claiming the right to use reprisals. That States do not reject such a claim, even if it directly concerns them, might be prima facie interpreted as their condoning of it. Yet one should be cautious before adopting this view. It is more natural that States remain silent simply because they do not take their non-state enemies’ legal arguments seriously enough; or that they deny the very existence of an armed conflict in their territory.

National legislation is also scarce. It encompasses only the Lieber Code (1863), signed by the US President Lincoln in the course of the US civil war. While not speaking explicitly about ‘reprisals’, the Lieber Code contains two articles on ‘retaliation’, which effectively relate to reprisals. Calling retaliation ‘the sternest feature of war’, Article 27 claims that ‘a reckless enemy often leaves to his opponent no other means of securing himself against the repetition of barbarous outrage’. Article 28 confirms that retaliation is ‘a means of protective retribution’ and adds that it can only be used as a measure of last resort, after careful consideration of all circumstances and taking into account the risk of escalation. The precedential value of the Lieber Code is limited not only by its age but also because it was drafted for a conflict, which though originally non-international, witnessed the recognition of belligerency. This recognition made the legal regime of IAC applicable to the US Civil war.

Military manuals are not particularly revealing either. None of them expressly excludes or confirms the applicability of reprisals in NIAC; most connect the institution with IAC only. Some manuals refer explicitly to States, Governments, Parties or Nations when specifying who is entitled to use reprisals, and against whom.Footnote ¹⁰⁸ Other manuals are less outspoken. Yet their use of terms such as combatants which only exist in the legal regulation of IAC suggests that they also see reprisals as an inter-State tool.Footnote ¹⁰⁹ Finally, some manuals, mainly from non-Western States, give no indication of their scope of application.Footnote ¹¹⁰ It is interesting to note that the manuals coming from the States which in 1974–77 supported the applicability of reprisals in NIAC do not make any express provision for this.

Resolutions adopted by the UN General Assembly and other UN bodies can also count as State practice.Footnote ¹¹¹ The General Assembly has so far adopted three resolutions which could be of relevance. Resolution 2675 (XXV) lists the basic principles for the protection of civilian populations in armed conflicts. One of those principles applicable both in IAC and NIAC states that ‘civilian populations or individual members thereof, should not be the object of reprisals’ (para 7). The other resolutions, 48/152 (1993) and 49/207 (1994) deal with the situation in Afghanistan. They call upon parties to the conflict ‘to respect accepted humanitarian rules, as set out in the Geneva Conventions … and the Additional Protocol … to protect all civilians from acts of reprisal’ (section 8 and section 9 respectively). This call was later repeated by the UN Security Council in the Resolution 1378 of 2001 (section 2). There are also references to the prohibition of reprisals against the civilians in the resolutions of the UN Commission on Human Rights adopted with respect to armed conflicts in Afghanistan,Footnote ¹¹² the DRCFootnote ¹¹³ and Somalia.Footnote ¹¹⁴

Taken at face value, the resolutions seem to convey a belief that reprisals can be used in NIAC (by all the parties) and that some of their forms, typically reprisals against the civilian population, are prohibited. Yet this conclusion should not be embraced hastily and without the consideration of several factors. Those factors include—in addition to the non-binding nature of resolutions—the tendency of the UN organs to use legal terminology in a loose way,Footnote ¹¹⁵ their inclination to address simultaneously IHL and human rights, and the fact that some resolutions deal with both IAC and NIAC. Moreover, the resolutions appear to be primarily motivated by humanitarian concerns, having for their purpose to affirm and uphold the absolute nature of some rules of IHL and human rights, rather than to pronounce on the modalities of the reprisal regime under IHL. From that perspective, they might indicate which limits would be imposed on reprisals, if those were applicable in NIAC. Yet they do not tell us whether this is the case.

B. Armed Opposition Groups Actions and Beliefs

Armed opposition groups seem more inclined to use or threaten to use what they (rightly or not) qualify as reprisals. The oldest recorded example goes as far back as to the 1860s, to the Mosby incident. The incident occurred during the US Civil War. In 1864 a Confederate fighter, Colonel John Mosby, responded in kind to seven of his men being hanged by Union troops and had seven Union soldiers sentenced to death and executed. In a letter addressed to the Union military commander, General Sheridan, Colonel Mosby declared that captured Union soldiers would be treated humanely ‘unless some new act of barbarity’Footnote ¹¹⁶ should occur. The letter made it clear that the execution of the Union soldiers occurred in reaction to the previous act of the Union forces and had for its aim to dissuade these forces from committing new ‘acts of barbarity’. After the defeat of the Confederation, the lawfulness of Colonel Mosby's action was not questioned, though he himself expressed regret over the painful decision he had had to take.

At first sight, the Mosby incident is a clear example of relevant practice accompanied by a belief (expressed by Mosby and tacitly condoned by the Union) that reprisals could be applied in NIAC. Yet the situation is not so straightforward. First, the incident occurred almost 150 years ago, in a different legal setting, a century before the IHL legal regime for NIAC was created. Moreover, the US civil war witnessed the recognition of belligerency, so the law applied in it was actually that of IAC. Finally, the incident was rather isolated, low-profile and did not provoke any official comments by either the Union's or the Confederate's leaders. Bearing these caveats in mind, the Mosby incident still has the merit of showing that in the nineteenth century, parties to a NIAC were thought to be able (and legally entitled, once the belligerence was recognized) to resort to reprisals. It also illustrates that under certain circumstances, such reprisals could work.

It is possible that reprisals were used, or claimed, during the Spanish Civil war (1936–39) and various civil wars of the post-WWII period, yet no records are available to confirm or discard this. An exception relates to the war in Algeria in the 1960s, in the course of which the Front de Libération Nationale (‘FLN’) denounced violations by French forces claiming that ‘it will be impossible for the FLN to respect the laws of war, if France persists in ignoring them’.Footnote ¹¹⁷ Post-Cold War period has turned out to be much richer in, or better documented regarding, the use of reprisals by armed groups. Several instances of the practice were recorded during the civil war in the former Yugoslavia. Two of them, later considered by the ICTY, gained particular prominence. The first instance involved the shelling of Zagreb by the armed forces of the Republika Srpska Krajina allegedly committed to stop violations of IHL by the Croatian forces.Footnote ¹¹⁸ The other instance concerned fighting in the Lašva Valley between the Bosnian Croats and Muslims. The Croats first carried out a massacre in the Muslim village Ahmiči resulting in the death of 103 civilians, claiming reprisals for previous attacks by Muslims upon Croatian civilians. In response, and claiming reprisals for Ahmiči, Muslim forces committed cruelties against Croatian civilians in the village of Miletiči.Footnote ¹¹⁹

As in the Mosby incident, there are several factors which call for caution in assessing the Yugoslav cases. First, the claims of reprisals were often made ex post facto, during the prosecution before the ICTY. While this does not necessarily mean that they do not reflect how the actors perceived the events at the time, it does not show that it did either. Second, it is disputable whether, during the war, armed opposition groups had a clear idea what belligerent reprisals meant. Declarations by their leaders suggest that they did not see a true difference between reprisals and vengeance and often used the former term when actually speaking about the latter concept. Finally, the war in Yugoslavia was an internationalized armed conflict, encompassing aspects of IAC and NIAC that are not easy to distinguish. Armed groups taking part apparently saw themselves as fighting on behalf of States or State-like entities.

The past two decades have seen other instances of the claimed use of reprisals by armed opposition groups in Afghanistan,Footnote ¹²⁰ Colombia,Footnote ¹²¹ Iraq,Footnote ¹²² Sri LankaFootnote ¹²³ and the Sudan.Footnote ¹²⁴ Most of them are vulnerable to the same caveats as the instances from the Yugoslav war. Moreover, since information often comes solely from media, it is not possible to verify how exactly armed opposition groups described and justified their actions. As was mentioned in the previous subsection, some NIAC of the second half of the twentieth century witnessed the conclusion of special agreements. In some cases, these agreements (Croatia, Bosnia and Herzegovina, and the Philippines) explicitly or implicitly regulate belligerent reprisals. The remarks made with respect to these agreements in the previous subsection are equally applicable here. There are no official declarations by armed opposition groups that would explain their position on the applicability of belligerent reprisals in NIAC.

A. Rules of IHL Implicitly Regulating Belligerent Reprisals in NIAC?

The main candidates for the implicit regulation of belligerent reprisals in NIAC in IHL are the primary norms applicable in NIAC, rules prohibiting reprisals in IAC and Common Article 1 of the GCs. Of the primary norms of IHL, Common Article 3 of the GCs and Article 4 of APII are the most frequently cited. It is claimed that these provisions, though not mentioning reprisals, implicitly prohibit them. This is inferred from three facts: first, that the provisions apply ‘in all circumstances’ (CA3, Article 4(1)) or ‘at any time and in any place’ (Article 4(2)); second, that Article 4(2)(b) of APII prohibits collective punishment; and third that the provisions are of absolute nature. The ICRC considers this to mean that the acts referred to are ‘prohibited absolutely and permanently, no exception or excuse being tolerated. Consequently, any reprisal which entails one of these acts is prohibited’.Footnote ¹²⁶ This view is shared by the ICTY,Footnote ¹²⁷ the ILCFootnote ¹²⁸ and IHL scholars.Footnote ¹²⁹

Yet not everyone is convinced. As for the first argument, Kalshoven rightly recalls that nothing in the travaux préparatoires of the GCs and APII indicates that States had reprisals in mind when drafting and adopting CA3 and Article 4 of APII.Footnote ¹³⁰ Indeed, they almost certainly did not, as the neglect of the issue in 1949 and the disagreement over it in 1974–77 show. This is confirmed by the fact that references to ‘at any time’ are present in other parts of the GCs and other treaties and have never been interpreted as addressing reprisals.Footnote ¹³¹ The second argument, concerning the prohibition of collective punishment, is also problematic. On the one hand, the ICRC stresses that ‘the concept of collective punishment … should be understood in its widest sense’ and that ‘to include the prohibition on collective punishments amongst the acts unconditionally prohibited by Article 4 is virtually equivalent to prohibiting “reprisals” against protected persons’.Footnote ¹³² On the other hand, collective punishments and reprisals differ from each other and IHL has a tradition of treating them separately.Footnote ¹³³ Moreover, as the drafters of APII could not reach any agreement on the applicability and regulation of reprisals in NIAC, it is doubtful that they could have agreed to address reprisals through a provision on collective punishment.

The third argument, based on the absolute nature of CA3 and Article 4 of APII, is more interesting and has been developed by Cassese,Footnote ¹³⁴ the ICRCFootnote ¹³⁵ and the ICTY.Footnote ¹³⁶ The argument is that these two provisions ‘lay down absolute obligations, namely obligations that are unconditional or … not based on reciprocity’.Footnote ¹³⁷ As such, they are non-synallagmatic and cannot be derogated from between individual parties to a conflict (being erga omnes in nature). Moreover, they are absolute and cannot be unilaterally derogated from at all (being part of jus cogens). However, while these claims may be correct, they miss the point. As circumstances precluding wrongfulness stricto sensu, belligerent reprisals are only prima facie violations of IHL. If the institution is applicable to NIAC, there is no violation of primary norms of IHL, because acts of reprisal fall outside of their scope. This is confirmed by the legal regime of IAC, where acts of reprisals, though prima facie violating IHL rules which are both erga omnes and jus cogens in nature, have never been thought to be outlawed automatically and have had to be the subject of explicit prohibitions. All in all, primary rules of IHL have a limited role to play in the debate on belligerent reprisals in NIAC.

KalshovenFootnote ¹³⁸ and QuénivetFootnote ¹³⁹ suggest that belligerent reprisals in NIAC could be regulated not by the primary norms of IHL applicable in NIAC but the rules regulating reprisals in IAC applied by analogy. It is difficult to assess this suggestion, as no arguments are given in its support. It might be argued, with the ICTY, that ‘what is inhumane, and consequently proscribed, in international wars, cannot but be inhumane and inadmissible in civil strife’.Footnote ¹⁴⁰ Yet, this claim—if taken as a legal and not merely as a moral maxim—is counterfactual. IHL has developed through two interconnected, yet independent bodies of rules, and not everything prohibited in IAC has been prohibited in NIAC. Moreover, the two bodies of law do not always share identical concepts and it is doubtful whether the applicability of reprisals in NIAC—a necessary precondition for any use of analogy with respect to their regulation—could be established by analogy to IAC.

Finally, it is sometimes suggested that Common Article 1 (‘CA1’) of the GCs provides a legal basis for the regulation of reprisals in NIAC. The provision stipulates that ‘the High Contracting Parties undertake to respect and to ensure respect for the present Convention in all circumstances’. The ICTY believes that this confirms ‘the exclusion of the application of the principle of reprisals in the case of … fundamental humanitarian norms’.Footnote ¹⁴¹ This argumentation is flawed. Even on its most ambitious interpretation, CA1 does nothing more than confirm the obligation of States ‘to do everything /they/ can to ensure that the rule … is respected by its organs and … by all’.Footnote ¹⁴² The obligation, which is, moreover, only addressed to States, relates to existing rules of IHL. Were IHL to contain rules on belligerent reprisals in NIAC, States would be obliged by CA1 to respect them and ensure their respect. Yet the provision does not say whether such rules exist in the first place or what they are. Thus, no substantive obligations can be derived from it.

B. Rules of International Law Implicitly Regulating Belligerent Reprisals in NIAC?

In addition to the rules of IHL, several rules of general international law have been invoked as a possible legal basis for the regulation of belligerent reprisals in NIAC. One suggestion is Article 60(5) of the VCLT, which provides that a material breach of a treaty which could lead to a termination or suspension of its operation has no impact on ‘provisions relating to the protection of the human person contained in treaties of a humanitarian character, in particular to provisions prohibiting any form of reprisals against persons protected by such treaties’. Another is Article 50(1)(c) of the ARSIWA, under which ‘countermeasures shall not affect obligations of a humanitarian character prohibiting reprisals’. The two provisions are functionally equivalent, the latter being modelled on the former. In the debate on reprisals in NIAC, they have been referred to by the ICTYFootnote ¹⁴³ and some States.Footnote ¹⁴⁴ Yet those references seem problematic.

It is clear from their wording that the two provisions merely confirm that existing IHL provisions or obligations prohibiting reprisals cannot be set aside in response to any previous violations of international law. In doing so, they do not pronounce on what these provisions or obligations actually are. What they tell us is that if belligerent reprisals were applicable in NIAC, and if some forms of such reprisals were prohibited, then such prohibitions could not be derogated from—by States—in response to any violations of the given IHL treaty (according to VCTL) or any other violations of international law (according to the ARSIWA). What they do not tell us is whether any such prohibitions exist and whether belligerent reprisals are applicable in NIAC in the first place.

Articles 50(1)(b) or (d) of ARSIWA are also sometimes invoked in this context.Footnote ¹⁴⁵ These provisions confirm the illegality of countermeasures affecting ‘obligations for the protection of fundamental human rights’ and ‘other obligations under peremptory norms of general international law’. The ILC gives these expressions a broad interpretation, reading into them a general prohibition of countermeasures which would either derogate from jus cogens or would be inhumane.Footnote ¹⁴⁶ Surprisingly, the examples the ILC quotes include belligerent reprisals. It is, however, unclear why these provisions should be applied in the area of IHL if it is already covered by Article 50(1)(c). Moreover, it is questionable whether the legal nature of primary norms has an effect upon the legal regime of reprisals, taking into account that, as circumstances precluding wrongfulness stricto sensu, they are outside the scope of application of these norms. Finally, even if it is accepted, the argument based on Articles 50(1)(b) and (d) merely tell us that some forms of reprisals would be unlawful in NIAC, if the institution was applicable. It would not tell us whether it is applicable or not.

A. Extralegal Considerations

Extralegal considerations focus on the practical necessity and on moral (un)desirability of belligerent reprisals. These aspects are more thoroughly discussed in the context of IAC, yet they form part of the debate on NIAC as well. This is illustrated by the arguments raised by States (France,Footnote ¹⁴⁷ CameroonFootnote ¹⁴⁸), the ICTYFootnote ¹⁴⁹ and scholars.Footnote ¹⁵⁰ The considerations of practical necessity encompass two questions, namely whether reprisals in NIAC are efficient and whether viable alternative enforcement tools exist to replace them.

The in/efficiency of reprisals is often simply postulated and deduced from general premises rather than evidenced and induced from the actual practice. Kalshoven speaks about ‘general futility’Footnote ¹⁵¹ of reprisals in NIAC without specifying the grounds of his claim. The same is true for Darcy's assertion that ‘reprisals are seen as an ineffective sanction’.Footnote ¹⁵² The relevant practice, though scarce, shows that reprisals have a mixed record. In some cases, they are efficient and succeed in stopping violation of IHL. This can be evidenced by the 1864 Mosby Incident, in which reprisals against detained Union fighters by the Confederation made the Union stop hanging detained Confederation fighters.Footnote ¹⁵³ In other cases, reprisals are inefficient and instead of restoring respect for IHL, they only spin the spiral of violence. The wars in the territory of the former Yugoslavia give examples of such developments. Moreover, the mere availability of reprisals may have a preventative effect, discouraging parties to NIAC from violating IHL out of the fear of repercussions;Footnote ¹⁵⁴ yet this very existence may also make IHL rules look more relative and, thus, less in need of strict observation. The same mixed result is recorded in IAC, which indicates that the efficiency factor cannot in itself decide for or against the applicability of the institution.

The second aspect focuses on the availability of alternative enforcement tools. Some, such as the ICTY, believe that ‘a means of inducing compliance … is at present more widely available and … is beginning to prove fairly efficacious: the prosecution and punishment of war crimes and crimes against humanity by national or international courts’.Footnote ¹⁵⁵ Others claim that despite the developments in international criminal law, reprisals are still the only tool which is available to all parties of an armed conflict and which can be easily employed in the course of such a conflict.Footnote ¹⁵⁶ Both the arguments contain some grains of truth. The range of enforcement tools available in NIAC has become more extensive over the past two decades, due to the establishment of international criminal tribunals and the introduction of several new concepts, such as those of human security and the responsibility to protect. At the same time, the implementation of these tools still has its limits and there can be place left for reprisals, which are cheap, easy to adopt and available in situ.

The moral undesirability of reprisals is acknowledged even by those who are in favour of their applicability. It is linked to the inherent inhumanity of the instrument and its problematic consequences, including the potential for abuse and the risk of escalation of violence. Reprisals are seen as ‘an inherently barbarous means of seeking compliance with international law’.Footnote ¹⁵⁷ Resting on the principle of collective responsibility, under which in the end people bear consequences for unlawful acts committed by others, they are at odds with the fundamental principles of humanity and individual autonomy.Footnote ¹⁵⁸ The situation might be particularly acute in NIAC, in which passions are often unbridled and where reprisals are directed against one's own citizens or compatriots.

The potential for abuse is present in virtually any human institution. Prima facie, this potential is more a factor to be taken into account while defining parameters of legal reprisals than a reason to reject the institution as such. Yet, since the values at stake here are particularly high, the fear that ‘instead of being a means of securing legitimate warfare /reprisals/ may become an effective instrument of its wholesale and cynical violation in matters constituting the very basis of the law of war’Footnote ¹⁵⁹ cannot be discarded easily. The same is true for the risk of escalation, which results mainly from the unilateral assessment of facts and their unilateral legal qualification. Rarely admitting violations of IHL by their side, parties to NIAC tend to impute the original wrong to the enemy, which leads to a spiral of reprisals and counter-reprisals going on without restraints.

With the moral undesirability of reprisals generally acknowledged, the view one takes on their applicability in NIAC depends on whether the practical necessity of the institution is seen as outweighing its inhumanity. Those for whom it does opt for the ‘permissive’ or ‘restrictive’ approach—stressing, in the latter case, the space left to use reprisals. This is the position of Newton, who concludes that reprisals ‘may be the most morally acceptable and humane strategy for serving a strategic imperative of civilized society’.Footnote ¹⁶⁰ Those who come to the contrary position, opt for the ‘extralegal’ or ‘restrictive’ approach—stressing, in the latter case, the limits imposed on the tool. Thus, Darcy holds that the doctrine of reprisals is ‘fast losing all credibility as a valid means of law enforcement’.Footnote ¹⁶¹

Of the two approaches, the second is more convincing. Belligerent reprisals are a tool of self-help based on the logic of collective actions. While this is not unusual in international law, reprisals form part of the legal regulation of jus in bello, which deals with issues of life and death—in the literal sense of these words. The stakes are thus very high here. At the same time, reprisals show a mixed record of efficiency, though it cannot be excluded that they work in some cases. Finally, more humane alternatives have emerged over the past decades, making the use of reprisals less necessary than ever before. In this situation, practical considerations cannot outweigh the moral ones, making the ‘permissive’ approach unacceptable. The choice thus remains between the ‘extralegal’ and ‘restrictive’ approaches.

B. General Theoretical Framework of IHL

The second set of considerations pertains to the general theoretical framework of IHL. This framework, built around certain premises and principles, should help interpret the silence of legal rules. Those in favour of the ‘extralegal’ approach usually assert that the silence entails the absence of the institution from the legal system. Under this view, the use of reprisals is seen not as a right, but as a privilege which is conferred upon parties to the conflict, if (and only if) treaty or customary rules explicitly provide for it. This view was adopted by the ICRC in the 2005 Study, which asserts that the right to resort to belligerent reprisals in NIAC would need to be inferred from (State) practice and opinio juris. The unavailability of such a practice and opinio juris automatically entails that ‘the very concept of lawful reprisal in non-international armed conflict has (n)ever materialized in international law’.Footnote ¹⁶² Since under this approach, there is no institution of reprisals in NIAC, all acts done in these conflicts remain within the realm of primary norms of IHL. Those in favour of ‘permissive’ and ‘restrictive’ approaches presuppose, in contrast, that the silence of legal rules over an institution means the absence of a prohibition of such an institution. This interpretation follows the Lotus principle according to which ‘under international law everything which is not prohibited is permitted’.Footnote ¹⁶³ Belligerent reprisals are not prohibited under international law and hence, they need to be permitted. From that perspective, the resort to reprisals is seen as a natural right.

The two theoretical positions seem prima facie sound and plausible. To decide between them, it is necessary to take into account several factors relating to the legal nature of belligerent reprisals and the character of IHL regulation in NIAC. First, reprisals are not only an enforcement tool but also a circumstance precluding wrongfulness. As such, they remove from the sphere of application of primary norms of IHL an act done in response to a previous violation of IHL. From that perspective, they are an exception to a rule, allowing for the derogation from certain norms. It is generally accepted under international law that exceptions are to be interpreted narrowly. If used in our case that means that the silence on reprisals in NIAC should be read in line with the ‘extralegal’ approach.

Second, IHL regulation of NIAC has traditionally emerged through a lengthy process, with States being reluctant both to bind themselves when suppressing internal insurgency and to give any status under international law to non-state actors fighting against them. It is generally accepted that armed opposition groups enjoy only a ‘functional international legal personality’ limited to what is ‘necessary to exercise the rights and duties laid down by it’.Footnote ¹⁶⁴ Moreover, IHL treaties expressly state that their application does not affect ‘the legal status of the Parties to the conflict’ (CA3) and that ‘sovereignty of a State … by all legitimate means, to maintain or re-establish law and order in the State or to defend the national unity and territorial integrity of the State’ (Article 3(1) of APII). This suggests that the silence of legal rules on an institution should indicate the inapplicability of this institution rather than the freedom to use it without any restraints—in line with the ‘extralegal’ approach.

Third, the institution of belligerent reprisals is based on the idea that any party to an armed conflict, as a whole, is accountable for illegal acts committed by its combatants/fighters or other persons under its control. Thus, the institution presupposes the concept of collective responsibility of parties to the conflict. In IAC, this does not pose a serious problem, since the concept of State responsibility is well developed under international law. In NIAC, the situation is more complicated, since it is not certain whether the concept of collective responsibility of armed opposition groups has already materialized.Footnote ¹⁶⁵ While IHL confers rights and imposes duties upon these groups, violations of primary norms are normally prosecuted at the individual level. The absence of a well-developed concept of collective responsibility of armed opposition groups discourages the adoption of such an interpretation of the silence of legal rules, which would need to be based on this concept, thus favouring the second position in line with the ‘extralegal’ approach.

Fourth, IHL as any other area of international law contains its own interpretative principles indicating how to act in case of uncertainty or doubt. Under IHL, the principle of humanity and the Martens clause are considered to be able to serve in this way. The former would suggest that in case of doubt in which several interpretations are possible, the least inhuman one should prevail. The latter provides that ‘in cases not covered by the law in force, the human person remains under the protection of the principles of humanity and the dictates of the public conscience’ (Preamble of APII). While sometimes given a broader use, the clause is as a minimum thought to imply that ‘in case of doubts, rules of international humanitarian law should be construed in a manner consonant with standards of humanity and the demands of public conscience’.Footnote ¹⁶⁶ If applied to doubts surrounding reprisals in NIAC, these principles would weigh in favour of their non-applicability that is in favour of the ‘extralegal’ approach again.