Treaties in Armed Conflict: The Sharp Split Between “General” and “Particular” Treaties
The recent completion and publication of the second edition of my text, The Law of Treaties, provides an opportunity to reflect on how treaty law applies to specific subject areas and aspects of international law. In this post, I identify a typology of law of armed conflict (LOAC) or international humanitarian law (IHL) treaty configurations and illustrate how the general law of treaties may, in some cases, apply to and influence the operation of IHL treaties, or in other cases, yield to regime-specific rules of formation and interpretation.
Treaty Configurations in International Law
Each subject of international law has its own configuration and cartography or, one might say, its own pitch of treaty typology and lay-out. Take some examples. In international human rights law, there are numerous general agreements (e.g., the 1966 UN Covenants – ICCPR and ICESCR) and many that are regional (e.g., the 1950 European Convention on Human Rights), but hardly any bilateral agreements. There is also a robust layer of municipal law, complementing international human rights law instruments.
By contrast, in the international law relating to the maintenance of peace, one finds the full spectrum of treaty typology. There are general (e.g., UN Charter), regional (e.g., NATO agreement), and bilateral treaties (e.g., on military cooperation). Meanwhile, in international aerial law, there is a tiny general layer (the 1944 Chicago Convention) and a highly developed bilateral layer of law-making (with many agreements setting out how the aerial relations between two States shall be organized: overflight; landing rights; frequency; etc.).
Overall, it can be said that each branch of international law has its own peculiar treaty setting. This fact not only describes the nature of these subject areas. It also affects how these respective treaties work.
Treaty Configurations in the Law of Armed Conflict
International humanitarian law is characterized by an often overlooked and sharp split between, on the one hand, a well-furnished layer of general law (with codifications setting out the widely applicable law such as the four Geneva Conventions of 1949 (GCs)), the complete absence of regional layers, and an extremely robust layer of particular (mostly bilateral) treaties called “special agreements.”
Usually, the latter are either normative or executory in nature. The smaller, normative portion of these particular treaties sets out rules that the parties agree to follow in their bilateral relationships. The parties in such cases normally include some or all of the belligerents engaged in an armed conflict. This is the purpose, for instance, of Article 3(3) agreements under the 1949 Geneva Conventions, where the parties to a non-international armed conflict (NIAC) are invited to conclude agreements to overcome the gaps left in the regulation of this type of armed conflict.
Already before the GCs of 1949 were concluded, such agreements were in use. In the Spanish Civil War of 1936-1939, the International Committee of the Red Cross (ICRC) proposed agreements to extend some prisoner of war (POW) protections in the 1929 Convention relative to the Treatment of Prisoners of War to captives, who were not entitled to POW status (see also art. 83(1)). These agreements were all the more remarkable because at that time there was no general law of NIAC.
However, most of these special agreements are merely executive in nature. That is, their object is to implement IHL duties and not to set out new rules, rights, or obligations. Article 109(2) GC III takes similar form, providing for special arrangements and agreements for the repatriation of seriously wounded and sick POWs during the armed conflict. These agreements allow implementation of the separately articulated and specific duty of repatriation.
There has been significant practice in concluding such special agreements. They are a highly significant body of IHL but are also often overlooked. Some of these agreements have enjoyed limited limelight. For example, in 1992, the parties to the armed conflict in the former Yugoslavia agreed to the application of a series of conventions on IHL, including provisions of the 1977 Additional Protocol I to the Geneva Conventions (see M. Sassòli, A. Bouvier, A. Quintin, How Does Law Protect in War?, p. 1717). Other agreements are less well known. For example, on April 19, 1920, Germany and Russia signed a treaty on the reciprocal repatriation of POWs and civilian detainees (art. 9, available at 2 League of Nations Treaty Series, p. 63). And on June 5, 1974, Israel and Syria signed an agreement on the disengagement and evacuation of their armed forces (art. f, available at 13 Int’l Legal Materials, p. 880). One may also mention the many truces or cease-fires that fall into this category.
Legal Consequences of the General and Particular Layers
This setting of IHL is not merely of a descriptive nature. It affects how treaty law applies to the subject area at issue. Some examples make the point clearer.
First, general IHL conventions are subject to “long” procedures of conclusion, signature, and separate ratification. Meanwhile the great bulk of particular treaties follows the path of “short” procedures of conclusion, with signature, and ratification taking place simultaneously or ratification being announced by an executive branch representative (usually from the military) having authority under international and municipal law to conclude such agreements directly. In other words, most of IHL (even if normally invisible) comes into force without separate procedures of ratification. This allows for responsiveness and agility, highly valued in the necessities of wartime.
Second, general IHL conventions are the archetype of “open conventions” in international law, i.e., conventions to which any third State is invited to participate by becoming a party through accession. Contrariwise, the great bulk of particular IHL treaties consists of “closed conventions” in which no third entity can participate.
Third, the general rules under Articles 31-33 of the 1969 Vienna Convention on the Law of Treaties (VCLT) (and their customary law parallels) apply to the interpretation of general IHL conventions, often with a particular emphasis on the purposive element of humanitarian protection. In the context of the particular conventions, which are mainly of bilateral nature, the intention of the parties, which is marginalized in the VCLT, can play a greater role.
Fourth, particular IHL conventions are especially prone to informal means of adaptation and modification. Bilateral treaties on technical matters are often revised by informal action, for instance, by a proposal that the other party accepts by some letter, act, or other simple acquiescence. For instance, in 2023, Switzerland proposed to Germany that the sessions of a treaty commission should not necessarily take place every year, as provided for in a bilateral railway agreement, but should be called when needed. The point was discussed immediately at a bilateral meeting and an informal agreement to that effect was put on a protocol (I corresponded with the ministry on the issue and have the file). The same flexibility occurs with the bulk of these particular agreements under IHL. Statistically speaking, most of IHL is thus extremely flexible and open to informal changes, even if it is not the most visible and eminent part of that subject area.
Fifth, particular IHL agreements can in most cases be unilaterally terminated by one party on account of material breach, because Article 60(1) of the VCLT or related customary international law applies. This is not possible in the case of multilateral IHL conventions, not only because Article 60(2) does not allow for unilateral suspension, but also because Article 60(5) applies. This rules out even the possibility of suspension in respect of provisions “relating to the protection of the human person contained in treaties of a humanitarian character … .” It is unclear whether Article 60(5) also applies to the layer of bilateral treaties. The answer is probably that only some qualify, but not all, including purely military agreements such as ceasefires.
Sixth, it stands to reason that general IHL conventions of a law-making nature often reflect customary international law. In contrast, bilateral agreements do not themselves give rise to rules of general international law, either because they lack normative status (as purely executive agreements) or because the parties inject special concessions disclaiming the formation of any sort of bilateral customary law.
Special Agreements
Special agreements are subject to particular regulation if they impact jus cogens norms. Common Articles 6/6/6/7 of the 1949 Geneva Conventions provide that the parties to a conflict may conclude special agreements. However, the common articles add that no such agreement shall adversely affect the situation of protected persons nor restrict the rights that it confers upon them. This provision means that more rights, privileges, and facilities may be granted to protected persons, but not less. In a sense, it can be said that “positive derogation” (granting more) is allowed, but “negative derogation” (granting less), or derogation proper, is not allowed.
Various legal issues arise here. The first relates to what is, or is not, a diminishment of the legal guarantees of protected persons. Most often, this is an issue of interpretation. But the position in the case of mixed agreements (i.e., agreements where some advantages are granted but some concomitant disadvantages accompany the former) is unclear. Under the letter of the law, these “disadvantages” will be permitted only when they do not impact the body of “rights” guaranteed under the relevant Geneva Convention. That would imply that a special agreement could stipulate away the rights protected persons hold under other IHL treaties, but not those of the Geneva Conventions and their Protocols.
A second question concerns what happens if the apparent “diminishment” of a right turns out in a specific case to be more advantageous to the protected person than the application of the black letter law. For example, Article 22(1) GC III stipulates that POWs shall be detained only in premises located on land. Egregious abuses during the Napoleonic wars, where persons detained at sea disappeared, provided some of the inspiration for this protective regulation. However, during the 1981-82 Falkland/Malvinas conflict between the United Kingdom and Argentina, there were no proper locations to detain the prisoners on land, where temperatures fall quite quickly into uncomfortable ranges. The United Kingdom thus decided to hold Argentinian POWs on British warships because they offered the best treatment of protected persons in the circumstances. The ICRC agreed but the letter of Article 6 GC III was thus disregarded to the benefit of its purpose.
A third question is more directly linked to the law of treaties. If a “negatively” derogatory agreement is concluded, what will be its fate? Is it void under the special regulation of common Articles 6/6/6/7 and therefore outside the provisions of the VCLT (see art. 73)? Such a conclusion may neutralize the “exhaustiveness” of the VCLT-recognized reasons for invalidity provided by Article 42(2). The VCLT clearly does not provide for an agreement to be void in that case.
The mentioned GC provisions do not delve into such consequences and give no hint that such an agreement would be void. This solution would be odd, because such agreements often contain other provisions that should be maintained. The best legal position seems to be that if a special agreement contravenes peremptory norms of general international law under Article 53 VCLT (and a series of IHL rules of universal nature have such status), the agreement is void and the otherwise valid provisions are not separable (VCLT, art. 44(5)). Contrariwise, if there is merely an impermissible diminution of rights under the GCs’ common Articles 6/6/6/7, the relevant provision of the agreement should be null (or if applied, lead to State responsibility), while the rest of the agreement remains in force and applicable.
Conclusion
As can be seen, the cartography of treaties in a particular subject area can allow a finer analysis of the applicable tools of treaty law. IHL is a quite special area of international law in this regard, one where universal and particular layers of treaties are pitched one in face of the other. Functionally, the particular layer of IHL treaties oils the wheels of the general layer, allowing it to function in the context of the particular needs of changing situations in the field. Such flexibility is particularly necessary in the changing circumstances of war, where unforeseen and quickly shifting events chain up one to the other.
Legally, the particular layer of treaties calls for the application of a series of treaty law rules which may partially differ from those applicable to the general layer. This illustrates the great adaptability of treaty rules, able to espouse the specific needs that arise in varying circumstances. It can thus be said that treaty law certainly corresponds well with the subject area of LOAC, even if our (illusory) hope remains that a better implementation of UN Charter law will dry up the ugly swamp of wars and conflicts.
***
Robert Kolb is a Professor of Public International Law at the University of Geneva and former legal staff member of the ICRC.
Photo credit: Paolo Bovo, U.S. Army
