Is the Time for Law of War Treaty Commentaries Over?

by | Feb 26, 2021


In two recent Articles of War posts, Sean Watts and Michael Meier criticized the Updated Commentary of Geneva Convention III. Rather than focusing on specific interpretations adopted or suggesting other interpretations, Michael Meier criticized the Commentary’s methodology as “flawed.” As a member of the editorial committee of this Commentary and co-editor of a previous commentary by the Geneva Academy of International Humanitarian Law and Human Rights—to which Sean Watts also contributed, and the methodology of which is certainly neither more consistent nor less open to the criticism made by Michael Meier—I was intrigued. The cumulative effect of their criticism would make any commentary of a law of war treaty simply impossible.

I cannot write in the name of the editorial committee of the Updated Commentary. Nor can I speak for the ICRC itself (where, in full disclosure, I spent over a decade working until 24 years ago). Rather, I am writing here as an international law professor who agrees that Geneva Convention III (GC III) must be interpreted in the same way as any other international treaty—according to the rules of the Vienna Convention on the Law of Treaties. I also admit that questions such as whether there are any specifics in the methodology of interpreting law of war treaties and the legal value of commentaries published by the ICRC deserve further academic research.[1]

Criticism of the ICRC Commentary

Michael Meier’s criticism of the Commentary focuses on four main points, the second and fourth of which are also expressed or implied by Sean Watts:

  • The ICRC substitutes its interpretation of the object and purpose of the Convention when the original drafters specifically declined to specify one (because they did not adopt a preamble);
  • Too much weight is afforded to statements and writings of the ICRC, academics, and non-governmental organizations while insufficient weight is given to State practice in applying the Convention, which I will refer to as “actual operational practice”;
  • Reliance is placed on subsequent treaties to which certain States are not a party; and
  • An expansive interpretation of the 1949 Geneva Conventions is adopted where it was not intended by the States. Here the interpretation of the positive external aspect of the obligation to “ensure respect” foreseen in Common Article 1 adopted by the Updated Commentary (paras 186-216) is mentioned by many as an example.

If this criticism were to be taken onboard wholesale, the cumulative effect would make an updated commentary—and any commentary of a law of war treaty—simply impossible (in particular if such a commentary were intended to be useful to the entire world and not only the United States).

I would, however, suggest that such criticism is not justified for a number of reasons.

Problems with the Criticisms

1. States did not adopt a substantive preamble. The Updated Commentary explains why there was no consensus on different proposals (paras 134-144). However, this lack of consensus cannot mean that GC III has no object and purpose. Interpretation according to the Vienna Convention on the Law of Treaties would otherwise be impossible.

The Updated Commentary suggests the object and purpose is to “ensure that prisoners of war are humanely treated at all times, while allowing belligerents to intern captured enemy combatants to prevent them from returning to the battlefield” (para 89). I have not found anyone suggesting a different object and purpose. If the object and purpose was simply to allow States to hinder individuals from continuing to fight, then there would have been no need for them to adopt a treaty. States could have detained without a treaty, at least in 1949 before international human rights law required a legal basis for any deprivation of liberty.

2. Actual operational State practice on GC III subsequent to its adoption (how States actually have treated POWs) could obviously assist in the interpretation of GC III provisions. Such practice, however, is relatively scarce. This is because (recognized) international armed conflicts have fortunately been relatively rare and, even when they have occurred, States often have not recognized the POW status of their captives. When these States have had POWs, their practice was sometimes in violation of the clear wording of GC III, but because that practice was clandestine or because other States spoke out against such practice, it did not abrogate or modify those rules.

In other cases, actual practice went, from a humanitarian point of view, beyond what the mere treaty text seems to require. Michael Meier, however, would argue that these additional restrictions could have been done for operational and policy reasons rather than out of a sense of legal obligation. Only rarely will we be able to assert that there was opinio juris. The greatest number of POWs recognized as such since the 1960 Pictet Commentary was published were held by India, Pakistan, Iran, Iraq, Ethiopia, Eritrea, and the United States. Should the new commentary have tried to find the lowest common denominator between the practice of those seven States?

3. The ICRC has a right to visit POWs and to interview them without witnesses under Article 126 of GC III. As such, the ICRC could give a good overview of the actual practice of States. But it too would meet the difficulty mentioned in the previous point.

The Updated Commentary takes practice found by the ICRC in its visits into account, without explicitly naming the countries concerned. The lack of specific mentions is unsatisfactory from an academic point of view, but it is understandable from an ICRC perspective. The ICRC wants to preserve its access and the integrity of its bilateral and confidential dialogue with detaining powers. To do so, the findings it makes during its visits must remain confidential. Nevertheless, the ICRC’s opinions on the interpretation of GC III can claim to be more credibly based on the State practice of those that have actually held POWs than anyone else’s opinions. Though, it is, of course, regrettable that the reader cannot check this.

The ICRC is not a State and its practice therefore does not count as subsequent practice according to the International Law Commission. But, shouldn’t one nevertheless consider its opinions when writing a commentary, inter alia because they may be “relevant when assessing the subsequent practice of parties to a treaty”?[2]

4. Astonishingly, critics argue that even (some) military manuals, produced by States, “are not necessarily an accurate reflection of a State’s view of its legal obligations under a particular treaty” as they can be formulated for policy reasons. This could, however, be even truer of most statements made by States in international fora. Michael Schmitt and Sean Watts have rightly criticized States for not making their position on international humanitarian law issues known often enough in other ways. Until these views are heard, should we therefore abandon our search for States’ opinions (except in twitter messages by Presidents), although such opinions are decisive, according to critics, and thus not interpret treaties at all?

5. It has been said that the Updated Commentary refers too frequently to academic writings. Even if the authors are (former) State officials, there is always a disclaimer. According to critics, the Commentary should not have given much weight to such writings. States themselves, however, never write academic articles. Should we therefore neglect academic articles, although States have recognized in Article 38(1)(d) of the Statute of the International Court of Justice “the teachings of the most highly qualified publicists of the various nations, as subsidiary means for the determination of rules of law”? If we do not take academic articles into account, where does it leave us? Given the difficulties we face in knowing actual operational practice and if one accepts the above-mentioned doubts around the legal value of military manuals, what remains upon which we can write a useful commentary?

6. Some say that we should not rely on subsequent treaties to which certain States are not a party. As the Geneva Conventions are the most universally accepted treaties—with 196 States Parties—whenever one makes reference to another treaty in commenting on them, there will inevitably be States not bound by that treaty. When we talk, for example, about gender equality issues should we therefore not refer to Convention on the Elimination of All Forms of Discrimination against Women, as that treaty “only” has 189 States Parties? On child protection issues, should we not take the UN Convention on the Rights of the Child into account just because “only” 193 States are parties? The United States is not a party to those treaties, but many other States are, and even the United States does not reject the substance of all their provisions. Is it not sufficient that the Updated Commentary points out: “It is important to note that treaties other than the Conventions themselves are referred to in the Commentaries on the understanding that they apply only […] to States that have ratified or acceded to them” (para 95)?

7. One might hope that if we may not refer to other treaties, we may refer to customary law. However, according to critics, the ICRC’s methodology to identify customary rules is as flawed as its interpretation of GC III. Actual operational practice accompanied by clear evidence of opinio juris—as LTC Ronald Alcala argues is required—is nearly nonexistent for nearly all States both for the purpose of finding customary rules and for the purpose of identifying subsequent practice aiding in the interpretation of treaty rules (let us assume that the same methodology must be used for both purposes). A simple reason is that most States have not held POWs since GC III was adopted.

8. On Common Article 1, it is indeed correct that the drafters of the Geneva Conventions did not consider the existence of a positive obligation to ensure respect by other States and armed groups in 1949. And, even today, a few States and academics have criticized such an understanding. However, should the Updated Commentary have dismissed the interpretation expressed by the UN Security Council, the UN General Assembly, and the International Court of Justice (referenced in the Commentary, para 189) that such a positive obligation to ensure respect by others exists as lex ferenda? And, should it have considered the current U.S. position as lex lata? If this is the correct methodology in public international law then we should also apply such an understanding to State pronouncements or State silence—which allegedly also indicates opinio juris—by Russia, China, Iran, Syria, Pakistan, or Sri Lanka. But then the Commentary would at best simply reproduce the treaty text, and at worse would deconstruct every rule.


As most of the above-mentioned problems are not peculiar to GC III, is the conclusion that there should be no more commentaries to law of war treaties, as they should only be interpreted by States, concluding a new treaty? Does anyone suggest that domestic law may be interpreted only by parliament?


Marco Sassòli is Professor of International Law at the University of Geneva, Switzerland.



[1] See the forthcoming PhD thesis by Sophie Rondeau, Les caractéristiques spéciales du droit international humanitaire en matière de sources, Geneva/Laval, 2021.

[2] ILC, Draft conclusion 5(2) on subsequent agreements and subsequent practice in relation to the interpretation of treaties.