“Change in International Law: Rules of Change or Changing Rules?” series


PDF Version
Vol 13, Issue 6
Editorial board: Patrycja Grzebyk (editor-in-chief), Lucas Lixinski, Alina Miron, Anne Saab and Peter-Tobias Stoll
Enzo Cannizzaro (Università di Roma “La Sapienza”)


Heraclitus by Johannes Moreelse (cc)



The present ESIL Reflection, the sixth in a series devoted to the dynamics of customary law, aims to explore one of these dynamics, namely the conflicts among customary rules, which not infrequently are a powerful factor in developing the law.

This exploration begins by putting forward logical arguments to demonstrate that the theoretical purity of the dualist doctrine of custom, rather than promoting the development of the law, works as an impediment to this development.

The Reflection goes on to make evident that the dualist doctrine is amenable to a slow process of change and, therefore, appropriate for a conservative community benefitting from internal cohesion among its members, who share common values and principles.

At the other end of the spectrum, sudden transformations in customary law could be explained by a model based on claims made by a State, or a qualified part of the international community, to change the law, which is met with either acceptance or resistance by other international law actors. However, this model, which can explain part of the practice in the aftermath of WWII, is unable to fully explain the recent powerful evolution of custom.

At the heart of this Reflection are numerous claims, including that customary rules are part of the international system of law; that within that system, customary rules interact among each other and with the other rules of law; that these interactions include conflicts and occasional collisions between preexisting custom; and finally, that these conflicts and these occasional collisions not infrequently create new law.

I. The classical doctrine of custom

The essential premise of this Reflection is that the classical dualist doctrine makes it very difficult, if not impossible, to logically explain how custom evolves. This is due to the idiosyncratic features of custom as a source of law. There is no need for lengthy explanations; simply put, the relation between custom and customary law inverts the typical relation between rule and facts. In the contemporary legal world, normally, rules precede, in logic and in time, the facts they intend to govern. Conversely, a customary rule follows, in logic and in time, the facts that gave birth to it, namely custom. Rules govern facts, whereas customary rules are governed by facts. This is tantamount to saying that facts are the creative factor of customary law. Unsurprisingly, customary rules are labeled as factual law.

But it is precisely custom’s factual nature that is a powerful obstacle to its evolution. To determine the content of customary rules over time one must determine custom anew. But how can new custom give birth to a new rule if the pre-existing customary rule controls the social behaviour and, consequently, prevents it from solidifying into a new custom? To develop the law, it is custom that should change the rule but, logically, if the rule is a prescriptive command, it is the rule that reinforces the social custom.

This appears to be quite a puzzling issue that replicates the famous Kelsenian dilemma on opinion iuris; either a subsequent conduct conforms to the existent customary rule, will perpetuate it in time and ensure for it a perennial life, or a subsequent conduct cannot receive the blessing of the opinion iuris and, therefore, will exist in breach of that rule[1].

But how to explain this oddity? How to reconcile the development of the law with the factual nature of custom, which prevents any change in law?

II. Evolution by adaptation?

Herein it is argued that even when considering more flexible perspectives, the relation between custom and customary rules remains problematic.

The more plausible perspective considers that custom is a social product and, therefore, cannot be encapsulated in too compelling a logic. An almost imperceptible difference between patterns of conduct can produce an equally almost imperceptible change to the law. This model is perfectly acceptable and, perhaps, it nicely describes the most intimate functioning of custom, as a source that follows, in all its sinuosity, the slow evolution of social behaviour.

But this model, which captures a part of the reality, is not immune from perplexities. It certainly explains one dimension of custom, typical of an integrated society in which social behaviour evolves at a very slow pace and law changes inadvertently. This image, which is very suggestive indeed, could depict the state of the law in a small bucolic community, whose members share interests and values; but it is the least appropriate imagine to represent the frantic nature of the contemporary international community. In addition, it could be accepted only if one admitted that the operator determining custom is not a neutral observer but is rather part of the process of change. How to explain otherwise that a change of the facts turns out to change the law and is not to be treated as a mere unlawful accident?

These logical difficulties, well known to social scholars, are, in turn, the consequence of the atomistic conception of custom. Whereas rules typically govern a pattern of conducts, custom seems to be a source destined for single-use and thrown away until the next conduct. This may explain why many a scholar relies on imperceptibility as the factor that allows custom to change over time, or, alternatively, on the qualification of a deviating behaviour as the expression of a will to change the law[2].

III. Evolution by claim

Be that as it may, the slow and gradual refinement of social custom can only produce a slow and gradual refinement of its rules. But a custom-change doctrine must be aware that, not infrequently, international law is subject to sudden and abrupt upheaval, which the traditional vision of custom cannot explain. This turmoil in custom is mainly the result of claims made by a State or a group of States that radically contest the existing law, deeming it inconsistent with their interests and advocating for a change. Of course, this claim can be resisted where it does not correspond to a more appropriate balance of interests than that provided for by the pre-existing law. In such a case, it paradoxically reinforces the pre-existing law instead of changing it. This occurred, for example, with the claim to change the law of self-defense in the Bush doctrine on preemptive war[3]. But if a claim corresponds to an acceptable balance of interests, it creates acquiescence and the law changes.

This is indeed a model of customary law change. Remaining in the traditional theoretical framework, it means that previous custom is deprived of the element of opinio, which is automatically transferred on to the new claim; practice immediately follows.

Thus, we have two models of customary law change mirroring two dimensions of custom. The slow evolution of customary rules, in correspondence with the equally slow evolution of social custom, is typical of a cooperative, conservative and integrated community. It is noteworthy that these attributes do not necessarily denote a community that shares collective interests and values. The typical example of the slow evolution of custom is the traditional customary regime on immunities, namely a common interest, not a collective one.

Conversely, the model of claim/resistance or claim/acquiescence reveals the trait of a competitive community, whose members vie for their own interest to the detriment of the interests of others. The evolution of the law is prompted by a claim to change, which is necessarily fast in nature, sometimes very fast, and this process is not immune from a dose of voluntarism; in a sense, it is negotiated custom. The chaotic development that has taken place in the realm of the law of the sea offers countless of examples. An example of a successful claim is the notorious Truman proclamation[4], which radically and almost instantaneously changed forever the classical tenets of this law, based on the revered principle of the freedom of the high seas. As previously referenced, an example of a failed attempt to change the law was probably the Bush doctrine on the preemptive war. The US claim aimed to change the law, but the only thing that changed was the geopolitical balance in one of the most troubled regions of the world, and the perception by a considerable part of the international community that the US stacks the deck.

IV. Is customary law a system of law?

These two models fit well, albeit for opposite reasons, in the doctrine of customary law change. Yet, it would be an error to rely entirely on these models to explain how custom evolves. Whereas the first model appears to be too slow to keep up with the rhythms of our times, the second is, conversely, too fast and pervaded by the idea that the will of States is the unmoved mover of international law.

Thus, we are back at square one. What is lacking in the chaotic and acephalus contemporary international community is a model capable of capturing the expectations of that community and transforming them into new law.

The main hurdle preventing the achievement of this objective is its alleged atomistic conception. This conception predicates that not only the genesis, but also the change of a customary rule can occur only on the basis of a consistent pattern of conduct, practice, and an equally consistent sense of legal obligation. This belief has ancient and deep roots. But the inevitable consequence of this conception is that customary law is not a full-fledged system of law but only a set of isolated rules, each governing the specific conduct from which it germinated. Indeed, the very essence of a system of law is the existence of systemic interactions among its rules, be they systemic interpretation, definition of their mutual scope, conflicts, occasional collisions, and so on.

All this is prevented by the idea of the two-way relation between custom and customary rules, namely the producer and its product. In particular, the factual nature of custom entails that its output, namely customary law, is gripped in a deadly embrace with the facts from which it sprouts. But two arguments help to dismiss this idea.

First, a theoretical argument: the widespread idea that custom’s factual nature is an idiosyncratic feature is misleading. In the normative theory of law, all sources of law are mere facts. A vote in Parliament approving a statute is a fact entitled to produce law by virtue of a Constitutional determination. In other words, the very existence of a legal order depends on an autopoietic factual determination – this is one of the functions of a constitution – which identifies the facts that create law.

The second is a factual argument, which is capable of dismissing the dogma that every customary rule necessarily evolves on the basis of the same factual procedure from which it originally sprouted. International practice demonstrates that dynamics among customary rules are part of the landscape of contemporary international law, and that international law is a system of law whose rules interfere and collide with each other, creating an intricate network of interactions that continuously transform its rules to adapt to the emerging need of the community they aspire to govern. In a nutshell, the factual nature of custom does not prevent customary rules from breaking free, from establishing relations among themselves and among other rules of the system of international law and from creating new rules and principles that are an essential complement to the system of international law.

V. Direct and indirect conflicts

The normative environment briefly outlined in the previous section will be further explored in the following sections with the aid of some notorious examples.

The existence of conflicts among customary rules can be demonstrated also by a normative argument related to the notion of jus cogens, which notoriously is part of general international law. The function of this relatively new type of rules is precisely the settlement of conflicts among international legal rules. Although Art. 53 and 64 of the Vienna Convention on the Law of Treaties of 1969 only refer to conflicts between jus cogens and treaties, it is commonly believed that also “ordinary” customary rules are invalid if they conflict with jus cogens. This is the effect of a conception of jus cogens as higher law, and not only as a limit to the contractual capacity of States[5].

The ILC expressly admitted that the scope of jus cogens also encompasses customary law, but in describing the effect of the conflict used a different phraseology than that used for conflicts between jus cogens and treaties. In Conclusion n. 14 it pointed out that a customary rule conflicting with jus cogens “does not come into existence”[6]. For the purposes of the current writing, this phraseology can convey the impression that conflicts between jus cogens and “ordinary” customary rules can arise only in the form of an Aristotelic conflicts: A/non-A. Therefore, custom that is inconsistent with jus cogens is unable to produce a customary rule. However, this assumption is illogical. There would be no need to refer to jus cogens to conclude that the conduct of States in a direct conflict with another pre-existing “ordinary” customary rule is unlawful and, logically, should prevent the formation of a new customary rule. Precisely in accordance with the atomistic conception, the very existence of a rule of customary law should prevent that a new rule in direct conflict with the pre-existing one comes into existence.

Luckily enough, examples of a direct and explicit conflict between customary rules are rare and mostly in the form of the two models referred to in the preceding paragraphs. Indeed, the examples offered by the ILC are not direct conflicts but occasional collisions between existing rules with different scopes that only occasionally clash.[7] In addition, the idea that a customary rule does not arise if it conflicts with a rule of jus cogens is at odds with the assumption, which emerges from Conclusion n. 20, that every international law rule must be interpreted consistently with jus cogens.

VI. Πόλεμος πάντων μὲν πατήρ ἐστι (Conflict is the father of all)[8]

In States’ practice and in international case law, conflicts among customary rules are rarely presented as such. In the Jurisdictional Immunities case, the ICJ ruled out the existence of a conflict between jus cogens and the customary regime of sovereign immunities without expressing its outrage for the temerarious allusion of a party to the possible existence of such a situation.[9] But it would be improper to infer too much from this passage. The Court hastily held that there was not a direct conflict, which is undeniable, but did not even contemplate the possible existence of an indirect conflict. Yet, whilst it is difficult to detect examples of direct conflicts in practice and in case law, it is not uncommon to detect implicit examples of indirect conflicts. Two examples will clarify this quite audacious assertion.

The first, coming from State practice, is the famous Torrey Canyon incident, which occurred in 1969. In that case, a forcible action by the UK, which bombed a foreign ship on the high seas off the coast of Cornwall to prevent an impending environmental disaster, was not contested by the international community but, surprisingly, encountered almost unanimous approval. This conduct immediately became the new behavioral standard under customary law and prompted a number of successful treaties modeled after the factual circumstances of the case.

How to explain this apparent oddity? One plausible explanation is based on the novelty of the factual situation, which required the law to be updated. To be sure, in this case there is no room for holding that the UK’s actions filled a gap in the law. Quite the contrary, this situation was disciplined by a multiplicity of rules, which hitherto had peacefully co-existed within their respective scopes, namely, on the one hand, the freedom of the high seas and the principle of the exclusive authority of the flag’s State over its ships, and, on the other, the prohibition to pollute in the vicinity of the coast of another State. But their application to a new case created a collision and required the re-balancing of the interests and values underlying the pre-existing rules. The British action implicitly deconstructed the previous balance between existing values and interests and recomposed them in a new normative balance that was more appropriate to the new situation. In other words, the new rule was implicit in the system of the pre-existing customary rules. The unusual events brought this implicit rule to the surface.

The technique of deconstructing the interests underpinning the clashing rules, and of recomposing them in a new rule apt to govern a new situation was used with relative frequency by the ICJ. For example, the ICJ’s solution in the 1951 advisory opinion on the reservations to the convention of genocide was arguably based on this technique[10], and the logical process followed by the Court speaks volumes on how the technique was employed. First, the Court carved out from practice the principle of the integrity of treaties, which underlays the rule that “that no reservation was valid unless it was accepted by all the contracting parties”. Second, it extracted the principle of universality of the new multilateral treaties, which “has already given rise to greater flexibility in the international practice concerning multilateral conventions”. The need to appropriately combine these two principles led the Court to draft a sort of judicial legislation on reservation, which immediately became the new law.

In the Fisheries Jurisdiction cases of 1974, the Court was asked to determine which law could reconcile the needs of the coastal State to reserve an equitable share of resources for its population, which had already materialized in the practice of the exclusive fishing zones, with the traditional rule of the freedom of the high seas. In this difficult situation, the Court deemed it opportune to premise a declaration of the appropriate method to be employed to settle the dispute: “the rules of international maritime law have been the product of mutual accommodation, reasonableness and co-operation. So it was in the past, and so it necessarily is today”[11]. Notoriously, the determination of the balance of interests underlying the respective rules invoked by the parties put in motion a process that ultimately appeared in the new rules of the exclusive economic zone.

VII. An inconclusive conclusion

There was a time when international lawyers predicted that customary law was destined to disappear or, at least, to be considerably downsized. Subsequent events disproved this prophecy; customary law is alive and well and seems capable of proliferation through parthenogenesis. Yet a full-fledged theory of the conflicts between customary rules has yet to be developed. Such a theory could definitively prove that conflicts prompt the development of customary law, a development that is based on the previous practice interpreted in light of the emerging needs of the international community.

I believe that this endeavor is well worth attempting. The international community needs general law that can keep up with the evolution of social custom. I further suspect that practice and case law have already accepted that customary law can be created by interaction between existing laws. The materials are there for whoever intends to collect and classify them.



[1] See H. Kelsen, Théorie du droit international coutumier, in Revue internationale de la théorie du droit (1939) 253, at 263.

[2] See, for recent scholarship, B.D. Lepard, Customary International Law. A New Theory with Practical Applications (CUP 2012), 277 ff.; J. Kammerhofer, Uncertainty in the Formal Sources of International Law: Customary International Law and Some of Its Problems, in 2004 EJIL 523 ff.; See also, in a wider context, P.-H. Verdier and E. Voeten,  Precedent, Compliance, and Change in Customary International Law: An Explanatory Theory, in 2014 AJIL p. 389 ff).

[3] See the US National Security Strategy (NSS), issued September 20, 2002

[4] President Harry S. Truman’s Executive Order 9633 of September 28, 1945.

[5] I refer, inter alia, to my contribution A Higher Law for Treaties?, in The Law of Treaties Beyond the Vienna Convention (E Cannizzaro ed., OUP 2011) 425 ff. The hierarchical superiority of jus cogens over all the “ordinary” international law rules has been acknowledged by the ILC. In the Commentary to the Conclusion n. 14, the ILC made it clear that “(p)eremptory norms of general international law (jus cogens) are hierarchically superior to other norms of international law and therefore override such norms in the case of conflict”; see the Report of the International Law Commission, Seventy-third session, doc. A/77/10, at 55.

[6] See doc. A/77/10 at 10. The ILC did not explain this phraseology in many words. In the Commentary to Conclusion n. 14 it wrote that “Unlike in the case of treaties, the terms ‘invalid’ or ‘void’ are not appropriate since the putative rule of customary international law does not come into existence in the first place”. Yet, in its Second report, the Special rapporteur, after held that the effect of jus cogens on customary law do not significantly differ from the effect on treaties (see paras. 138-140, and after held that “(t)he notion that customary international law rules that conflict with norms of jus cogens are invalid flows from the hierarchical superiority and is reflected in the practice of States”, drafted the Conclusion n. 15 in terms analogous to those used by the final Conclusion n. 1: “A customary international law rule does not arise if it conflicts with a peremptory norm of general international law (jus cogens)”.

[7] See the document A/77/10, p. 56-57.

[8] Heraclitus, fragment 53. Polemos is mainly translated into “war”. This word is here used in the sense given by Heidegger, as a situation conflictual and dynamic, opposed to the static and perpetual ideas of Plato: see the Cambridge Heidegger Lexicon (M. Wrathall ed., CUP, 2021).

[9] Jurisdictional Immunities of the State (Germany v. Italy: Greece intervening) [2012] ICJ Rep 99, para 93.

[10] Reservations to the Convention on the Prevention and Punishment of the Crime of Genocide (Advisory Opinion) [1951] ICJ Rep 15.

[11] Fisheries Jurisdiction (United Kingdom of Great Britain and Northern Ireland v. Iceland) Merits [1974] ICJ Rep 3, at para. 53 and Fisheries Jurisdiction (Federal Republic of Germany v. Iceland) Merits [1974] ICJ Rep 175, at para. 45.



Cite as: Enzo Cannizzaro, ‘How Custom Evolves’, ESIL Reflections 13:6 (2024).

© 2024. This work is licensed under a CC BY-NC-ND 4.0 licence.