‘Customary Law Interpretation as a Tool’ Series

PDF Version
Vol 11, Issue 5
Editorial board: Federico Casolari, Patrycja Grzebyk (editor-in-chief), Ellen Hey, Guy Sinclair, and Ramses Wessel
Başak Etkin* (Université Paris II – Panthéon-Assas)



Heraclitus – painting by Raffaello Sanzio da Urbino (cc)


Herakleitos never actually said “You cannot step into the same river twice”. In Plato’s Cratylus (402a), Socrates quotes Herakleitos to Hermogenes as having said so. The famous line is therefore merely heard through the grapevine.


  1. The Analogy


Analogies (and metaphors) are very common in IL[1], and if used well they can be part of the rational reconstruction of a notion. The river analogy is particularly useful for our understanding of norms and interpretation, specifically that of customary (international) law. It exposes the flow of custom in a surprisingly simple way, which then highlights key differences with treaty law as well as some of the many inconsistencies of the two-element approach. Thinking about customary rules as a flux reveals some of these overlooked aspects.

What Herakleitos actually said about the river is somewhat disputed: three of his fragments seem to mention the idea of the river, but fragment 12 is widely accepted as the genuine river fragment.[2] Alas, because of the confusion around the competing fragments, his doctrine is widely misunderstood. The river doctrine is more about the unity in change rather than the constancy of change, as presented in fragment 12.[3] David W. Graham explains that “some things stay the same only by changing. […] On this reading, [Herakleitos] believes in flux, but not as destructive of constancy; rather it is, paradoxically, a necessary condition of constancy, at least in some cases (and arguably in all)”.[4] What makes the river a river is the constant changing of waters.[5]

So is the case for customary international law¾and potentially even for all legal rules more generally. One cannot adjudicate the same custom twice. As a unity, the customary rule R which we observe at time t1 is the same customary rule as the one which we observe at t2; as a diversity, since its contents keep changing, it is not the same.[6]

For instance, at times t1 and t2, a judge may be adjudicating the same rule R, as in, referring to the same source without identifying a new customary law. Yet, between those two moments in time, the very same rule R may have been subject to interpretation by other authorities like states or international organisations, its penumbra reduced and core expanded. Thus, the judge would be referring to the same rule R, although its contents would have changed in the meantime. The rule is therefore the same and not only in name, given that it is seen as in continuity, yet different all at once. As this sort of change is inherent to the life of a rule, as is the flow to a river, it is this constant change or flux that makes it a unity.[7] For instance, in the Jurisdictional Immunities of the State case, the International Court of Justice decided that an accusation of jus cogens violations would not deprive a state of immunity.[8] From that point on the customary rule on immunities had changed, and future cases, such as the proceedings instituted by Germany against Italy on 29 April 2022, now have to take this into consideration. The rule is the same, that of state immunity, yet different, with the addition of a new aspect.

So, what does this mean for interpretation of customary rules?

Firstly, it means that interpretation of a (customary) rule is merely a part of its flux and that all interpretation is somewhat constructive. The lines between modification and evolution through interpretation are blurred, both indicating a type of change¾with the latter implying somewhat a sense of direction, given the ordinary meaning of the term ‘evolution’. Interpreting a (customary) rule is necessarily changing the rule, by including or excluding a certain aspect of it. As may already be clear, change here is understood in a specific but very broad way. Some argue that interpretation does not change the rule but merely reveals it. I, on the other hand, would suggest using the word ‘change’ for the outcome of all interpretative processes. In my view, the true content of the rule is indeterminate before it is interpreted, whereas the idea of a reveal suggests that it was fixed beforehand in an improperly deterministic fashion.  While interpretation is done in an active sense by an actor invested with the authority to interpret the rule, be it for instance a court or a state, change (as an umbrella term for modification, evolution and beyond as used in this paper) is the result of that process, and all of these phenomena contribute to the rule’s flux.

A seemingly artificial division of what is understood by interpretation of customary international law will allow for some clarification: is there interpretation of the customary rule independent from the interpretation of the two elements that prove its existence? There must be, because state practice and opinio juris are not the rule itself, they are merely evidence of the rule, or the microscope with which we observe it.[9] Even if the ILC calls them ‘constitutive’, the rule itself, already constituted, exists in a normative realm before its initial observation, and may or may not be recognised by the court. [10] The rule, or its new interpretation, exists and does not exist until the judge rules on the matter, as there is just no certainty on whether the rule will be identified or not (and on which version of it will be chosen among the many possibilities). The problem here is that as there is no formal ascertainment of custom,[11] as a result the difference between the existence of a customary rule and its content is more difficult to establish. If the object were a treaty provision, the authorities would first look to see whether there was a treaty in force and then to the content of the rule. This two-step approach could be replicated for a customary rule as follows: firstly, if the customary rule in question has never been recognised before a court or another body in the past, the authorities would first have to establish its existence in a quantifying manner (“is there enough state practice?” etc.). Then and only then could questions around content be answered; yet establishing the existence through the two elements would already have prejudiced the content¾the practice and belief are necessarily tied to a certain understanding of the rule. Taking into account or leaving out a particular behaviour or a specific belief about the rule will necessarily taint the final product, as the evidence presented to a jury would affect the verdict. Therefore, unlike treaty law and as dictated by the flux of the river analogy, the existence/content division somewhat collapses, although this questionable distinction may be useful for theorising other issues. Secondly, once it is identified, there can also be interpretation of the rule itself independently from the interpretation of the two elements, which will also affect the content, in light of its object and purpose for example. Interpreting the rule changes it in an unknowable, even if not unforeseeable, manner. Hence, in all cases, interpretation is constitutive of content, and essentially a similar exercise to identification. Especially given the lack of official terms as in written law, there is no ordinary meaning to give to the terms, and therefore all interpretative exercise of a customary rule will be constructive and constitutive of meaning, rather than extractive of what that is already there, as might be the case with treaty interpretation. Interpretation thus plays a decisive role in every step of a customary international rule’s life[12], including¾and maybe especially¾identification, as this is the closest moment to an origin. Interpretation and identification are therefore¾while distinct¾intertwined in a single stream, which makes the flux a particularly suitable analogy.

Finally, as interpretation is intertwined with change, so might it be with the identification of a new, separate, customary rule. The question is, at what point during the interpretative process does the interpreter bifurcate the flux and generate a new rule? Treating customary international law the same as treaty law would be a legal fiction, but interpretation of customary international law can, while not replicating it mutatis mutandis, resemble treaty interpretation. In treaty law, the fact that the text stays the same makes for an easy answer; one cannot go beyond the text¾whatever that means. When it comes to customary international law, on the other hand, the answer is a sorites paradox. A customary rule can have a core and a penumbra just like treaty law, but there is no text to limit the change that the rule goes through. Through the incremental change the norm goes through à la sorites, the river analogy would mean that one single customary norm can become unrecognisable and entirely different and still somehow be the same. Potentially, a law of the sea rule can turn into a rule on jurisdictional immunities over time, as interpretation intertwined with new practice and opinio juris together in a double helix carry the norm to new dimensions.[13] Unbound by a treaty’s black letter confines, the customary rule is free to change, in the above-mentioned ‘large’ understanding of change. However, change can be less incremental and more abrupt, in which case, unlike the incremental change à la sorites, the phenomenon cannot be explained simply by the interpretative flux and a proportionally abrupt change in practice or opinio juris must be established.


  1. The Limitations


Strangely, the river analogy attains its limit when confronted with the question of time, the very thing it is supposed to describe. Unlike a river which starts at a source and ends where it joins another body of water, time has no beginning and no end. Thus is the case for custom, which partially exists beyond the confines of time. How so?

As mentioned above, contrary to treaty law, there is no formal ascertainment of customary rules, and it is therefore difficult to pinpoint an exact start date. In the Chagos Archipelago advisory opinion before the International Court of Justice[14], the court had to determine if the right to self-determination had crystallised in 1965, and decided that it had, without giving the exact date when it did. To do so, the court went back in time to situate the rule by looking at its formal evidence through the two-element approach. However, the two elements are proof of the norm but not the norm itself, they are merely how customary rules present themselves to us, just as treaty law presents itself through the medium of a treaty in black and white. The norm these elements translate into words, as in treaty law, are intangible and unobservable. What we can and do observe is the formal evidence of customary rules, which are the two elements. At the moment of the first formal evidence, practice P by state S, the state claims that they are obeying a rule that already exists. Therefore, there is a beginning to the formal evidence, but not to the norm itself.[15] Interpretation happens in time and constitutes the flux, but the river analogy cannot account for the timeless nature of customary rules qua rules that exist in an atemporal realm, as the paradox of timelessness clashes with the temporal interpretative rivers.

One explanation, once again resorting to ancient Greek philosophy, is to imagine customary rules as Plato’s abstract objects, known as Forms. Like the latter, customary norms exist outside space and time, and are entirely non-physical. In order to reveal them, we, international lawyers, invented the two-element doctrine. Through practice and opinio juris we see proof of the rule, not the rule itself. This fiction allows us to situate the rule in time, as we live in a temporal and spatial world, and thus was the case in the Chagos Archipelago opinion. The rule does not have a formal beginning[16], but its evidence does. The soft discursive structure of opinio juris together with the sharp staticity of state practice bring the abstract rule into the spatio-temporal world.

Maybe not a limit, but a nuance that the river analogy misses is the direction of fit of customary rules. Direction of fit, as explained by speech act theorist John Searle, allows us to contrast commands and assertions through the distinction between fitting a situation to words and fitting words to a situation. When applying treaty law, it is the situation, the ‘world’, that has to fit the word (world-to-word) primarily, according to what has been decided and put in the precise terms of a provision¾at least in a simpler case where the rule is merely applied without resorting to interpretation; in a more complex case, ‘world-to-word’ application may not be black and white since the contents of a treaty obligation too can change during the interpretative process. In customary law, at the moment of identification, the direction of fit is opposite: words need to fit the world (word-to-world). The authority which decides that a customary rule already exists in the world finds the words that will fit that rule. In a sense, this is similar to how the two elements translate an intangible norm into concrete evidence. We must choose the right words to translate the observable situation in the world into language that will describe the normative rule. However, there is a limit to this explanation too. Once the customary rule is identified and transcribed into words, then the direction of fit can no longer simply be word-to-world, but also world-to-word as after the moment of identification we are provided with the words that the world is supposed to fit. For the interpretative process, one could say the direction of fit is doubled, as the interpreting authority, while trying to fit world to word, also fits word to world. In treaty law and also in customary law once it is identified, adjudicating authorities look at the world to judge whether it fits the words. Nevertheless, especially in evolutive interpretation, they also look at the words to make them fit the world in its current state, and their declaration of what the rule now is makes the rule precisely that¾which makes it a double direction of fit.


  1. Conclusion


While not perfect, the river analogy is useful, not only because it puts into words the flux that is customary law, but especially because it helps show the cracks in the two-element theory. As the two elements are merely proof of the customary rule, and in the absence of formal ascertainment, an additional effort is necessary to reveal the differences between identification, existence, modification and evolution, which are more clear-cut in treaty law. What is certain is that all interpretation generates change and contributes to the flux of the customary rule. Breaking from the confines of treaty law logic by studying the particularities of customary law, given that the latter does not fit the restraints of the former, would allow us a better understanding of the nebula that is customary international law.



[1] Gorobets, K., 2019. The Unity of International Law: An Exercise in Metaphorical Thinking. Philosophy of Law and General Theory of Law, 2019(2), pp. 80-109; Cohen, H., ‘Metaphors of International Law’, in Bianchi A. and Hirsch M. (eds) International Law’s Invisible Frames: Social Cognition and Knowledge Production in International Legal Processes (OUP, 2021).

[2] The three fragments mentioned are numbers 12, 49(a) and 91(a). T. M. Robinson is of this view, and according to D. W. Graham, K. Reinhardt, G. S. Kirk, M. Marcovich all agree. From Robinson, T. M., Heraclitus: Fragments: A Text and Translation with a Commentary [1987] University of Toronto Press:

Fragment 12: ποταμοῖσι τοῖσιν αὐτοῖσιν ἐμβαίνουσιν ἕτερα καὶ ἕτερα ὕδατα ἐπιρρεῖ (“As they step into the same rivers, different and (still) different waters flow upon them.”)

Fragment 49a: “ποταμοῖς τοῖς αὐτοῖς ἐμβαίνομέν τε καὶ οὐκ ἐμβαίνομεν, εἶμέν τε καὶ οὐκ εἶμέν” (“We step and do not step into the same rivers; we are and are not.”)

Fragment 91a: ποταμῷ γὰρ οὐκ ἔστιν ἐμβῆναι δὶς τῷ αὐτῷ καθ ́ Ἡράκλειτον (“For, according to Heraclitus, it is not possible to step twice into the same river.”)

[3] In the 2017 film directed by Luca Guadagnino, Call Me By Your Name, based on the novel of the same name by André Aciman, Oliver (the protagonist’s love interest who is also a graduate student) writes on this topic: “The meaning of the river flowing is not that all things are changing so that we cannot encounter them twice, but that some things stay the same only by changing”.

[4] Graham, D.W., “Heraclitus”, The Stanford Encyclopedia of Philosophy (summer 2021 edition), E. N. Zalta (ed.).

[5] Hume expresses the same idea: “Thus as the nature of a river consists in the motion and change of parts; tho’ in less than four and twenty hours these be totally alter’d; this hinders not the river from continuing the same during several ages.” in D. Hume, A Treatise of Human Nature, reprinted from the Original Edition in three volumes and edited, with an analytical index, by L.A. Selby-Bigge [1896] Clarendon Press,, p. 258. For a French expression of the same idea: “Plus ça change, plus c’est la même chose” (attributed to journalist and novelist Alphonse Karr, 1808-1890).

[6] Adapted from Robinson, T. M., Heraclitus: Fragments: A Text and Translation with a Commentary [1987] University of Toronto Press, p. 112.

[7] On the discussion about temporal parts and how objects persist, see: Hawley, K., “Temporal Parts”, The Stanford Encyclopedia of Philosophy (summer 2020 Edition), E. N. Zalta (ed.); Merkouris, P., 2014. (Inter)Temporal Considerations in the Interpretative Process of the VCLT: Do Treaties Endure, Perdure or Exdure? in Netherlands Yearbook of International Law, 2014(45), pp. 121–156.

[8] Jurisdictional Immunities of the State (Germany v. Italy: Greece intervening), Judgment, ICJ Reports 2012, p. 99.

[9] Kammerhofer J., “Between Pragmatism and Disenchantment: The Theory of Customary International Law after the ILC Project” in Merkouris, Kammerhofer, Arajärvi (eds), The Theory and Philosophy of Customary International Law and Its Interpretation (CUP 2022), p. 19.

[10] See below on Plato’s Forms.

[11] “At least theoretically, one can well prove the existence of a rule through a formal evidentiary process although the rule concerned is ascertained by virtue of non-formal criteria […]”, d’Aspremont J., Formalism and the Sources of International Law: A Theory of the Ascertainment of Legal Rules (CUP 2013), p. 152, more generally, see, 151-154.

[12] For a brilliant discussion on custom’s inherent plasticity and the role interpretation plays at different stages of a customary rule’s life which underlie this analysis: Chasapis Tassinis, O., 2020. Customary International Law: Interpretation from Beginning to End. European Journal of International Law, 31(1), pp. 235-267.

[13] Different aspects of a rule have to develop (or not) depending on the case. For instance, the customary rules on immunities were implicitly taken into account for the interpretation of the rule of exclusive jurisdiction of the flag state (a codification of customary law) at least twice: M/V “Norstar” (Panama v. Italy), Judgment, ITLOS Reports 2018–2019, paras. 216-218, 225; PCA, The ‘Enrica Lexie’ Incident, Italy v India, Award, Case No 2015-28 (21 May 2020), paras. 524-527. If this aspect were further explored in future affairs, one can imagine the customary rule on the exclusive jurisdiction of the flag state potentially becoming a rule on immunities.

[14] Legal Consequences of the Separation of the Chagos Archipelago from Mauritius in 1965, Advisory Opinion, ICJ Reports 2019, p. 95.

[15] On the paradox of opinio juris, see: Lefkowitz, D., 2008. (Dis)solving the Chronological Paradox in Customary International Law: A Hartian Approach. Canadian Journal of Law & Jurisprudence, 2008(21), pp. 129-148; Besson S., ‘Theorizing the Sources of International Law’ in Besson and Tasioulas (eds), The Philosophy of International Law (OUP 2010); Maiko Meguro, ‘Distinguishing the Legal Bindingness and Normative Content of Customary International Law’, ESIL Reflections 6:11 (2017).

[16] For another take on customary international law breaking the barriers of time and space, see d’Aspremont J., ‘The Custom-Making Moment in Customary International Law’ in Merkouris, Kammerhofer, Arajärvi (eds), The Theory and Philosophy of Customary International Law and Its Interpretation (CUP 2022).


Cite as: Başak Etkin, ‘The Changing Rivers of Customary International Law – The Interpretative Process as Flux’, ESIL Reflections 11:5 (2022).

© 2022. This work is licensed under a CC BY-NC-ND 4.0 licence.