‘Customary Law Interpretation as a Tool’ Series

PDF Version

Vol 11, Issue 4
Editorial board: Federico Casolari, Patrycja Grzebyk (editor-in-chief), Ellen Hey, Guy Sinclair, and Ramses Wessel
Ivo Tarik de Vries-Zou* (University of Groningen)

© Image by Ivo Tarik de Vries-Zou (cc)



To identify a treaty means to affirm that certain subjects intended to become bound under international law. The interpretation of a treaty thereafter goes on to detail the exact terms by which the subjects agreed to become bound. An interpretation thus reconstructs the common intent of subjects and must be based on material that evinces this common intent.[1] Whether an interpreter may give the terms of a treaty a static or evolutive meaning, or a narrow or broad scope, likewise depends on the intentions of all parties.[2] It is unsurprising then that if an interpreter were to mostly use material produced by one party during negotiations, to which no other state had access, she would face the criticism of honouring only the intentions of the former party, instead of a common intent.[3] If, alternatively, she read a treaty in the light of an object and purpose that the parties by all accounts did not wish to give their treaty, she would be criticised for discounting their intentions completely.

The origin and nature of the object to be interpreted will, in other words, dictate how it must be interpreted. Many writers thus consider that, as custom does not have its origin in an intent to become bound, its interpretation is also divorced from any quest for authorial intent.[4] But many of these accounts do not consider what an interpretation of custom would then otherwise expand on. The authors often hold that interpretation here entails any ‘ascription or elucidation of meaning’, such that it occurs as soon as one gives a customary rule a particular content.[5] This portrayal of the nature of interpreting custom says very little. It says as little as saying the interpretation of a treaty expands on the contents thereof; this is true, but is also only half the story, and the less interesting part at that. The crux of the matter is precisely that an interpreter may not construct the content of a treaty in whatever way she would like, but only as per the common intent of the parties.

Despite describing the interpretation of custom in reductive terms, writers still profess to know the means for interpreting custom. It is often noted that elements of treaty interpretation apply by analogy to the interpretation of custom, elements that would normally be used to ascertain a common intent. Custom may, for example, be interpreted textually, teleologically, or systemically.[6] In effect, the literature attempts to list the means of interpreting custom without a definite answer on what end such interpretation must serve.

This Reflection submits that it is necessary to consider closely the nature of what is to be interpreted before postulating the means of interpretation. It is thus demonstrated below that even treaties and unilateral acts, which both have their basis in the intent of subjects, are still different in nature and thus interpreted according to different rules. It is argued thereafter that custom is not based on intent, but rather the common convictions of states. It would follow that custom must then be interpreted so as to honour such convictions, much in the same way as a common intent conditions the reading of a treaty. The Reflection ends by noting how such convictions may serve as a yardstick for evaluating any particular interpretation of custom.

Beyond Treaties

In the Fisheries Jurisdiction case of 1998, the International Court of Justice described a declaration made under Art. 36(2) of its Statute as a ‘unilateral act’ which also sets up ‘a consensual bond and the potential for a jurisdictional link’ with any other states that have made such declarations.[7] Since such declarations are then not exactly treaties, the Court considered that the principles of treaty interpretation, as embodied in Arts. 31 and 32 of the Vienna Convention on the Law of Treaties, could ‘only apply analogously to the extent compatible with the sui generis character of the unilateral acceptance of the Court’s jurisdiction’.[8]

Specifically, the Court stated that it would have to pay ‘due regard to the intention of the State concerned at the time when it accepted the compulsory jurisdiction of the Court’.[9] While the interpretation of a treaty thus relies on evidence of a common intent, the Court noted it would decide the intent of the author of a declaration using not only the text thereof but also ‘evidence regarding the circumstances of its preparation’.[10] The Court referred in this respect to ‘Canadian ministerial statements, parliamentary debates, legislative proposals and press communiqués’.[11] Interpreters of a treaty would, by contrast, have to give material which is mostly internal to a state mere supplementary value in their interpretation, lest they interpret a treaty primarily as per the intentions of only one party.[12]

The Court also noted that how it would read the said declaration depended on its distinct nature. Spain, the applicant, argued that the classic maxim of contra proferentem—the rule by which an agreement drafted by one party may, where ambiguous, be read so as to benefit that party the least—could apply to the interpretation of a declaration.[13] While the Court said it would, as with a treaty, interpret the terms of a declaration ‘in a natural and reasonable way’, the maxim had only ‘a role to play in the interpretation of contractual provisions’, and it was thus of no application in the interpretation of a declaration.[14]

What the above example shows is that the norms for interpreting treaties must be adapted depending on the object of concern.[15] Thus, the Court could only say what material could be used to interpret the declaration, and in what way this had to be read, by first considering its character as a unilateral act. Similar considerations would have to steer any account of how to interpret custom. To know within what parameters an interpreter of custom must operate, she shall first have to consider the peculiar nature of custom and how this differs from the essential nature of an international agreement or unilateral act.

The Nature of Custom

In contrast to treaties and unilateral acts, custom does not flow directly from the intentions of states. Norms of customary law cannot, as with a treaty, be traced back to a moment in which states showed a concerted will to bring these into being. While treaties and comparable acts are at heart promises that subjects make about how they shall act in the future, no formal promises are required in order to evaluate an ongoing practice as mandatory under customary law. All that is needed is a common conviction that such is the case.

This point may, somewhat paradoxically, be best illustrated with cases wherein a group of states desired to see a change in the norms of custom at the behest of another group. An example is the contest over the norms for the nationalisation of natural resources that took place in the second half of the twentieth century. Newly independent states passed resolutions in the General Assembly of the United Nations which stated that, if any measures of nationalisation would lead to a dispute about compensation, this would have to be ‘settled under the domestic law of the nationalizing State’ rather than international law.[16] Thus, these texts did not say what the law should be, but what the law was; they asserted what rights a nationalising state has. This characterisation of what states were already bound by was then also voted against by most industrialised states, which thereby prevented the proclaimed rights from becoming custom.[17] One group of states did not, in other words, share the convictions of another group.

A similar struggle is observed in the area of nuclear weapons. In its Nuclear Weapons opinion, the International Court of Justice took note of how some participants to the proceedings cited resolutions of the General Assembly as confirming a specific customary ban on the use of nuclear weapons. But many of these had received ‘substantial numbers of negative votes and abstentions’.[18] Regarding the content of one resolution from 1961, the Court stated that this did not as such express a prohibition of nuclear weapons. The General Assembly had rather applied general norms of custom to such weapons, suggesting it did not believe a particular ban thereof to exist; ‘if such a rule had existed, the General Assembly could simply have referred to it’.[19]

Of course, had a resolution expressed a specific ban, those states in possession of nuclear weapons would undoubtedly have opposed its adoption and precluded the newly pronounced rule from gaining force as custom. And any conviction would also need to be observed in practice. The Court thus observed that the ‘nascent opinio juris’ against any use of nuclear weapons, as proclaimed in resolutions of the General Assembly, stood opposed to the ‘still strong adherence to the practice of deterrence’ by a number of states.[20]

However, the key takeaway is that if states wish to alter custom, they will claim the desired norm already exists, in the hope that this conviction about the state of international law is not readily opposed by other affected states. The concept of conviction thereby behaves differently from that of intent. If a state intends to be bound by a rule and to act accordingly from tomorrow on, that is the end of the matter. If a number of states agree to behave henceforth in a certain manner, this is the end of the matter as well. We might suspect the states will not in fact follow through on their promises, but that does not negate their intent to become bound. Yet if a state claims that under custom it must act in a given way and that other states must do so too, its appraisal is only a preface to, if not a mere footnote in, an unending process of claims and counterclaims.[21] Every individual exposé of a conviction, be it through word or deed, is thus at once relevant and falsifiable. In effect, material that is thus pertinent to identifying a customary norm is sprawled over time and space. If we were even to imagine custom as a treaty, we would bring to life, in the words of Robert Kolb, ‘the legal monster of a pact infinitesimally split in time’.[22]

We observe this distinction between intent and conviction further in the fact that only custom will be subject to the arduous process of codification. As Wouter Werner notes in his Reflection in this series, while the ‘paratext’ of treaties will say that ‘the parties have agreed’ to new norms and detail ‘why it was necessary or desirable to create something new’, reports that restate customary law ‘are preceded by prefaces and introductions, which spell out why it was deemed necessary to restate existing law’ and ‘set out which methods were used to find the law as it is’.[23] The latter texts thus claim to have tamed the legal monster of custom, at least to a degree, by compiling a registry that accurately reflects the common convictions of states. Of course, some treaties codify custom, but these betray precisely the point that states can intend to become bound conventionally to rules they have already accepted through another modality of consent.

An End in Sight

It has been established above that a meaningful account of the means of interpreting custom must appreciate the peculiar nature of this source of international law. Given that custom, as shown in the immediately preceding part, has its basis in the common convictions of states, it is proposed that the interpretation of any of its norms also has to elaborate on such a conviction. This would be the conviction, as clarified above, that the process of identifying a customary rule would verify as that generally shared by states, and one which they follow in practice, to the exclusion of contradictory convictions.

To regard the interpretation of custom as such does not mean we conflate interpretation with identification. The relationship proposed is actually comparable to the link between identifying and interpreting a treaty. By finding that a text embodies a treaty, we not only say that a handful of subjects intended to be bound thereunder, but we also delimit what materials we may rely on when interpreting this treaty, or, in other words, when we specify what exactly the said subjects wished to be bound by. Similarly, if certain material shows that states are convinced a norm governs their practice as custom, it would be at variance with common sense to not use the material to also specify what they believe this norm requires of them.[24]

Such material would of course in many cases be vast. For, as previously explained, and unlike norms within a treaty, ‘the legal principles of customary international law, even if they relate to the same subject matter, are not united into a single entity by a single will of the contracting parties’.[25] The convictions of states could find their place in a resolution of the General Assembly but could as well be contained in a treaty, military handbooks, or the judicial decisions of national courts, to give only a handful of examples.

States can of course also communicate their convictions through acts or omissions. As one writer notes, ‘does not every behaviour of any actor manifest the interpretation he has made upstream of what can or cannot be done, of what must or must not be done, whatever the order of consideration according to which this appraisal is made (social, political, opportunity, legal system)?’[26] Another scholar has gone so far as to say that, in the context of interpreting custom, the relevant practice of states is ‘the ultimate point of reference one has when clarifying a particular legal rule’.[27]

In some cases, though, the practice of states would militate that a particular instrument is especially significant to the interpretation of a norm. One example would be the Articles on the Responsibility of States for Internationally Wrongful Acts, which, though not binding, are widely regarded to reflect custom. In Gabčíkovo-Nagymaros, the parties thus agreed the conditions for a state of necessity were codified in Art. 33 of the draft version of the Articles.[28] But the major point of contention rested on how the conditions applied to the facts and, to that end, the Court made sure to clarify the meaning of some of the terms used in Art. 33, before applying these.[29]

Evaluative Dimension

As a final note, it must be said that a common conviction, just as the concept of a common intent, has to it an evaluative dimension. It is said that custom is open to ‘systemic interpretation’, for example, in that a rule must be interpreted in connection to other norms in the system of international law.[30] Yet an interpreter must decide their relationship in line with the persuasions of states. Take as an example the principles of uti possidetis and self-determination. In 1986, the International Court of Justice observed that while the two seem to conflict, the territorial security that the former promised had ‘induced African States judiciously to consent to the respect of colonial frontiers, and to take account of it in the interpretation of the principle of self-determination’.[31] In turn, if an interpreter were to interpret the norms as conflicting with each other, her interpretation would not be beyond the particular reproach that it does not reflect the convictions of states themselves.

It is also said that custom may be interpreted teleologically, with an object and purpose in mind that the relevant rule must fulfil. But just as the object and purpose of a treaty shall have its basis in the actual terms thereof, or any other material that showcases the common intent of the parties, thus must an interpreter aim to establish the teleology of a norm on the basis of the common conviction of states. It is after all only in its application and handling by states that a norm is given a life and purpose. With the principle of uti possidetis, the Court thus observed that the rule had emerged in South American practice with the main ‘aim of securing respect for the territorial boundaries at the moment when independence is achieved’.[32] Since the rule had then a broad purpose, the Court said that it could also apply to newly independent African states, which had recognised its application in their practice.[33] The Court thus discerned the purpose of the custom from its origins in a local practice and buttressed its application to a new landscape by noting the generality of its object and purpose, and the convictions of the affected states. At no point did the Court suggest that the object and purpose of the principle descended from heaven, so to speak, or could overrule a common conviction of states.[34]


The literature has in the past few years become increasingly concerned with the interpretation of custom. Authors have in the first place had to consider whether custom is even subject to a process of interpretation. In many cases, their response is that it must be, if only because any application of a customary norm presupposes its interpretation, and not its identification alone. A norm may be identified as being customary, but whether it can be applied to a certain context will still require a specification of its content. It has been argued above, though, that not any specification of the content of a norm will do. Just as the interpretation of a treaty must elaborate on the common intent of parties, so must the interpretation of custom be concerned with an investigation into the common convictions of states. Such a common conviction is after all the basis of a customary norm. Accordingly, the means for interpreting custom must be made to elucidate such common convictions. Each interpretation of custom must be subjected to the critique of whether it actually accords with the common convictions of states. In other words, it is only by closely considering the nature of custom that we can begin to develop the appropriate means of interpretation, and at the same time a standard to evaluate the conclusions of an interpreter of customary law.

* Ivo Tarik de Vries-Zou is a doctoral researcher at the University of Groningen. This contribution is based on a discussion during the workshop The Role of Interpretation in the Practice of Customary International Law, organized by the TRICI-Law project. This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 Research and Innovation Programme (Grant Agreement No. 759728).

[1] ‘Report of the International Law Commission on the Work of Its Seventieth Session’ (30 April–1 June, 2 July–10 August 2018) UN Doc A/73/10, 24.

[2] See Dispute regarding Navigational and Related Rights (Costa Rica v. Nicaragua) (Judgment) [2009] ICJ Rep 213, paras. 63–66.

[3] Such material will only be of supplementary value in treaty interpretation. See Oliver Dörr, ‘Article 32’ in Oliver Dörr and Kirsten Schmalenbach (eds), Vienna Convention on the Law of Treaties: Commentary (Springer 2018) 617–18.

[4] For discussions on this, see Kostiantyn Gorobets, ‘Practical Reasoning and Interpretation of Customary International Law’ in Panos Merkouris, Jörg Kammerhofer, and Noora Arajärvi (eds), The Theory, Practice, and Interpretation of International Law (Cambridge University Press 2022) 374; Peter G Staubach, The Rule of Unwritten International Law: Customary Law, General Principles, and World Order (Routledge 2018) 111–12. But see also Panos Merkouris, Article 31(3)(c) VCLT and the Principle of Systemic Integration (Brill 2015) 267–68.

[5] Orfeas Chasapis Tassinis, ‘Customary International Law: Interpretation from Beginning to End’ (2020) 31 European Journal of International Law 235, 240. See Marina Fortuna, ‘Different Strings of the Same Harp: Interpretation of Rules of Customary International Law, Their Identification and Treaty Interpretation’ in Merkouris, Kammerhofer, and Ara­järvi (n 4) 394, 402–4; Riccardo Di Marco, ‘Customary International Law: Identification versus Interpretation’ in ibid 415, 420, 422; Serge Sur, ‘La créativité du droit international’ (2012) 363 Recueil des Cours de l’Académie de Droit International 9, 294. But cf Gorobets (n 4) 375; Staubach (n 4) 136–37.

[6] See, for example, Fortuna (n 5) 404. See also Alexander Orakhelashvili, The Interpretation of Acts and Rules in Public International Law (Oxford University Press 2008) 497, who writes, at 297, that intent ‘can and must be identified with regard to all acts’ in the process of their interpretation, and yet, at 497, says, without any mention of intent, that the methods of treaty interpretation are also applicable to custom because the ‘rationale underlying the interpretation of a treaty and of custom is the same—the preservation of the integrity of the legal rule.’

[7] (Spain v. Canada) (Jurisdiction) [1998] ICJ Rep 432, para. 46.

[8] ibid.

[9] ibid para. 49.

[10] ibid.

[11] ibid.

[12] cf Oil Platforms (Islamic Republic of Iran v. United States of America) (Preliminary Objections) [1996] ICJ Rep 803 (diss op Schwebel) 882–84.

[13] Fisheries Jurisdiction (n 7) para. 43.

[14] ibid paras. 49, 51.

[15] See ‘Report of the International Law Commission on the Work of Its Fifty-Eighth Session’ (1 May–9 June, 3 July–11 August 2006) UN Doc A/61/10, 178.

[16] UNGA Res 3281 (XXIX) (12 December 1974). See also UNGA Res 3201 (S-VI) (1 May 1974); UNGA Res 3171 (XXVIII) (17 December 1973).

[17] Texaco Overseas Petroleum Company and California Asiatic Oil Company v. The Government of the Libyan Arab Republic (1979) 53 ILR 389, paras. 83, 88.

[18] Legality of the Threat or Use of Nuclear Weapons (Advisory Opinion) [1996] ICJ Rep 226, para. 71.

[19] ibid para. 72.

[20] (n 19) para. 73.

[21] Similarly, Pauline Westerman, ‘Opinio Juris: Test, Filter, Ideal or Map?’ in Pauline Westerman, Kostiantyn Gorobets, and Andreas Hadjigeorgiou (eds), Conceptual (Re)Constructions of International Law (Elgar 2022, forthcoming).

[22] ‘Selected Problems in the Theory of Customary International Law’ (2003) 50 Netherlands International Law Review 119, 143.

[23] Wouter G Werner, ‘Custom as Rewritten Law: The Text and Paratext of Restatement Reports’, ESIL Reflections 11:3 (2022) (emphasis in original).

[24] Similarly, Nina Mileva, ‘The Role of Domestic Courts in the Interpretation of Customary International Law: How Can We Learn from Domestic Interpretive Practices?’ in Merkouris, Kammerhofer, and Arajärvi (n 4) 460.

[25] Albert Bleckmann, ‘Zur Feststellung und Auslegung von Völkergewohnheitsrecht’ (1977) 37 Zeitschrift für ausländisches öffentliches Recht und Völkerrecht 504, 527 (tr author; ‘Zwar sind die Rechtssätze des Völkergewohnheitsrechts, selbst wenn sie sich auf dieselbe Materie beziehen, nicht durch einen einheitlichen Willen der Vertragspartner zu einer Einheit zusammengefaßt’).

[26] Denis Alland, ‘L’interprétation du droit international public’ (2013) 362 Recueil des Cours 41, 88 (tr author; ‘est-ce que tout comportement d’un acteur quelconque ne manifeste pas l’interprétation qu’il a opérée en amont de ce qui peut ou ne peut pas se faire, de ce qui doit ou ne doit pas se faire, et ce quel que soit l’ordre de considération en fonction duquel l’appréciation est portée (usage social, politique, opportunité, système juridique)?’).

[27] Gorobets (n 4) 375.

[28] (Hungary/Slovakia) (Judgment) [1997] ICJ Rep 7, paras. 50, 52, 54. Note that the state of necessity, in the final version of the Articles, is considered in Art. 25.

[29] ibid para. 54. See also Fortuna (n 5) 410–13.

[30] Merkouris (n 4) 265–67.

[31] Frontier Dispute (Judgment) [1986] ICJ Rep 554, para. 25.

[32] ibid paras. 20, 23.

[33] ibid paras. 21–22, 24.

[34] cf Tassinis (n 5) 254–55; Staubach (n 4) 139ff.


Cite as: Ivo Tarik de Vries-Zou, ‘Common Convictions and the Interpretation of Custom’, ESIL Reflections 11:4 (2022).

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