Revision Hub  /  Module · Sources of International Law

Sources — Custom

Lecture · Sources I Dr Catriona Drew Reading: Roberts (2001); Koskenniemi (1989); Mendelson (1995); Cheng (1965); Charlesworth & Chinkin (2000)

Snapshot

The provocation usually quotes Article 38(1)(b) — "international custom, as evidence of a general practice accepted as law" — and asks whether the two-element doctrine is coherent. Thesis: custom remains the load-bearing source for rules binding states without treaty consent, but is genuinely indeterminate. Practice must be general and uniform (Asylum; Anglo-Norwegian Fisheries), weighted toward "specially-affected" states (North Sea para 73). Opinio juris is a separate element (North Sea para 77; Nicaragua paras 184, 207). The inductive orthodoxy sits uneasily beside the modern deductive practice of reading custom off UNGA resolutions and treaty texts (Texaco; Nuclear Weapons; Climate Change AO 2025). Roberts names this "traditional v modern"; Koskenniemi reads it as structurally indeterminate; Charlesworth asks whose practice counts. Answer-line: a flexible Nicaragua-Roberts sliding-scale, anchored in opinio juris, alive to the deductive-turn critique.

1. ISSUE — what the question really asks

A "discuss" question on custom tests one of three live tensions. Structural: are the two elements really separable, or does opinio juris just relabel practice? Identification: how do we know a rule has crystallised, and how short can the time-frame be? Legitimacy: when "general practice" is built from a small group of powerful states and ratified by opinio juris read off UNGA voting records, is custom still law-making — or law-laundering? Pick the tension the quotation foregrounds. Avoid history-of-custom preambles. Isolate the doctrinal hook (practice, opinio juris, specially-affected states, persistent objector, instant custom) and signpost the thesis.

2. RULE — Article 38(1)(b) and the two elements

Custom is the second listed source in Article 38(1) of the Statute of the International Court of Justice (1945). The text is short, the doctrine it supports enormous.

"The Court, whose function is to decide in accordance with international law such disputes as are submitted to it, shall apply: … (b) international custom, as evidence of a general practice accepted as law."

— Article 38(1)(b), Statute of the International Court of Justice 1945.

The orthodox reading splits the source into two elements that must both be present — a material element (state practice) and a psychological element (opinio juris sive necessitatis). ILC Draft Conclusions on Identification of CIL (2018), Conclusion 2, restates this.

2.1 State practice — the material element

Practice must be constant and uniform (Asylum 1950) and general but not universal (Anglo-Norwegian Fisheries; North Sea). North Sea para 73 adds the critical refinement: practice must include "specially affected" states. The materials are broad (diplomatic correspondence, legislation, military manuals, UNGA voting, conduct in IOs); ILC Conclusion 6 formalises the list. Inaction counts where reaction was due; private actors do not (Conclusion 4(3)).

2.2 Opinio juris — the psychological element

The locus classicus is North Sea para 77: practice alone is not enough; states must feel they are conforming to a legal obligation. "Ceremonial and protocol" acts do not generate custom. The methodological worry is under-determination — how do we know what a state "feels"? Mendelson (1995) calls opinio juris "the central problem": indispensable yet unknowable. The Court reads opinio juris off the same materials as practice, which Koskenniemi diagnoses as structural circularity.

2.3 Duration, treaty interplay, persistent objector, regional custom

The "time immemorial" rule is dead. North Sea para 73 confirms a short time-frame is no bar if practice is "extensive and virtually uniform" — the hook for Cheng's instant custom thesis (1965); Lauterpacht is the parent of the debate.

Treaty/custom interplay is governed by North Sea paras 71–76 and ILC Conclusion 11 (2018): a treaty rule may (i) codify a pre-existing custom, (ii) crystallise an emerging one, or (iii) generate a new one through the practice it produces. The North Sea test: norm-creating character (para 72); representative participation including specially-affected states (para 73); duration not a bar (para 74). The transition is "not lightly to be regarded as having been attained" (para 71).

The persistent objector rule lets a state that objected clearly and consistently from emergence escape the rule (Asylum; Anglo-Norwegian Fisheries); no defence to jus cogens. Regional custom is the other exception: Asylum recognises the category; Right of Passage (1960) confirms it can exist between two states.

3. Key cases

Five cases do almost all the work; cite them with paragraph numbers wherever you can.

North Sea Continental Shelf Cases (FRG v Denmark; FRG v Netherlands) ICJ Reports 1969, p. 3 — paras 71–78
Holding
Equidistance under Article 6 of the 1958 Geneva Convention had not become custom binding the non-party FRG. The Convention was not norm-creating (para 72); participation excluded specially-affected states (para 73); duration not itself a bar (para 74); no evidence states acted from legal obligation (para 78). The transition is "not lightly to be regarded as having been attained" (para 71).
Para 77
The acts must be "carried out in such a way, as to be evidence of a belief that this practice is rendered obligatory by the existence of a rule of law requiring it … There are many international acts, e.g. in the field of ceremonial and protocol, which are performed almost invariably, but which are motivated only by considerations of courtesy, convenience or tradition, and not by any sense of legal duty."
Asylum Case (Colombia v Peru) ICJ Reports 1950, p. 266
Holding
Colombia could not prove an "American international law" custom of unilateral qualification of political offences for diplomatic asylum. The party invoking a regional custom must show "constant and uniform usage" accepted as binding. Practice was inconsistent; even if the custom existed, Peru had repudiated it — persistent-objector argument avant la lettre.
Anglo-Norwegian Fisheries Case (UK v Norway) ICJ Reports 1951, p. 116
Holding
The alleged customary 10-mile rule for bay closures had not "acquired the authority of a general rule of international law"; Norway had "always opposed any attempt to apply it to the Norwegian coast". The leading authority for the persistent-objector rule; "general" practice is representative, not universal.
Military and Paramilitary Activities in and against Nicaragua (Merits) ICJ Reports 1986, p. 14 — paras 184, 186, 207
Paras 184 / 207
UNGA Resolution 2625 (XXV) (Friendly Relations Declaration 1970) treated as evidence of opinio juris on non-intervention; practice need not be in "absolutely rigorous conformity" with the rule.
Para 186 — violation v new practice
"The Court deems it sufficient that the conduct of states should in general be consistent with such rules and that instances of state conduct inconsistent with a given rule should generally have been treated as breaches of that rule not as an indication of the recognition of a new rule." A state appealing to exceptions within the rule "confirms rather than weakens" it.
SS Lotus (France v Turkey) PCIJ Reports, Series A, No 10 (1927)
Holding
"The rules of law binding upon States therefore emanate from their own free will… Restrictions upon the independence of States cannot therefore be presumed." Residual sovereign permission where no clear rule prohibits action. North Sea para 78 silently applied it; Nuclear Weapons AO (1996) resisted a permissive reading.
Nuclear Weapons AO (1996) · Climate Change AO (2025) ICJ Reports 1996, p. 226 para 67; ICJ 23 July 2025 paras 131, 313–315
Nuclear Weapons 1996
No customary prohibition on threat/use because the community was "profoundly divided" on whether fifty years' non-recourse constituted opinio juris. Divided opinio juris defeats custom even where practice is virtually universal.
Climate Change 2025
Two customary duties independent of the climate treaties: prevent significant transboundary environmental harm; co-operate. Treaty and customary obligations "inform each other" but "establish independent obligations that do not necessarily overlap" (para 314). The deductive turn in action.

4. APPLICATION — how the elements interact

4.1 Must opinio juris be proved separately?

The North Sea majority insisted on independent proof (paras 76–78): practice existed but there was "no evidence" states acted from legal obligation. Judge Sørensen, dissenting, argued sustained uniform practice should be presumptive of opinio juris, shifting the burden to the state denying the rule. Nicaragua came close to Sørensen — UNGA resolutions as opinio juris evidence; no rigorous-conformity requirement (para 186). Roberts names the divergence traditional (inductive) v modern (deductive).

4.2 Violation v new practice; jus cogens

The classic challenge: 130+ states reportedly practise torture; does that undermine the customary prohibition? Nicaragua para 186: a state that tortures but denies the conduct, or claims an exception within the rule, commits a breach that confirms the rule; only open defence of torture as a right would erode it — the substantive defect v procedural breach distinction.

Some rules harden into peremptory norms. Article 53 VCLT 1969 defines jus cogens as a norm "accepted and recognized by the international community of States as a whole" from which no derogation is permitted. Torture is canonical: Pinochet (No. 3) [2000] 1 AC 147 treated it as peremptory with consequences for immunity ratione materiae; Al-Adsani v UK (2001) 34 EHRR 11 accepted its jus cogens character (though upholding state immunity from civil suit). Persistent objection cannot defeat jus cogens.

4.3 UNGA resolutions and instant custom

Cheng (1965) argued a near-unanimous UNGA resolution with clear opinio juris could crystallise custom instantly. Texaco v Libya (1977) used resolutions on the customary expropriation standard; Nicaragua read non-intervention off Resolution 2625; Nuclear Weapons rejected instant custom on nuclear use; the 2025 Climate Change AO derives transboundary-harm and co-operation duties from treaty practice and resolutions. Danger: the more readily we read opinio juris off voting records, the closer custom moves to legislative wish-thinking.

5. CRITIQUE — Roberts, Koskenniemi, Charlesworth

Anthea Roberts (2001) 95 AJIL 757 distinguishes two methodologies inside Article 38(1)(b). Traditional custom is inductive: gather acts, infer the rule, prove opinio juris separately. Modern custom is deductive: start from authoritative statements (UNGA resolutions, treaty preambles), reason down to the rule, treat opinio juris as proved by the same materials. The divergence reflects two orientations — apologetic (descriptive) v utopian (progressive). Recommendation: a sliding scale — dense practice needs less explicit opinio juris; articulate opinio juris needs less dense practice.

Martti Koskenniemi, From Apology to Utopia (1989; reissue 2005), argues custom is structurally indeterminate. Every argument oscillates between an apologetic pole (the rule must reflect what states do) and a utopian pole (the rule must transcend what states do). The two-element doctrine is not a synthesis but a structure that lets the lawyer reach whichever conclusion is politically required.

Hilary Charlesworth, with Christine Chinkin in The Boundaries of International Law (MUP 2000), asks the prior question: whose practice counts? The traditional list privileges executive branches, foreign ministries and powerful states, and excludes domestic spheres that are feminised or privatised. "Specially-affected states" compounds this. Custom reinscribes existing distributions of power.

Defendable thesis: a moderate sliding-scale (Roberts), disciplined by Mendelson on opinio juris, humble before Charlesworth's point that the materials we count are themselves political.

6. Academic voices

Roberts argues custom has two faces. Traditional custom proceeds inductively from practice and demands separate proof of opinio juris; modern custom proceeds deductively from authoritative statements (UNGA resolutions, treaty texts), treating the same materials as evidence of both elements. The reconciliation: a sliding scale — dense, consistent practice needs less explicit opinio juris; articulate, near-universal opinio juris needs less dense practice.

— Anthea Roberts, "Traditional and Modern Approaches to Customary International Law: A Reconciliation" (2001) 95 American Journal of International Law 757.

Koskenniemi argues every customary argument oscillates between an apologetic pole (the law must reflect what states do) and a utopian pole (the law must transcend what states do, or it is not law against them). The two-element doctrine does not resolve the tension; it is a structure that lets the lawyer reach whichever conclusion the political moment requires.

— Martti Koskenniemi, From Apology to Utopia: The Structure of International Legal Argument (CUP, reissue 2005).

Charlesworth and Chinkin argue the conventional materials of practice — diplomatic correspondence, military manuals, foreign-ministry statements — privilege the conduct of executive branches and powerful states, and marginalise the domestic spheres where the practice of women and weak states is most visible. "Specially-affected states" deepens the bias. Custom is a sieve, not a mirror.

— Hilary Charlesworth & Christine Chinkin, The Boundaries of International Law: A Feminist Analysis (MUP, 2000).

Mendelson argues opinio juris is the central problem of customary international law: indispensable yet unknowable. The methodological discipline is to demand that opinio juris be argued and evidenced, not silently inferred from the same materials called practice.

— Maurice Mendelson, "The Subjective Element in Customary International Law" (1995) 66 British Yearbook of International Law 177.

7. Common pitfalls

Treating Article 38(1)(b) as a complete answer. The "rules about the rules" come from North Sea, Asylum, Anglo-Norwegian Fisheries, Nicaragua, and ILC Conclusions (2018). Quote the Article and move quickly to the case law.
Citing "state practice" without granularity. Examiners reward "specially-affected states" (North Sea para 73); inaction where reaction was due (ILC Conclusion 6); verbal v physical practice.
Treating opinio juris as a presumption from practice. North Sea paras 76–78 require both elements be proved; Sørensen dissents; Nicaragua is more flexible. Engage the disagreement.
Confusing violation with new practice. Nicaragua para 186: a state appealing to exceptions within the rule confirms it; only open repudiation contributes to a new rule.
Reading Lotus as a freedom-of-action banner. It is a residual rule for genuine indeterminacy; Nuclear Weapons AO (1996) resists mechanistic application to morally weighty silences.

8. Exam answer skeleton

  1. ISSUE. Identify the live tension (structure, identification, legitimacy). Thesis: a Nicaragua-Roberts sliding-scale, anchored in opinio juris.
  2. RULE. Article 38(1)(b); two elements (North Sea para 77); practice general, uniform, specially-affected states (para 73); treaty/custom (ILC Conclusion 11); persistent objector; regional custom; Lotus default.
  3. APPLICATION. Distinguish violation from new practice (Nicaragua para 186). For modern customs, engage the deductive turn (Texaco; Nuclear Weapons; Climate Change AO 2025). Note jus cogens (Art 53 VCLT; Pinochet; Al-Adsani) where relevant.
  4. CRITIQUE. Roberts (traditional v modern); Koskenniemi (apology v utopia); Charlesworth (whose practice).
  5. CONCLUDE. Take a position. Refuse the binary between rigid two-element orthodoxy and unconstrained instant-custom optimism.

9. Checklist