Defining International Law

"Considered legally binding." Not "binding", but considered binding, by the States themselves. That is a consent-based, positivist view: the law is what States accept as law.

"Civilised States." The nineteenth-century division of the world into civilised, barbarous and savage nations, dealt with in [What International Law Is, and Where It Came From]. The phrase is indefensible now, and later editions of the same treatise dropped it.

"In their intercourse with each other." States only, and only in their external dealings. Nothing about individuals, nothing about organisations, and nothing about how a State treats its own people.

The other classical definitions

An examiner expects more than one, and the four below are the ones the standard reading list carries.

J.L. Brierly: the law of nations, or international law, may be defined as the body of rules and principles of action which are binding upon civilised States in their relations with one another. Note what Brierly changes: he says binding upon, not "considered binding by". That is a small phrase carrying a large claim, because it does not make the obligation depend on the State's own opinion.

Charles Fenwick: international law may be defined in broad terms as the body of general principles and specific rules which are binding upon the members of the international community in their mutual relations. Fenwick's phrase members of the international community is wider than "States", and it is the opening through which organisations later walked.

J.G. Starke, whose Introduction to International Law your syllabus lists in two editions, gives the definition that is closest to the modern one and is the safest to reproduce: international law may be defined as that body of law which is composed for its greater part of the principles and rules of conduct which States feel themselves bound to observe, and therefore commonly do observe, in their relations with each other, and which includes also the rules of law relating to the functioning of international institutions and organisations, their relations with each other and their relations with States and individuals, and certain rules of law relating to individuals and non-State entities so far as the rights and duties of such individuals and non-State entities are the concern of the international community.

Torsten Gihl: the body of rules of law which apply within the international community or society of States. Short, and it moves the emphasis from the parties to the community they form.

Starke's definition is long, and its length is the point. It is long because it has to name three things Oppenheim's did not: international institutions, individuals, and non-State entities.

Why the classical definitions broke down

Four developments made "States only" impossible to maintain, and each one is a chapter of this book.

International organisations acquired legal personality. In 1949 the International Court of Justice was asked whether the United Nations could bring an international claim for injury to its agent, and in the Reparation for Injuries advisory opinion, worked in [Subjects of International Law], it held that the Organization has a large measure of international personality and the capacity to operate on the international plane, though it is certainly not a super-State. A definition that mentions only States cannot accommodate a claimant that is not a State.

Individuals acquired duties. The Nuremberg and Tokyo tribunals after 1945 tried individuals for crimes under international law, and the principle that individuals can be punished directly by international law is now embodied in the Rome Statute of the International Criminal Court. If international law can hang a man, it plainly reaches him.

Individuals acquired rights. Human rights treaties give individuals rights against their own State, and several of them let an individual petition an international body. That is the exact opposite of the Westphalian arrangement in which how a State treated its own subjects was nobody else's business.

Some rules stopped depending on consent. A small class of rules, called jus cogens or peremptory norms, binds every State whether or not it agreed, and no treaty contrary to them is valid. The prohibitions on genocide, slavery, torture and aggression are the standard examples. Article 53 of the Vienna Convention on the Law of Treaties gives them their statutory form, and they are worked out in [Validity of Treaties, and Jus Cogens]. A definition built entirely on consent cannot explain them.

The modern definition

Putting those four developments into a sentence gives the definition to write in an exam:

International law is the body of rules and principles which States, international organisations and, in defined respects, individuals and other entities accept as binding in their mutual relations, derived from treaties, custom and general principles of law, and including a small core of peremptory norms which bind regardless of consent.

Every clause of that sentence is doing work. States, international organisations and individuals answers the who. Accept as binding in their mutual relations keeps the consensual character that is still true of most of the subject. Treaties, custom and general principles names the sources, which are the subject of [The Sources of International Law]. A small core of peremptory norms admits the exception without overstating it.

A table you can reproduce under time pressure

A worked example

Two years ago the government of State A, a party to the Genocide Convention and to the Charter, orders its army to expel an ethnic minority from a border province. Thousands die. A commander, Colonel B, gives the orders on the ground. An international organisation, the World Food Programme, has a warehouse in the province, and one of its staff, an Italian national, is shot at a checkpoint.

Take Oppenheim's definition and ask what it can explain. It can explain that State A has broken a treaty it accepted as binding, and that other State parties may complain. That is all. It cannot explain Colonel B's personal criminal liability, because individuals are not in the definition. It cannot explain the organisation's claim for its warehouse and its staff member, because organisations are not in it either. And if State A had never ratified the Genocide Convention, Oppenheim's consent-based definition would struggle to explain why the prohibition binds it anyway.

Now take Starke's, with jus cogens added. Colonel B is liable because international law imposes duties directly on individuals. The World Food Programme may bring a claim for the injury to its agent under the reasoning in the Reparation for Injuries advisory opinion, and separately the agent's national State may protect him, the competition between the two being reconciled case by case. And the prohibition of genocide binds State A regardless of ratification, because it is peremptory.

That is what a change of definition buys, and it is what an examiner is testing when the question says "critically examine".

What it does NOT mean

A definition is not a source. No case has ever been decided by quoting Oppenheim. Definitions organise the subject; article 38 of the Statute of the International Court of Justice supplies the law.

Saying individuals are subjects does not mean they are equal subjects. An individual cannot make a treaty, cannot sit in the General Assembly and cannot bring a case before the International Court, whose Statute in article 34 says only States may be parties in cases before the Court. The correct statement is that individuals have some rights and some duties directly under international law, not that they stand where States stand.

"Civilised nations" is not a live test. Where the phrase survives, in article 38(1)(c), it is read as meaning nothing more than the legal systems of the world generally. No State is excluded from international law today for want of civilisation, and any answer that repeats the phrase without noticing the problem loses more than it gains.

Limits and criticism

Every definition is circular to a degree. Oppenheim defines the law by what States consider binding, which assumes they already know which rules are law. Brierly avoids the circle by asserting the rules are binding, which is the very thing an objector denies. There is no way out of this by better drafting; it is the same difficulty that makes [Is International Law Really Law?] a real question.

Definitions lag behind practice. Oppenheim wrote before the League, before the United Nations, before human rights treaties and before the International Criminal Court. Starke's definition is capacious enough for all of them because it was written after them, which is less a virtue of the author than a fact about dates.

The subject may now be too varied to define usefully. Trade law, the law of the sea, human rights and the law of armed conflict have their own institutions, their own courts and their own habits, so much so that international lawyers argue about the fragmentation of the subject. A single sentence that covers all of them will do so only at a level of generality that tells a student very little.

Quick revision

The phrase international law was coined by Bentham, 1780. The older name is the law of nations.

Oppenheim, 1905: the body of customary and conventional rules considered legally binding by civilised States in their intercourse with each other. Two sources, consent-based, States only, and the offensive adjective.

Brierly: rules binding upon civilised States. Says binding upon, not considered binding by.

Fenwick: binding upon the members of the international community.

Starke: the fullest of the classical definitions, expressly covering international institutions, individuals and non-State entities.

Gihl: rules applying within the international community or society of States.

Four developments broke the States-only definitions: international organisations acquired personality (the Reparation for Injuries advisory opinion, 1949, in [Subjects of International Law]); individuals acquired duties (Nuremberg, the International Criminal Court); individuals acquired rights (human rights treaties); and jus cogens, in article 53 of the Vienna Convention, binds without consent.

Modern definition to write: rules and principles binding on States, international organisations and, in defined respects, individuals, derived from treaties, custom and general principles, and including a peremptory core that binds regardless of consent.

Test yourself

1. State Oppenheim's definition and identify three defects in it.

"Law of Nations or International Law is the name for the body of customary and conventional rules which are considered legally binding by civilised States in their intercourse with each other." Three defects: it names only two sources and omits general principles of law, which article 38(1)(c) of the Statute of the International Court of Justice recognises; it limits the subjects to States, which cannot accommodate international organisations or the direct duties international law imposes on individuals; and it rests obligation entirely on what States consider binding, which cannot explain peremptory norms that bind without consent. The reference to "civilised States" is a fourth defect and is now indefensible.

2. What is the difference between saying rules are "considered legally binding by States" and saying they are "binding upon States"?

The first makes the law depend on the opinion of the States bound, which is a positivist, consent-based account; if a State does not consider a rule binding, on that view it is not bound. The second asserts an obligation independent of that opinion, which is what is needed to explain jus cogens and, more generally, why a new State is bound by customary rules it never voted for. Oppenheim takes the first position, Brierly the second.

3. Which definition would you use to answer a question about the liability of an army officer for war crimes, and why?

Starke's, because it is the only classical definition that expressly includes rules of law relating to individuals and non-State entities so far as their rights and duties are the concern of the international community. Oppenheim's, Brierly's and Gihl's are limited to States and cannot reach an individual at all.

4. Why is the phrase "civilised nations" still in the Statute of the International Court of Justice, and how should a student handle it?

Article 38(1)(c) was drafted in 1920 for the Permanent Court and reproduced in 1945 without change, so the phrase survives by inertia rather than by intention. A student should state that it is understood today as meaning the legal systems of the world generally, that it excludes no State, and that it is a relic of the era when international law was made by a small group of European powers and applied to everyone else.

5. "Any definition of international law is either too narrow to be accurate or too wide to be useful." Discuss briefly.

There is force in it. Narrow definitions such as Oppenheim's are accurate about the core, relations between States, but cannot accommodate organisations, individuals or peremptory norms. Wide definitions such as Starke's cover all of those but do so at such a level of generality that they tell a beginner little about what the rules actually are. The reply is that a definition is not meant to do the work of the subject: it fixes who is bound and by what, and article 38 and the chapters that follow supply the content. A definition that gets the who and the what right is doing its job even if it is uninformative about the rest.

6. Distinguish the sources of international law from the definitions of it.

A definition states what the subject is and whom it binds. A source states where a rule comes from and how you prove it exists. Bentham, Oppenheim, Brierly, Fenwick, Starke and Gihl offer definitions. Article 38(1) of the Statute of the International Court of Justice lists the sources: international conventions, international custom, the general principles of law recognised by civilised nations, and, as subsidiary means, judicial decisions and the teachings of the most highly qualified publicists. An answer that recites definitions when asked for sources, or the reverse, answers a different question from the one set.