TRADITIONAL AND MODERN APPROACHES TO
CUSTOMARY INTERNATIONAL LAW: A RECONCILIATION
By Anthea Elizabeth Roberts*
I. THE PROBLEM OF TRADITIONAL AND MODERN CUSTOM
The demise of custom as a source of international law has been widely forecasted.1 This
is because both the nature and the relative importance of custom’s constituent elements are
contentious. At the same time, custom has become an increasingly significant source of law
in important areas such as human rights obligations.2 Codification conventions, academic
commentary, and the case law of the International Court of Justice (the Court) have also
contributed to a contemporary resurrection of custom.3 These developments have resulted
in two apparently opposing approaches, which I term “traditional custom” and “modern
custom.” The renaissance of custom requires the articulation of a coherent theory that can
accommodate its classic foundations and contemporary developments. This article seeks to
provide an enriched theoretical account of custom that incorporates both the traditional
and the modern approaches rather than advocating one approach over the other.
The Statute of the International Court of Justice describes custom as “evidence of a gen-
eral practice accepted as law.”4 Custom is generally considered to have two elements: state
practice and opinio juris.5 State practice refers to general and consistent practice by states,
while opinio juris means that the practice is followed out of a belief of legal obligation.6 This
distinction is problematic because it is difficult to determine what states believe as opposed
to what they say. Whether treaties and declarations constitute state practice or opinio juris
is also controversial. For the sake of clarity, this article adopts Anthony D’Amato’s distinc-
tion between action (state practice) and statements (opinio juris).7 Thus, actions can form
custom only if accompanied by an articulation of the legality of the action.8 Opinio juris
The research for this article was initially undertaken for my LLB dissertation at the Australian National
University and was completed while I was teaching in the university’s Faculty of Law. I wish to thank Professor
Hilary Charlesworth for her insightful comments on various drafts of the manuscript.
E.g., N. C. H. Dunbar, The Myth of Customary International Law, 1983 AUSTL. Y.B. INT’L L. 1; J. Patrick Kelly, The
Twilight of Customary International Law, 40 VA. J. INT’L L. 449 (2000).
THEODOR MERON, HUMAN RIGHTS AND HUMANITARIAN NORMS AS CUSTOMARY LAW (1989).
Eduardo Jiménez de Aréchaga, Custom, in CHANGE AND STABILITY IN INTERNATIONAL LAW-MAKING 1, 2–3
(Antonio Cassese & Joseph H. H. Weiler eds., 1988) [hereinafter CHANGE AND STABILITY]; W. Michael Reisman, The
Cult of Custom in the Late 20th Century, 17 CAL. W. INT’L L.J. 133 (1987).
INTERNATIONAL COURT OF JUSTICE STATUTE Art. 38(1)(b) [hereinafter ICJ STATUTE].
North Sea Continental Shelf (FRG/Den.; FRG/Neth.), 1969 ICJ REP. 3, 44 (Feb. 20).
RESTATEMENT (THIRD) OF THE FOREIGN RELATIONS LAW OF THE UNITED STATES §102(2) (1987) [hereinafter
RESTATEMENT]; IAN BROWNLIE, PRINCIPLES OF PUBLIC INTERNATIONAL LAW 4–11 (5th ed. 1998); MICHAEL BYERS,
CUSTOM, POWER, AND THE POWER OF RULES 130 (1999); ANTHONY A. D’AMATO, THE CONCEPT OF CUSTOM IN
INTERNATIONAL LAW 49 (1971) (citing FRANÇOIS GÉNY, MÉTHODE D’INTERPRÉTATION ET SOURCES EN DROIT PRIVÉ
POSITIF §110 (1899)).
D’AMATO, supra note 6, at 89–90, 160. This distinction has been criticized by many commentators, who include
certain statements as forms of state practice. RESTATEMENT, supra note 6, §102; BROWNLIE, supra note 6, at 5–6;
Michael Akehurst, Custom as a Source of International Law, 1974–75 BRIT. Y.B. INT’L L. 1, 2, 35. The distinction is also
impliedly inconsistent with some case law of the Court. Fisheries Jurisdiction (UK v. Ice.), Merits, 1974 ICJ REP.
3, 47, 56–58, 81–88, 119–20, 135, 161 ( July 25); North Sea Continental Shelf, 1969 ICJ REP. at 4, 32–33, 47, 53.
D’AMATO, supra note 6, at 74–75.
758 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
concerns statements of belief rather than actual beliefs.9 Further, treaties and declarations
represent opinio juris because they are statements about the legality of action, rather than
examples of that action.10 As will be demonstrated below, traditional custom and modern
custom are generally assumed to be alternatives because the former emphasizes state prac-
tice, whereas the latter emphasizes opinio juris.11
What I have termed traditional custom results from general and consistent practice fol-
lowed by states from a sense of legal obligation.12 It focuses primarily on state practice in the
form of interstate interaction and acquiescence. Opinio juris is a secondary consideration
invoked to distinguish between legal and nonlegal obligations.13 Traditional custom is evolu-
tionary14 and is identified through an inductive process in which a general custom is derived
from specific instances of state practice.15 This approach is evident in S.S. Lotus,16 where the
Permanent Court of International Justice inferred a general custom about objective
territorial jurisdiction over ships on the high seas from previous instances of state action and
By contrast, modern custom is derived by a deductive process that begins with general state-
ments of rules rather than particular instances of practice.18 This approach emphasizes opinio
juris rather than state practice because it relies primarily on statements rather than actions.19
Modern custom can develop quickly because it is deduced from multilateral treaties and
declarations by international fora such as the General Assembly, which can declare existing
customs, crystallize emerging customs, and generate new customs.20 Whether these texts
become custom depends on factors such as whether they are phrased in declaratory terms,
supported by a widespread and representative body of states, and confirmed by state prac-
tice.21 A good example of the deductive approach is the Merits decision in Military and
Paramilitary Activities in and Against Nicaragua.22 The Court paid lip service to the traditional
test for custom but derived customs of non-use of force and nonintervention from state-
ments such as General Assembly resolutions.23 The Court did not make a serious inquiry into
Id. at 35–39; Akehurst, supra note 7, at 36–37.
Cf. D’AMATO, supra note 6, at 89 (arguing that treaties are actions). But see Akehurst, supra note 7, at 3, 43
(criticizing D’Amato for inconsistency).
E.g., Ted Stein, Remarks [on customs and treaties], in CHANGE AND STABILITY, supra note 3, at 12, 13.
North Sea Continental Shelf, 1969 ICJ REP. at 44; RESTATEMENT, supra note 6, §102(2).
North Sea Continental Shelf, 1969 ICJ REP. at 44; Right of Passage over Indian Territory (Port. v. India),
Merits, 1960 ICJ REP. 6, 42–43 (Apr. 12); Asylum (Colom./Peru), 1950 ICJ REP. 266, 276–77 (Nov. 20); S.S. “Lotus”
(Fr. v. Turk.), 1927 PCIJ (ser. A) No. 10, at 28 (Sept. 7).
The Paquete Habana, 175 U.S. 677, 686 (1900); JAMES BRIERLY, THE LAW OF NATIONS 62 (6th ed. 1963).
Delimitation of the Maritime Boundary in the Gulf of Maine Area (Can. v. U.S.), 1984 ICJ REP. 246, 299 (Oct.
12); OSCAR SCHACHTER, INTERNATIONAL LAW IN THEORY AND PRACTICE 35–36 (1991); GEORG SCHWARZENBERGER,
THE INDUCTIVE APPROACH TO INTERNATIONAL LAW 33 (1965); Georg Schwarzenberger, The Inductive Approach to
International Law, 60 HARV. L. REV. 539, 566–70 (1947).
S.S. “Lotus,” 1927 PCIJ (ser. A) No. 10, at 18, 29; see also Nottebohm (Liech. v. Guat.), Second Phase, 1955 ICJ
REP. 4, 22 (Apr. 6); S.S. Wimbledon, 1923 PCIJ (ser. A) No. 1, at 25 (Aug. 17).
Hiram Chodosh, Neither Treaty nor Custom: The Emergence of Declarative International Law, 26 TEX. INT’L L.J. 87,
102 n.70 (1991).
Bruno Simma & Philip Alston, The Sources of Human Rights Law: Custom, Jus Cogens, and General Principles,
1988–89 AUSTL. Y.B. INT’L L. 82.
Bin Cheng, United Nations Resolutions on Outer Space: ‘Instant’ International Customary Law? 5 INDIAN J. INT’L L.
23 (1965), reprinted in INTERNATIONAL LAW: TEACHING AND PRACTICE 237 (Bin Cheng ed., 1982).
North Sea Continental Shelf, 1969 ICJ REP. at 44; Eduardo Jiménez de Aréchaga, Remarks [on general
principles and General Assembly resolutions], in CHANGE AND STABILITY, supra note 3, at 48.
Akehurst, supra note 7, at 6–7; Jonathan I. Charney, Universal International Law, 87 AJIL 529, 544–45 (1993).
Military and Paramilitary Activities in and Against Nicaragua (Nicar. v. U.S.), Merits, 1986 ICJ REP. 14 ( June
27) [hereinafter Nicaragua]; see also Western Sahara, Advisory Opinion, 1975 ICJ REP. 12, 30–37 (Oct. 16); Legal
Consequences for States of the Continued Presence of South Africa in Namibia (South West Africa) Notwith-
standing Security Council Resolution 276 (1970), Advisory Opinion, 1971 ICJ REP. 16, 31–32 ( June 21) [herein-
after Namibia Advisory Opinion].
E.g., Declaration on Principles of International Law Concerning Friendly Relations and Co-operation Among
States in Accordance with the Charter of the United Nations, GA Res. 2625, UN GAOR, 25th Sess., Supp. No. 28,
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 759
state practice, holding that it was sufficient for conduct to be generally consistent with state-
ments of rules, provided that instances of inconsistent practice had been treated as breaches
of the rule concerned rather than as generating a new rule.24
The tests and justifications for traditional and modern custom appear to differ because
the former develops slowly through state practice, while the latter can arise rapidly based
on opinio juris.25 This difference has spurred considerable discussion over two related issues.
First, the legitimacy of traditional and modern custom has been debated at length.26 David
Fidler characterizes the various approaches to this issue as the dinosaur, dynamo, and dan-
gerous perspectives.27 The dinosaur approach focuses on traditional custom and argues that
massive changes in the international system have rendered it an anachronism. For example,
Jonathan Charney claims that the increasing number and diversity of states, as well as the
emergence of global problems that are addressed in international fora, makes traditional
custom an inappropriate means for developing law.28 The dynamo perspective concentrates
on modern custom and embraces it as a progressive source of law that can respond to moral
issues and global challenges. For example, Theodor Meron, Richard Lillich, and Lori Bruun
argue that modern custom based on declarations by international fora provides an
important source of law for human rights obligations.29 Finally, the dangerous perspective
views modern custom as a departure from the traditional approach that has created an
opportunity for legal and political abuse. Thus, Michael Reisman characterizes the increased
dependence on custom as a “great leap backwards” designed to serve the interests of power-
ful states.30 Similarly, Arthur Weisburd holds that modern custom often lacks the legitimacy
of state consent because it is formed despite little, or conflicting, state practice.31
Second, the divergence between traditional and modern custom has been criticized as
undermining the integrity of custom as a source of law. Patrick Kelly argues that custom is
an indeterminate and malleable source of law, simply a “matter of taste.”32 According to
D’Amato, the modern approach trashes the theoretical foundations of custom by inverting
the traditional priority of state practice over opinio juris.33 Sir Robert Jennings insists that
“most of what we perversely persist in calling customary international law is not only not
customary law: it does not even faintly resemble a customary law.”34 The phrases “modern,”
at 121, UN Doc. A/8028 (1970); Conference on Security and Co-operation in Europe, Final Act, Aug. 1, 1975, 73
DEP’T ST. BULL. 323 (1975), reprinted in 14 ILM 1292 (1975) [hereinafter Helsinki Accord].
Nicaragua, 1986 ICJ REP. at 98, para. 186.
Georges Abi-Saab, Remarks [on custom and treaties], in CHANGE AND STABILITY, supra note 3, at 9; Louis
Henkin, Human Rights and State “Sovereignty,” 25 GA. J. INT’L L. 37 (1995/96); Simma & Alston, supra note 18, at
90; Ted Stein, The Approach of a Different Drummer: The Principle of the Persistent Objector in International Law, 26 HARV.
INT’L L.J. 457, 457 (1985).
Consider, for example, the strong and conflicting responses to the Nicaragua case. Symposium, Appraisals of
the ICJ’s Decision: Nicaragua v. United States (Merits), 81 AJIL 77 (1987).
David Fidler, Challenging the Classical Concept of Custom, 1996 GER. Y.B. INT’L L. 198, 216–31.
Charney, supra note 21, at 543.
MERON, supra note 2; Lori Bruun, Beyond the 1948 Convention—Emerging Principles of Genocide in Customary
International Law, 17 MD. J. INT’L L. & TRADE 193, 216–17 (1993); Richard B. Lillich, The Growing Importance of
Customary International Human Rights Law, 25 GA. J. INT’L & COMP. L. 1, 8 (1995/96).
Reisman, supra note 3, at 135.
Arthur A. Weisburd, Customary International Law: The Problem of Treaties, 21 VAND. J. TRANSNAT’L L. 1 (1988)
[hereinafter Weisburd, Customary IL]. For further discussion, see Anthony A. D’Amato, Custom and Treaty: A
Response to Professor Arthur A. Weisburd, 21 VAND. J. TRANSNAT’L L. 459 (1988) [hereinafter D’Amato, Response];
Arthur A. Weisburd, A Reply to Professor Anthony A. D’Amato, id. at 473; Anthony A. D’Amato, A Brief Rejoinder,
id. at 489.
Kelly, supra note 1, at 451.
Anthony A. D’Amato, Trashing Customary International Law, 81 AJIL 101 (1987).
Robert Y. Jennings, The Identification of International Law, in INTERNATIONAL LAW, supra note 19, at 3, 5; see also
North Sea Continental Shelf (FRG/Den.; FRG/Neth.), 1969 ICJ REP. 3, 224 (Feb. 20) (Sørensen, J. ad hoc,
dissenting); Simma & Alston, supra note 18, at 83.
760 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
“new,”35 “contemporary,”36 and “instant”37 custom appear inherently contradictory and ob-
scure the real basis for forming this law. Hilary Charlesworth contends that modern custom
can be rationalized only by dispensing with the traditional rhetoric of custom.38 Bruno
Simma and Philip Alston argue that the modern approach has created an “identity crisis”39
for custom and would be better understood as a general principle of international law.40
Likewise, Charney, Daniel Bodansky, and Hiram Chodosh conclude that modern custom
is really a new species of universal declaratory law because it is based on authoritative
statements about practice rather than observable regularities of behavior.41
Both the legitimacy and the integrity of traditional and modern custom have received
considerable attention and polarized positions are evident. However, few commentators
have transcended these debates by attempting to provide an overall theory of custom.
Frederic Kirgis rationalizes the divergence in custom by analyzing the requirements of state
practice and opinio juris on a sliding scale.42 At one end, highly consistent state practice can
establish a customary rule without requiring opinio juris. However, as the frequency and
consistency of state practice decline, a stronger showing of opinio juris will be required. Kirgis
argues that the exact trade-off between state practice and opinio juris will depend on the
importance of the activity in question and the reasonableness of the rule involved.43 Simma
and Alston claim that this approach reinterprets the concept of custom so as to produce the
“right” answers.44 However, John Tasioulas argues that the sliding scale can be rationalized
on the basis of Ronald Dworkin’s interpretive theory of law, which balances a description
of what the law has been with normative considerations about what the law should be. This
perspective shows why the Court may be less exacting in requiring state practice and opinio
juris in cases that deal with important moral issues.
This article builds on the work of Kirgis and Tasioulas and offers a defense of custom by
seeking to reconcile the traditional and modern approaches. Part II analyzes the com-
peting values of descriptive accuracy and normative appeal that are used to justify inter-
national law. These values characterize traditional and modern custom, respectively,
because of their inductive and deductive methodologies and facilitative and moral content.
Part III examines custom on a sliding scale and rejects this interpretive approach because
it does not accurately describe the process of finding custom and would create customs that
are apologies for power or utopian and unachievable.45 Part IV presents an alternative
vision of Dworkin’s interpretive theory of law applied to custom, which incorporates the
justifications of descriptive accuracy and normative appeal and seeks to balance them in a
Rawlsian reflective equilibrium.46 Part V outlines the advantages of the reflective interpre-
tive approach over the sliding-scale methodology. Rearticulating the theoretical founda-
Curtis A. Bradley & Jack L. Goldsmith, Customary International Law as Federal Common Law: A Critique of the
Modern Position, 110 HARV. L. REV. 815, 838 (1997); Kelly, supra note 1, at 454 n.20, 484.
Stein, supra note 11, at 12.
Bin Cheng, in INTERNATIONAL LAW, supra note 19, at 249.
Hilary C. M. Charlesworth, Customary International Law and the Nicaragua Case, 1984–87 AUSTL. Y.B. INT’L L. 1.
Simma & Alston, supra note 18, at 88, 96.
Id. at 102–06; see also ICJ STATUTE Art. 38(1)(c).
Daniel Bodansky, Customary (and Not So Customary) International Environmental Law, 3 IND. J. GLOBAL LEGAL
STUD. 105, 116–19 (1995); Charney, supra note 21, at 543, 546–47; Chodosh, supra note 17.
Frederic L. Kirgis, Jr., Custom on a Sliding Scale, 81 AJIL 146 (1987); John Tasioulas, In Defence of Relative
Normativity: Communitarian Values and the Nicaragua Case, 16 OXFORD J. LEGAL STUD. 85 (1996).
Kirgis, supra note 42, at 149.
Simma & Alston, supra note 18, at 83.
For an explanation of the phrases “apology for power” and “utopian and unachievable,” see MARTTI KOSKENNIEMI,
FROM APOLOGY TO UTOPIA: THE STRUCTURE OF INTERNATIONAL LEGAL ARGUMENT 2 (1989).
RONALD DWORKIN, LAW’S EMPIRE (1986); JOHN RAWLS, A THEORY OF JUSTICE 20 (1972).
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 761
tions of custom in this more principled and flexible fashion will provide international actors
with a coherent theory for applying custom. It will also help to justify the traditional and
modern approaches to custom as two aspects of a single source of law, rather than char-
acterize one approach as illegitimate and the choice between them as undermining the
integrity of custom.
II. THE DESCRIPTIVE AND NORMATIVE APPROACHES TO CUSTOM
Descriptive Accuracy and Normative Appeal
H. L. A. Hart and R. M. Hare distinguish between descriptive and prescriptive statements and
laws.47 Descriptive laws can be discovered by observation and reasoning because they are
statements about what the practice has been. By contrast, prescriptive laws are not deter-
mined primarily by observations of fact because they state demands about what the practice
should or ought to be. Legal rules are always prescriptive because they make demands about how
people and states should behave. However, their prescriptive nature can be justified by what
the practice has been and/or what the practice should be. A law is primarily descriptive if it
conforms to the premise: the law is what the practice has been. A law is primarily normative
if it is formulated on the assumption: the law is what the practice ought to be. What the law
is (prescription) can be justified by what the practice has been (description) or what the prac-
tice ought to be (normativity). Thus, we should distinguish between what the practice has
been, what the law is, and what the practice ought to be: “has/is/ought” (description/
prescription/normativity). These distinctions, however, are often confused in two ways.
First, some theorists mistakenly amalgamate descriptive and prescriptive considerations.
By aligning the distinction between description and normativity with the positivist “is/
ought” dichotomy, which distinguishes between what the law is and what the law should be,
they conflate the “has/is/ought” structure with the “is/ought” dichotomy. This is misguided
because it merges descriptive and prescriptive elements by fusing considerations of what the
practice has been with considerations of what the law is. Moving from has to is involves some
level of law creation because it requires the formulation of an abstract rule from actual prac-
tice, despite the existence of silences, ambiguities, and contradictions in that practice.48
Determining what the law is from what the practice has been relies heavily on the choice of
characteristics under which precedents are classified and the degree of abstraction em-
ployed.49 Description and prescription are not coextensive because one reflects practice
(description) and the other directs practice (prescription).
Second, prescriptive and normative rules are often confused because both express an
imperative to act. Prescriptive laws express a legal imperative to act (you should do x because
x is legally required), while normative rules express a moral imperative to act (you should do
x because x is morally required). Many laws embody both a legal and a moral imperative, for
example, the prohibition of murder. However, the two considerations are distinct because
laws may be prescriptive but not normative; thus, traffic regulations are legally required but
often morally indifferent. Likewise, a rule may be normative but not prescriptive; for example,
adultery may be considered morally wrong, but it is not illegal in Western societies. Hence,
H. L. A. HART, THE CONCEPT OF LAW 183 (1961); R. M. HARE, MORAL THINKING, ch. 4 (1981); R. M. HARE, THE
LANGUAGE OF MORALS (1952).
The same issue arises with codification of international law. CHARLES DE VISSCHER, THEORY AND REALITY IN
PUBLIC INTERNATIONAL LAW 39 (P. E. Corbett trans., rev. ed. 1968); Hersch Lauterpacht, Codification and
Development of International Law, 49 AJIL 16 (1955); Oscar Schachter, Recent Trends in International Law Making,
1988–89 AUSTL. Y.B. INT’L L. 1, 2–3; Julius Stone, The Vocation of the International Law Commission, 57 COLUM. L. REV.
16, 18–19 (1957).
Ulrich Fastenrath, Relative Normativity in International Law, 4 EUR. J. INT’L L. 305, 316–19 (1993).
762 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
what the law is (prescription) can be justified by what the practice should be (normativity),
but the two considerations are not always coextensive.
The values of descriptive accuracy and normative appeal provide important, and some-
times competing, justifications for international law. Descriptive accuracy (which focuses on
what the practice has been) is valuable in justifying the content of international law because
laws should correspond to reality.50 Laws must bear some relation to practice if they are to
regulate conduct effectively, because laws that set unrealistic standards are likely to be dis-
obeyed and ultimately forgotten. This consideration particularly applies to decentralized
systems of law, such as international law, where traditional enforcement mechanisms are
unavailable or underdeveloped.51 Descriptive accuracy is also essential to predictive power52
because a theory that accurately describes practice enables more reliable predictions of
future state behavior.53
The alternative justification for international law provided by normative appeal (which
focuses on what the practice ought to be) involves procedural and substantive aspects. First,
procedural normativity requires that the process for forming laws be transparent, so that
states are aware of the real basis for forming customs and can regulate their actions accord-
ingly.54 It also entails an opportunity for states to participate in law formation and have their
positions considered.55 Legal rules are more likely to engender respect in a decentralized
system, possibly even when the outcome is less favorable, if they result from a process per-
ceived as legitimate.56 Second, substantive normativity requires that laws be coherent and
that their content be morally good or at least neutral, depending on their subject matter.
Claims about “morality” are contentious because it remains unclear whether morality is
objective or culturally relative. By morality, I am referring to commonly held subjective
values about actions that are right and wrong, which a representative majority of states has
recognized in treaties and declarations.57
The justifications for traditional and modern custom align with description and norma-
tivity, respectively, in keeping with the methodologies that characterize each approach.
Descriptive accuracy and normative appeal respectively represent inductive and deductive
methods of justification.58 Martti Koskenniemi argues that the inductive pattern privileges
description over normativity, while the deductive pattern does the reverse.59
Traditional custom is closely associated with descriptive accuracy because norms are con-
structed primarily from state practice—working from practice to theory.60 Reliance on state
practice provides continuity with past actions and reliable predictions of future actions. It
Bin Cheng, Custom: The Future of General State Practice in a Divided World, in THE STRUCTURE AND PROCESS OF
INTERNATIONAL LAW: ESSAYS IN LEGAL PHILOSOPHY, DOCTRINE, AND THEORY 513, 539 (Ronald St. J. Macdonald &
Douglas M. Johnston eds., 1983).
Bodansky, supra note 41, at 116–19.
Jerome Frank, Law and the Modern Mind, in LLOYD’S INTRODUCTION TO JURISPRUDENCE 679 (Michael Freeman
ed., 6th ed. 1994); Oliver Wendell Holmes, The Path of Law, 10 HARV. L. REV. 457, 457 (1896–97); Karl Llewellyn,
Some Realism About Realism, 44 HARV. L. REV. 1222 (1931).
Simma & Alston, supra note 18, at 89.
Bin Cheng, supra note 50, at 539; Charlesworth, supra note 38, at 27.
THOMAS R. TYLER, WHY PEOPLE OBEY THE LAW 170–73 (1990); Kelly, supra note 1, at 517–18, 530–31; Stephen
Machura, Introduction: Procedural Justice, Law and Policy, 19 MICH. J. INT’L L. 345, 353–54 (1998).
TYLER, supra note 55, at 163, 170–73, 175–78; Thomas M. Franck, Legitimacy in the International System, 82 AJIL
705, 706 (1988); Kelly, supra note 1, at 531; Machura, supra note 55, at 353–54.
These values are intersubjective, i.e., values commonly held by a group of subjects, in this case states. Jan
Narveson, Inter-subjective, in THE OXFORD COMPANION TO PHILOSOPHY 414 (Ted Honderich ed., 1995) [hereinafter
See also ascending and descending methods of justification. WALTER ULLMANN, LAW AND POLITICS IN THE
MIDDLE AGES 30–31 (1975).
KOSKENNIEMI, supra note 45, at 41.
North Sea Continental Shelf (FRG/Den.; FRG/Neth.), 1969 ICJ REP. 3 (Feb. 20); RESTATEMENT, supra note
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 763
results in practical and achievable customs that can actually regulate state conduct. By con-
trast, modern custom demonstrates a predilection for substantive normativity rather than
descriptive accuracy. Modern custom derives norms primarily from abstract statements of
opinio juris—working from theory to practice. Whereas state practice is clearly descriptive,
opinio juris is inherently ambiguous in nature because statements can represent lex lata (what
the law is, a descriptive characteristic) or lex ferenda (what the law should be, a normative
characteristic). The Court has held that only statements of lex lata can contribute to the
formation of custom.61 However, modern custom seems to be based on normative state-
ments of lex ferenda cloaked as lex lata, for three reasons.
First, attempts to distinguish lex lata and lex ferenda often rely on differentiating between
codification and progressive development of the law.62 Codification means formulating and
systematizing rules that already exist in the form of practice accepted as law (lex lata), while
progressive development is an overtly political act concerned with formulations of what the
law should be (lex ferenda).63 However, even though codification is meant to be scientific, not
political,64 formulating a general rule from actual practice necessarily entails some level of
law creation. Codification involves legal judgments because it seeks to form a coherent rule
in the face of inconsistent practice, which requires some crystallization of lex ferenda.
Second, it is difficult to distinguish between codification and progressive development by
examining the language of treaties and declarations and their travaux préparatoires.65 For
example, treaties and resolutions might form customs when expressed in obligatory lan-
guage, such as “must,” rather than aspirational terms, such as “should.”66 However, the
Court has relied on resolutions couched in both mandatory and nonmandatory language.67
Further, mandatory language is still ambiguous because it can represent a moral, treaty, or
customary obligation. Thus, whether a declaration or treaty provides a statement of what the
customary law is or should be cannot be determined by reference to mandatory or per-
missive words alone.
Finally, treaties and resolutions often use mandatory language to prescribe a model of
conduct and provide a catalyst for the development of modern custom.68 Treaties and dec-
larations do not merely photograph or declare the current state of practice on moral issues.
Rather, they often reflect a deliberate ambiguity between actual and desired practice, de-
signed to develop the law and to stretch the consensus on the text as far as possible.69 For
example, some rights set out in the Universal Declaration of Human Rights of 1948 are
expressed in mandatory terms and have achieved customary status even though infringe-
ments are “widespread, often gross and generally tolerated by the international com-
munity.”70 As a result, modern custom often represents progressive development of the law
masked as codification by phrasing lex ferenda as lex lata.71
North Sea Continental Shelf, 1969 ICJ REP. at 3, 38.
MARK VILLIGER, CUSTOMARY INTERNATIONAL LAW AND TREATIES: A MANUAL ON THE THEORY AND PRACTICE OF
THE INTERRELATION OF SOURCES 102 (2d ed. 1997); R. R. Baxter, Multilateral Treaties as Evidence of Customary Inter-
national Law, 1965–66 BRIT. Y.B. INT’L L. 275, 286.
ICJ STATUTE Art. 15.
Schachter, supra note 48, at 2.
Akehurst, supra note 7, at 45–47.
For example, the range of terms in the Manila Declaration on the Peaceful Settlement of International Dis-
putes, GA Res. 37/10 (Nov. 14, 1982). See also Luigi Condorelli, Remarks [on lex lata and lex ferenda], in CHANGE
AND STABILITY, supra note 3, at 79.
E.g., Nicaragua, 1986 ICJ REP. at 107, paras. 203–04 (citing Declaration on Principles of International Law Con-
cerning Friendly Relations and Co-operation Among States in Accordance with the Charter of the United Nations,
supra note 23; Helsinki Accord, supra note 23).
Jiménez de Aréchaga, supra note 20; Alain Pellet, The Normative Dilemma: Will and Consent in International Law-
Making, 1988–89 AUSTL. Y.B. INT’L L. 22, 35.
MERON, supra note 2, at 43–44.
SCHACHTER, supra note 15, at 335.
MERON, supra note 2, at 42.
764 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
Thus, state practice is descriptive, while opinio juris can be descriptive or normative. Tradi-
tional custom and modern custom are primarily justified by descriptive accuracy and
normative appeal, respectively, because of their emphasis on state practice and normative
Facilitative and Moral Customs
The best balance between the justifications of descriptive accuracy and normative appeal
depends on the facilitative or moral content of the custom involved. International laws are
ranged on a spectrum between facilitative and moral rules.72 At one extreme, there are
completely facilitative rules, which promote coexistence and cooperation but do not deal
with substantive moral issues (such as that ships must pass on the left). Next come primarily
facilitative rules that regulate interaction but also give rise to some moral considerations
(such as the fair distribution of resources in a continental shelf). Moving to the middle, one
finds rules that involve important facilitative and moral considerations (such as environmen-
tal rules). Toward the other extreme, the laws are primarily moral rather than facilitative
(such as some human rights obligations) and peremptory rules that prohibit actions whether
or not they affect coexistence and cooperation (such as jus cogens laws prohibiting geno-
cide). While it is possible for a law to facilitate interaction between states and also have a
strong moral content (such as the prohibition on the use of force), laws often tend more
toward one or the other end of the spectrum.
Facilitative customs are more descriptive than normative because they turn a description
of actual practice into a prescriptive requirement for future action.73 Moral customs are more
normative than descriptive because they prescribe future action based on normative evalu-
ations of ideal practice. Traditional customs primarily facilitate coexistence and cooperation
between sovereign states without having a peremptory moral character; for example, the law
on diplomatic immunity.74 State practice is dominant in establishing traditional facilitative
customs because these rules turn empirical descriptions of past practice into prescriptive
requirements for future practice.75 However, international law has expanded since 1945 to
include many moral 76 issues such as human rights,77 the use of force,78 and environmental
protection.79 Louis Henkin characterizes this development as a move from state values to
human values, and from a liberal state system to a welfare system.80 State practice is less
important in forming modern customs because these customs prescribe ideal standards of
conduct rather than describe existing practice.81 For example, the customary prohibition
on torture expresses a moral abhorrence of torture rather than an accurate description of
HART, supra note 47, at 225.
HANS KELSEN, PRINCIPLES OF INTERNATIONAL LAW 307–08, 418 (1952).
Kelly, supra note 1, at 479–80.
Simma & Alston, supra note 18, at 89.
By morality, I am again referring to commonly held subjective values about actions that are right and wrong,
and have been recognized by a representative majority of states in treaties and declarations. While these values
are often called “normative,” I call them “moral” so as not to confuse the distinction between descriptive/
normative justifications and facilitative/moral content. I use the term “normative” to describe a statement about
what the practice should be, rather than to reflect the inherent moral content of a rule.
Bruun, supra note 29, at 216–17; Lillich, supra note 29, at 8.
E.g., Nicaragua, 1986 ICJ REP. 14.
Bodansky, supra note 41.
Henkin, supra note 25, at 34–35.
SCHACHTER, supra note 15, at 11; FERNANDO R. TESÓN, HUMANITARIAN INTERVENTION: AN INQUIRY INTO LAW
AND MORALITY 14 (1988); Theodor Meron, On a Hierarchy of International Human Rights, 80 AJIL 1, 19–20 (1986);
Oscar Schachter, Entangled Treaty and Custom, in INTERNATIONAL LAW AT A TIME OF PERPLEXITY 717, 733–34 (Yoram
Dinstein ed., 1989).
RESTATEMENT, supra note 6, §702.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 765
The moral content of modern custom explains the strong tendency to discount the impor-
tance of contrary state practice in the modern approach. Irregularities in description can
undermine a descriptive law, but a normative law may be broken and remain a law because
it is not premised on descriptive accuracy. For example, jus cogens norms prohibit funda-
mentally immoral conduct and cannot be undermined by treaty arrangement or incon-
sistent state practice.83 Since the subject matter of modern customs is not morally neutral,
the international community is not willing to accept any norm established by state practice.84
Modern custom involves an almost teleological approach, whereby some examples of state
practice are used to justify a chosen norm, rather than deriving norms from state practice.
As noted above, this approach was evident in the Nicaragua case, where the Court held it
sufficient for the conduct of states to be generally consistent with statements of rules, pro-
vided that contrary state practice had generally been “treated as breaches of that rule, not
as indications of the recognition of a new rule.”85 Thus, the importance of descriptive
accuracy varies according to the facilitative or moral content of the rule involved.
The relative importance of procedural and substantive normativity also varies between
facilitative and moral customs. These two types of normativity are reminiscent of Dworkin’s
distinction between fair institutions (procedural normativity) and just outcomes (substantive
normativity).86 The two concepts are related in international law because unfair institutions
that concentrate power in the hands of a few states are less likely to produce just outcomes
as judged by the majority of states. They are also independently important because inade-
quacies in either procedural or substantive normativity will diminish respect for asserted
customs and reduce their ability to effect compliance. However, their relative importance
varies in accordance with the customs concerned. As primarily facilitative rules do not deal
with substantive moral issues, it suffices for them to be procedurally normative and sub-
stantively neutral. For example, traffic regulations should be enacted by a fair process but
do not need to be substantively normative because they generally facilitate interaction rather
than deal with substantive moral issues. Ideally, rules that deal with moral issues should be
procedurally and substantively normative, but the international community may allow the
substantive morality of some customs to outweigh defects in their process. For example, the
substantive morality of jus cogens norms outweighs the procedural right of sovereign states
to become persistent objectors to them.87
The relative importance of substantive and procedural normativity links to criticisms of
modern custom as creating quasi legislation on the basis of declarations and treaties.88 Dec-
larations of international fora are formally nonbinding, while states can determine whether,
and to what extent, they wish to be bound by treaties.89 By contrast, custom is generally
binding except for the limited and contentious persistent objector rule.90 Transforming
Vienna Convention on the Law of Treaties, opened for signature May 23, 1969, Art. 53, 1155 UNTS 331; see also
RESTATEMENT, supra note 6, §102 cmt. k; DAVID HARRIS, CASES AND MATERIALS ON INTERNATIONAL LAW 42, 835 (5th
ed. 1998); H. W. A. THIRLWAY, INTERNATIONAL CUSTOMARY LAW AND CODIFICATION 110 (1972); Charlesworth, supra
note 38, at 4; cf. ANTONIO CASSESE, INTERNATIONAL LAW IN A DIVIDED WORLD 179 (1989).
SCHACHTER, supra note 15, at 11; see also Schachter, supra note 81, at 733–34.
Nicaragua, 1986 ICJ REP. at 98, para. 186.
DWORKIN, supra note 46, at 177.
RESTATEMENT, supra note 6, §102 cmt. k; THIRLWAY, supra note 83, at 110; Charlesworth, supra note 38, at 4;
cf. CASSESE, supra note 83, at 179.
Henkin, supra note 25, at 37; Fred L. Morrison, Legal Issues in the Nicaragua Opinion, 81 AJIL 160, 162 (1987).
Vienna Convention on the Law of Treaties, supra note 83, Art. 26; Schachter, supra note 81, at 727–28.
According to the persistent objector rule, states that have persistently objected during the emergence of a
custom are not bound by it. Gerald Fitzmaurice, The General Principles of International Law Considered from the
Standpoint of the Rules of Law, 92 RECUEIL DES COURS 1, 49–50 (1957 II); Stein, supra note 25, at 457. However, the
International Court of Justice has endorsed the persistent objector rule only twice, and arguably both times in
obiter dicta. Fisheries case (UK v. Nor.), 1951 ICJ REP.116, 131 (Dec. 18); Asylum (Colom./Peru), 1950 ICJ REP. 266,
277–78 (Nov. 20); see also Nuclear Tests (Austl. v. Fr.), 1974 ICJ REP. 253, 286–93 (Dec. 20) (Gross, J., sep. op.).
Further, a state cannot be a persistent objector to jus cogens rules and theorists have generally concluded that the
766 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
declarations and treaties into custom changes their nature because customs can bind non-
parties to treaties and declarations91 and are not affected by reservations92 or denunciation.93
Thus, Prosper Weil argues that the requirements of treaties have “not been frontally
assaulted but cunningly outflanked.”94 However, the emphasis on community consensus over
individual state consent in modern custom reflects the priority of substantive normativity
over procedural normativity in important moral issues. Modern custom evinces a desire to
create general international laws that can bind all states on important moral issues.95 Accord-
ing to Kirgis, “The alternative would be an international legal order containing ominous
silences—where treaty commitments cannot be found—concerning the ways in which states
impose their wills on other states or on individuals.”96 The international community dis-
counts the importance of dissenting states and contrary state practice because it is not
prepared to recognize exceptions to the maintenance of certain fundamental values. Rec-
ognizing exceptions to such rules would “shock the conscience of mankind”97 and be con-
trary to “elementary considerations of humanity.”98 The substantive normativity of modern
custom can therefore be used to justify a reduced focus on procedural normativity and
The importance of descriptive accuracy and normative appeal, and procedural and sub-
stantive normativity, varies according to the facilitative and moral content of traditional and
modern custom. The reduced focus on state practice in the modern approach is explained
by its use to create generally binding laws on important moral issues.
Apology and Utopia
Descriptive accuracy and normative appeal provide important bases for justifying inter-
national law. While facilitative and moral rules respectively tend more toward descriptive
accuracy and normative appeal, neither value is sufficient to justify all international laws.
According to Koskenniemi, international law mediates between the competing tendencies
of apology (description) and utopia (normativity).
A law which would lack distance from State behaviour, will or interest would amount
to a non-normative apology, a mere sociological description. A law which would base
itself on principles which are unrelated to State behaviour, will or interest would seem
utopian, incapable of demonstrating its own content in any reliable way.99
The conflicting values of description and normativity, and their respective risks of being
an apology for power or utopian and unachievable, represent the fundamental tension in
legal argument.100 Theorists oscillate between these two extremes and each position remains
open to challenge from the opposite perspective.101 Thus, international legal argument is
practical application of the rule is limited. BYERS, supra note 6, at 181; D’AMATO, supra note 6, at 187–99, 233–63;
Jonathan I. Charney, The Persistent Objector Rule and the Development of Customary International Law, 1986 BRIT. Y.B.
INT’L L. 1, 11–16.
It may also discourage states from voting for aspirational instruments. Thomas M. Franck, Some Observations
on the ICJ’s Procedural and Substantive Innovations, 81 AJIL 116, 119 (1987).
Nicaragua, 1986 ICJ REP. at 113–14, paras. 217–18; MERON, supra note 2, at 6–7, 27; Pierre Imbert, Reservations
and Human Rights Conventions, 6 HUM. RTS. REV. 28 (1981); Schachter, supra note 81, at 727–28.
Vienna Convention on the Law of Treaties, supra note 83, Art. 43.
Prosper Weil, Towards Relative Normativity in International Law? 77 AJIL 413, 438 (1983).
CASSESE, supra note 83, at 31, 110, 398; Charlesworth, supra note 38, at 1–3; Tasioulas, supra note 42, at 116–17.
Kirgis, supra note 42, at 148.
Reservations to the Convention on the Prevention and Punishment of the Crime of Genocide, Advisory
Opinion, 1951 ICJ REP. 15, 23 (May 28).
Corfu Channel case (UK v. Alb.), Merits, 1949 ICJ REP. 4, 22 (Apr. 9).
KOSKENNIEMI, supra note 45, at 2.
BYERS, supra note 6, at 49.
KOSKENNIEMI, supra note 45, at 42.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 767
always dynamic. David Kennedy argues that “[e]ither international law has been too far
from politics and must move closer to become effective, or it has become dangerously inter-
mingled with politics and must assert its autonomy to remain potent.”102 The dynamic be-
tween description and normativity represents the “deep doctrinal schizophrenia” of law,103
or a “disciplinary hamster wheel” from which theorists appear unable to escape.104
While the justifications for traditional and modern custom mainly align with descriptive
accuracy and normative appeal, neither approach is completely descriptive or normative
because both recognize the importance of state practice and opinio juris to varying degrees.
However, as the two approaches to custom tend toward opposite ends of the theoretical
spectrum, the strongest criticisms of each come from the counterstandpoint. Traditional
custom embodies the value of descriptive accuracy but the risk of apologism, so that one of
the main criticisms of traditional custom is that it lacks democratic legitimacy (a normative
criticism). By contrast, modern custom is normatively appealing but risks creating utopian
rules, so it is criticized for producing norms that are divorced from reality (a descriptive
criticism). The critiques of traditional and modern custom rest on this basis.
A critique of traditional custom. Traditional custom lacks procedural normativity. The process
of custom formation is inherently uncertain, with no clear guide to the amount, duration,
frequency, and continuity of state practice required to form a custom.105 The unwritten
nature of traditional custom makes its content inherently insecure, while requiring repeated
practice is “too clumsy and slow”106 to accommodate the fast-paced evolution of law.107
Traditional custom is meant to be based on general and consistent state practice, but selec-
tive analysis inheres in this approach because of the impossibility of thoroughly analyzing
the practice of almost two hundred states.108 This selectivity results in a “democratic defi-
cit”109 because most customs are found to exist on the basis of practice by fewer than a dozen
states.110 For example, the finding of a custom of neutrality of international canals in the
Wimbledon case was based on the Suez and Panama Canal Treaties alone.111 If norms are
based primarily on actions,112 then only states with the ability to act can form and reject
customs.113 States might participate directly by making statements114 or indirectly if inaction
amounts to acquiescence.115 However, all states must then have perfect knowledge of state
practice, unlimited resources to respond, and the awareness that failure to respond will have
legal consequences, all of which is patently not so.116 Thus, a majority of states rarely par-
ticipate in the creation of customs that limit their sovereignty.117
David Kennedy, When Renewal Repeats: Thinking Against the Box, 32 N.Y.U. J. INT’L L. & POL. 335, 355 (2000).
DWORKIN, supra note 46, at 271.
Kennedy, supra note 102, at 407.
BYERS, supra note 6, at 156–62; D’AMATO, supra note 6, at 56–66.
WOLFGANG FRIEDMANN, THE CHANGING STRUCTURE OF INTERNATIONAL LAW 122 (1964).
Fidler, supra note 27, at 216.
Charney, supra note 21, at 537; Fidler, supra note 27, at 203, 217.
Kelly, supra note 1, at 519.
Charney, supra note 21, at 537; Chodosh, supra note 17, at 102; Oscar Schachter, New Custom: Power, Opinio
Juris and Contrary Practice, in THEORY OF INTERNATIONAL LAW AT THE THRESHOLD OF THE 21ST CENTURY: ESSAYS IN
HONOUR OF KRZYSZTOF SKUBISZEWSKI 531, 536 ( Jerzy Makarczyk ed., 1996); Weisburd, supra note 31, at 6.
S.S. Wimbledon, 1923 PCIJ (ser. A) No. 1, at 25 (Aug. 17); see also Chodosh, supra note 17, at 102 n.69; Kelly,
supra note 1, at 470 n.93.
E.g., D’AMATO, supra note 6, at 74–87; W. Michael Reisman, International Incidents: Introduction to a New Genre
in the Study of International Law, 10 YALE J. INT’L L. 1, 1–4 (1984).
BYERS, supra note 6, at 37.
D’AMATO, supra note 6, at 96–97.
KAROL WOLFKE, CUSTOM IN PRESENT INTERNATIONAL LAW 135 (2d ed. 1993); Akehurst, supra note 7, at 37;
I. C. MacGibbon, The Scope of Acquiescence in International Law, 1954 BRIT. Y.B. INT’L L. 143, 145–46 [hereinafter
MacGibbon, Acquiescence I ]; I. C. MacGibbon, Customary International Law and Acquiescence, 1957 BRIT. Y.B. INT’L
L. 115, 138 [hereinafter MacGibbon, Acquiescence II ].
Kelly, supra note 1, at 522.
Id. at 519.
768 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
These procedural inadequacies have also led to problems regarding the substantive
normativity of traditional custom. New, developing, and socialist states have objected to
customs as having been created by wealthy European and imperialist powers.118 Instead of
being apolitical, traditional custom is arguably hegemonic, ideologically biased, and a legiti-
mating force for the political and economic status quo. For example, new states are bound
by existing customs even though they did not participate in their formation.119 The legal
fiction of free and equal states also masks the reality of extreme variations of de facto power.120
Powerful states wield disproportionate and often decisive influence in determining the
content and application of custom.121 Further, courts tend to focus primarily on Western
state practice.122 For example, the Permanent Court cited decisive precedents in the Lotus
case based on actions by only six Western states.123 Recently, some international lawyers have
predicted that NATO’s intervention in Kosovo will establish a right to unilateral humanitar-
ian intervention, despite protests by China, India, and the Russian Federation.124 Procedural
inadequacies also give rise to substantive unfairness. For example, Mohammed Bedjaoui
argues that the freedom of the high seas was developed to meet the needs of wealthy states
with large fleets rather than the interests of states whose shores were approached.125 Its
undemocratic nature allows traditional custom to become an apology for exercises of power
by strong Western nations.126 The resulting sense of injustice diminishes respect for tradi-
tional custom, as well as its pull toward compliance.
A critique of modern custom. Modern custom demonstrates strengths and weaknesses in
procedural normativity. Reliance on statements accelerates the formation of custom,127 which
allows it to regulate global, as opposed to regional or national, issues in a timely manner.128
Deriving customs primarily from treaties and declarations, rather than state practice, is
potentially more democratic because it involves practically all states. Most states can parti-
cipate in the negotiation and ratification of treaties and declarations of international fora,
such as the United Nations General Assembly. The notion of sovereign equality (one state,
one vote) helps to level the playing field between developed and developing countries.
While formal equality cannot remedy all inequalities in power, international fora provide
less powerful states with a cost-efficient means of expressing their views.129 However, allowing
a majority of states to bind a minority does not sit well with the notion of state sovereignty.
Further, the precise criteria for forming modern custom have not been clearly articulated
in cases such as Nicaragua, where the Court still claimed to be applying the traditional test.130
It is also not clear that states know that treaties will become customary law, or that they wish
RESTATEMENT, supra note 6, §102(2); LOUIS HENKIN, HOW NATIONS BEHAVE 121 (2d ed. 1979); Fidler, supra
note 27, at 213–14, 218; B. Graefrath, Remarks [on custom and treaties], in CHANGE AND STABILITY, supra note 3,
at 19, 19–20; Schachter, supra note 81, at 721.
D’AMATO, supra note 6, at 191–93; Akehurst, supra note 7, at 27.
Tasioulas, supra note 42, at 123.
MOHAMMED BEDJAOUI, TOWARD A NEW INTERNATIONAL ECONOMIC ORDER 51–52 (1979); DE VISSCHER, supra
note 48, at 149; SCHACHTER, supra note 15, at 6.
Kelly, supra note 1, at 469.
S.S. “Lotus” (Fr. v. Turk.), 1927 PCIJ (ser. A) No. 10, at 29 (Sept. 7); see also Chodosh, supra note 17, at 102
n.70; Weisburd, Customary IL, supra note 31, at 6.
E.g., Antonio Cassese, Ex iniuria ius oritur: Are We Moving Towards International Legitimation of Forcible Humani-
tarian Countermeasures in the World Community? 10 EUR. J. INT’L L. 23 (1999); Jonathan I. Charney, Anticipatory
Humanitarian Intervention in Kosovo, 93 AJIL 834, 835–37 (1999).
BEDJAOUI, supra note 121, at 136.
KOSKENNIEMI, supra note 45, at 355.
Bradley & Goldsmith, supra note 35, at 840 n.167.
Charney, supra note 21, at 543; cf. Right of Passage over Indian Territory (Port. v. India), Merits, 1960 ICJ REP.
6, 39 (Apr. 12) (on regional customs).
Right of Passage over Indian Territory, 1960 ICJ REP. at 37.
Charlesworth, supra note 38, at 27.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 769
them to,131 though this may be changing. Likewise, votes in the General Assembly usually
receive little media scrutiny and are generally not intended to make law.132 For example, the
General Assembly resolution on torture133 was adopted unanimously, while a much smaller
number of states ratified the Convention Against Torture134 and others entered significant
reservations to it.135
On the substantive normative level, modern custom provides coherent statements of law
on moral issues such as human rights, but it has been criticized for “normative chauvin-
ism.”136 For example, some decry human rights obligations as reflecting a Western tendency
to give primacy to individual rights over communal or group needs.137 Similarly, the conclu-
sions of the Restatement (Third) of the Foreign Relations Law of the United States about customary
human rights correlate with civil and political rights enshrined in the U.S. Bill of Rights,138
but omit economic, social, and cultural rights not protected by those provisions.139 Treaties
and international fora are harder to charge with Western ideological hegemony because
developing nations have been in the majority since the 1950s.140 However, as resolutions and
treaties do not automatically become customary law,141 the practice of powerful states argu-
ably determines which ones are so transformed.142
The greatest criticism of modern custom is that it is descriptively inaccurate because it
reflects ideal, rather than actual, standards of conduct.143 The normative nature of modern
custom leads to an enormous gap between asserted customs and state practice. For example,
customary international law prohibits torture, yet torture is endemic.144 A similar criticism
is made of the “emptiness” of jus cogens norms, which are often flouted in practice.145 These
laws lack efficacy because states have not internalized them as standards of behavior to guide
their actions and judge the behavior of others.146 The regulatory function of modern custom
is doubtful because it appears merely to set up aspirational aims rather than realistic require-
ments about action. For example, while many Asian and African states have accepted human
rights treaty obligations, the Tunis declaration by forty-two African states in 1992 and the
Bangkok declaration by thirty-four Asian states in 1993 suggest that the majority of non-
Western states view asserted human rights customs as aspirational goals rather than enforce-
Franck, supra note 91, at 119; I. C. MacGibbon, Means for the Identification of International Law, in INTER-
NATIONAL LAW, supra note 19, at 10, 13; Pellet, supra note 68, at 31.
Dean Rusk, A Comment on Filartiga v. Pena-Irala, 11 GA. J. INT’L & COMP. L. 311 (1981); Phillip R. Trimble,
A Revisionist View of Customary International Law, 33 UCLA L. REV. 655, 680 n.60 (1986).
Resolution Against Torture and Other Cruel, Inhuman, or Degrading Treatment or Punishment, GA Res.
46, UN GAOR, 39th Sess., Supp. No. 51, at 197, UN Doc. A/39/51 (1984).
Convention Against Torture and Other Cruel, Inhuman, or Degrading Treatment or Punishment, Dec. 10,
1984, 1465 UNTS 85, 23 ILM 1027 (1984), as modified 24 ILM 535 (1985).
For example, the United States. U.S. Reservations, Declarations, and Understandings, Convention Against
Torture and Other Cruel, Inhuman, or Degrading Treatment or Punishment, 136 CONG. REC. S17,486 (daily ed.
Oct. 27, 1990).
Simma & Alston, supra note 18, at 94.
Rhoda E. Howard, Dignity, Community, and Human Rights, in HUMAN RIGHTS IN A CROSS-CULTURAL PERSPEC-
TIVE 81 (Abdullahi Ahmed An-Na’im ed., 1992); see also Kelly, supra note 1, at 466, 491; cf. Lawrence Friedman,
Borders: On the Emerging Sociology of Transnational Law, 32 STAN. J. INT’L L. 65, 83–85 (1996).
RESTATEMENT, supra note 6, §702.
Simma & Alston, supra note 18, at 94–95.
BEDJAOUI, supra note 121, at 141.
See pt. I supra; see also Charney, supra note 110, at 547; cf. D’AMATO, supra note 6, at 104, 110, 164.
Kelly, supra note 1, at 520–21.
Bodansky, supra note 41, at 110–11; James S. Watson, Legal Theory, Efficacy and Validity in the Development of
Human Rights Norms in International Law, U. ILL. L.F. 609, 610–13, 641 (1979).
RESTATEMENT, supra note 6, §702; Fidler, supra note 27, at 227 n.129; Rosalyn Higgins, Problems and Process:
International Law and How We Use It, in INTERNATIONAL HUMAN RIGHTS IN CONTEXT 232 (Henry J. Steiner & Philip
Alston eds., 2d ed. 2000).
Arthur A. Weisburd, The Emptiness of the Concept of Jus Cogens, as Illustrated by the War in Bosnia-Herzegovina,
17 MICH. J. INT’L L. 49 (1995).
HART, supra note 47, at 86–91.
770 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
able obligations.147 Some theorists characterize modern customs as “soft laws” or sublegal
obligations that do not amount to law.148 Indeed, norms that are honored in the breach do
not yield reliable predictions of future conduct and are likely to bring themselves, and
possibly custom as a whole, into disrepute.149
The dynamic relationship between apology and utopia. The advantages and disadvantages of
traditional and modern custom are “dynamic” because criticisms of traditional custom tend
to correspond to advantages of modern custom and vice versa. An improvement in norma-
tive appeal will often correspond to a decline in descriptive accuracy, while an enhancement
of descriptive accuracy will often result in a deterioration of normative appeal. Whereas
some rules are descriptively accurate and normatively appealing, the two justifications often
have a relationship of inverse proportionality. For example, relying on declarations by inter-
national fora is potentially more democratic than relying on the actions of powerful states
(a normative improvement in modern custom), but since these declarations are supported
by a majority of states rather than a predominance of power, they are less likely to be enforced
in practice (a descriptive deterioration from traditional custom).150 Similarly, the relation-
ship between procedural and substantive normativity is also dynamic. For example, one
could bind all states on important moral issues (a substantive normative advantage), but
doing so would bind some states without their consent (a procedural normative problem).151
Thus, the values of description and normativity represent the fundamental tension in legal
argument152 and both approaches remain open to challenge from the opposite perspective.153
As seen, international law reveals a dynamic tension between descriptive accuracy and
normative appeal. Traditional custom has descriptive strengths but lacks democratic legiti-
macy, while modern custom deals with substantive moral issues but produces laws that do
not reflect reality. This tension leads to the issue of whether the two approaches can be
reconciled in a single interpretive theory.
III. CUSTOM ON AN INTERPRETIVE SLIDING SCALE
The divergence between the traditional and modern approaches to custom has been criti-
cized as making custom indeterminate and malleable.154 This part analyzes one attempt at
reconciling them by applying Dworkin’s interpretive theory of law to custom on a sliding scale.
Law as Interpretation: Dworkin
Dworkin argues that legal decisions aim at reformulating past decisions and practice in
the most coherent and morally attractive way, consistent with the facts of legal history. Dworkin
outlines three steps in the interpretive process: preinterpretation, interpretation, and post-
In the preinterpretive stage, the rules and facts that form the practice to be interpreted
are identified. The term preinterpretation is somewhat misleading because the collection of
Report of the Regional Meeting for Africa of the World Conference on Human Rights, UN Doc. A/
CONF.157/AFRM/14, at 1 (1992); Report of the Regional Meeting for Asia of the World Conference on Human
Rights, UN Doc. A/CONF.157/ASRM/8, at 3 (1993).
Weil, supra note 94, at 415.
Jonathan I. Charney, Customary International Law in the Nicaragua Case Judgment on the Merits, 1988 HAGUE Y.B.
INT’L L. 16, 24.
BEDJAOUI, supra note 121, at 141–44.
Fidler, supra note 27, at 220–22, 224–25.
BYERS, supra note 6, at 49.
KOSKENNIEMI, supra note 45, at 42.
Kelly, supra note 1, at 451.
DWORKIN, supra note 46, at 66.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 771
data is theory dependent.156 Consequently, this stage requires a high degree of consensus
to provide a common focus for interpretation.157
In the interpretive stage, the interpreter formulates a general explanation for the main
elements of the practice. This explanation is called the “dimension of fit” and it imposes a
rough threshold requirement that accepts an interpretation as eligible only if the raw data
of legal practice adequately support it.158 There can be three outcomes at this stage: no
eligible interpretations, if every possible interpretation is inconsistent with the bulk of raw
material available; easy cases, when there is only one eligible interpretation; and hard cases,
where a range of eligible interpretations result because the tests are unsettled or the prac-
Postinterpretation only becomes relevant in hard cases, where interpreters must choose
the best eligible interpretation according to the “dimension of substance.”159 The best inter-
pretation is the one that makes the practice appear in the best light, judged according to
the substantive aspirations of the legal system. This criterion involves consideration of moral
and political ideals, as well as higher-order convictions about how these ideals should be
prioritized when they conflict.160
The interpretive stage is backward looking and descriptive, while the postinterpretive
stage is forward looking and normative. Dworkin’s interpretive theory of law combines a
descriptive historical investigation about what the law has been (fit) with a normative moral
inquiry about what it should be (substance).161 According to Dworkin: “propositions of law
are not simply descriptive of legal history in a straightforward way, nor are they simply
evaluative in some way divorced from legal history. They are interpretive of legal history,
which combines elements of both description and evaluation but is different from both.”162
What, however, is the relationship between fit and substance? Dworkin originally proposed
a lexical ordering between them,163 but in Law’s Empire he revised the relationship. Fit pro-
vides a rough threshold criterion.164 Interpretations need not fit every aspect of the existing
practice, but they must fit enough for the interpreter to be “interpreting that practice, not
inventing a new one.”165 On the other hand, if there are multiple eligible interpretations,
then fit and substance must be balanced against each other.166 Thus, any inadequacies of
fit will count against an interpretation at the substantive stage because “an interpretation
is pro tanto more satisfactory if it shows less damage to integrity than its rival.”167 Similarly,
“defects of fit may be compensated . . . if the principles of that interpretation are par-
ticularly attractive.”168 The interpreter must therefore balance the relative strengths of fit
and substance in different interpretations to determine the best interpretation. Dworkin
likens this process to multiple authors writing consecutive chapters in a chain novel. Each
author must contribute a chapter that provides sufficient continuity with the earlier chapters
Id. at 65–66.
Id. at 255.
Id. at 256.
Ronald Dworkin, Law and Morals, Natural Law, and Legal Positivism, in OXFORD COMPANION, supra note 57,
at 473, 473–74, 606, 606–07, and 476, 476–77, respectively; Ronald Dworkin, Law as Interpretation, 60 TEX. L. REV.
527, 528 (1982) [hereinafter Dworkin, Interpretation].
Dworkin, Interpretation, supra note 161, at 528.
RONALD DWORKIN, TAKING RIGHTS SERIOUSLY 340–41 (3d ed. 1981); RONALD DWORKIN, Is There Really No Right
Answer in Hard Cases? in A MATTER OF PRINCIPLE 143 (1985).
DWORKIN, supra note 46, at 255–57.
Id. at 66.
John Finnis, On Reason and Authority in Law’s Empire, 6 LAW & PHIL. 357, 373–74 (1987).
DWORKIN, supra note 46, at 246–47.
Id. at 257.
772 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
(fit) and develops the book in the best way (substance).169 However, Dworkin details no
method for weighing fit and substance other than the metaphor balance.170 Hence, anyone
using Dworkin’s interpretive theory must correctly identify fit and substance and develop
a working theory about how to balance them.
Custom on a Sliding Scale: Kirgis and Tasioulas
The international legal system is decentralized, resulting in multiple interpreters of custom.
The existence and content of custom is usually determined by states and academics, though
the Court remains the ultimate arbiter in some cases.171 These interpreters are meant to
determine customs on the basis of state practice and opinio juris.172 However, the traditional
and modern forms of custom both appear to emphasize one element at the expense of the
other. Kirgis argues that the two approaches can be understood by viewing state practice
and opinio juris as interchangeable along a sliding scale:173
On the sliding scale, very frequent, consistent state practice establishes a customary
rule without much (or any) affirmative showing of an opinio juris, so long as it is not
negated by evidence of non-normative intent. As the frequency and consistency of the
practice decline in any series of cases, a stronger showing of an opinio juris is required.
At the other end of the scale, a clearly demonstrated opinio juris establishes a customary
rule without much (or any) affirmative showing [of state practice].174
Yet when will a custom be formed on the basis only of state practice or opinio juris,
instead of requiring both elements? According to Kirgis, the answer depends on the im-
portance of the activity in question and the reasonableness of the rule involved. If an
activity is not very destructive, then the Court will be more exacting in requiring both state
practice and opinio juris to form a custom. By contrast, Kirgis states, “[t]he more de-
stabilizing or morally distasteful the activity—for example, the offensive use of force or
the deprivation of fundamental human rights—the more readily international decision
makers will substitute one element for the other, provided that the asserted restrictive
rule seems reasonable.”175
The sliding scale has been criticized as reinterpreting the concept of customary law to
ensure that it provides the right answers.176 Tasioulas, however, claims that the sliding scale
can be rationalized by means of Dworkin’s interpretive theory of law. While Dworkin is
primarily concerned with interpreting previous case law, custom is principally concerned
with interpreting the raw data of state practice and opinio juris. At the level of fit, the Court
must attempt to formulate eligible interpretations that can explain the raw data of state
practice and opinio juris. If there is little or no state practice or opinio juris, there will be no
eligible interpretations. In easy cases, there will be strong state practice and opinio juris,
which will produce only one eligible interpretation. In hard cases, there will be inconsistent
levels of state practice and opinio juris, giving rise to multiple eligible interpretations. For
example, high state practice and low opinio juris may indicate that torture is permitted, and
low state practice and high opinio juris may indicate that torture is prohibited. Two eligible
interpretations result, one permitting torture and the other prohibiting it. To determine
Id. at 228–32.
Finnis, supra note 166, at 374.
Bodansky, supra note 41.
ICJ STATUTE Art. 38(1)(b); North Sea Continental Shelf (FRG/Den.; FRG/Neth.), 1969 ICJ REP. 3, 44 (Feb. 20).
See Nicaragua, 1986 ICJ REP. 14; Namibia Advisory Opinion, 1971 ICJ REP. 16; Corfu Channel case (UK v.
Alb.), Merits, 1949 ICJ REP. 4, 34 (Apr. 9).
Kirgis, supra note 42, at 149.
Simma & Alston, supra note 18, at 83.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 773
the best interpretation in these cases, Tasioulas argues that the Court must balance the
dimensions of fit and substance on a sliding scale. Thus, an interpretation may be the best,
even though it fares poorly on the dimension of fit (for example, because it has strong opinio
juris but weak state practice), if the norm possesses very strong substantive appeal.177 By
contrast, the Court will be more exacting in requiring both elements in the absence of
strong substantive considerations.178
A Critique of Custom on a Sliding Scale
Although Tasioulas does not articulate this, his interpretive theory of custom actually
involves two sliding scales. The first weighs state practice and opinio juris against each other
at the level of fit to create eligible interpretations. If there are multiple eligible interpreta-
tions, the best interpretation is determined by employing a second sliding scale to balance
the dimensions of fit and substance. While interpretation is less mechanical in practice,
these stages should be differentiated to demonstrate why custom should not be analyzed on
a sliding scale.
Criticism one: apology and utopia. The first sliding scale allows inadequacies in state practice
to be compensated for by strong opinio juris and vice versa. However, this procedure paves
the way for normative criticism because it allows the strength of one element to outweigh
deficiencies in the other. Finding traditional custom on the strength of state practice alone
allows it to become an apology for state power.179 Similarly, deducing modern custom purely
from opinio juris can create utopian laws that cannot regulate reality.
Arguably, interpretations are prevented from lapsing into apology and utopia because the
second sliding scale balances deficiencies in fit against the strength of the dimension of
substance. But this argument fails for two reasons. First, the dimension of substance becomes
relevant only if there is more than one eligible interpretation. Thus, a custom may be the
only eligible interpretation on the basis of strong state practice without any reference to its
procedural or substantive normativity. In this case, custom would become an apology for
exercises of power. Second, the sliding scale allows strong opinio juris to create an eligible
interpretation at the dimension of fit, which is then weighed according to the second slid-
ing scale for its substantive appeal. However, opinio juris is often an expression of lex ferenda,
so that the two sliding scales would allow modern custom to be formed by normative con-
siderations of lex ferenda at the level of fit and substance. In this case, utopian customs
could be produced without any guarantee of the descriptive accuracy associated with fit
and state practice.
Criticism two: asymmetrical application. Kirgis and Tasioulas claim that the more morally
distasteful an activity, the more readily the Court will substitute opinio juris for state practice
and vice versa.180 This works for modern custom, where there is strong opinio juris but weak
state practice, for example in the Nicaragua case. However, the Court seems willing to find
traditional custom, where there is strong state practice and weak opinio juris, without con-
sidering whether the activity is morally distasteful, for example in S.S. Wimbledon.181 Thus,
it is wrong to conclude that the stronger the substantive considerations involved, the more
likely the Court is to substitute one element for the other. Instead, questions of substance
seem likely to arise in situations where the problems with fit concern a lack of affirming state
practice rather than a lack of opinio juris.
Tasioulas, supra note 42, at 113.
E.g., North Sea Continental Shelf, 1969 ICJ REP. 3.
KOSKENNIEMI, supra note 45, at 2.
Kirgis, supra note 42, at 149.
S.S. Wimbledon, 1923 PCIJ (ser. A) No. 1, at 25 (Aug. 17); see also North Sea Continental Shelf, 1969 ICJ REP.
at 231 (Lachs, J., dissenting); id. at 241, 246–47 (Sørensen, J. ad hoc, dissenting); Nottebohm case (Liech. v. Guat.),
Second Phase, 1955 ICJ REP. 4, 22 (Apr. 6).
774 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
Employing a sliding scale between fit and substance does not adequately describe the
process of finding custom because it applies asymmetrically to traditional and modern
custom. It does so for two reasons. First, the second sliding scale is employed only if more
than one interpretation meets the threshold requirement of fit.182 Traditional custom fares
well on fit because it focuses on state practice, which is descriptively accurate. Modern cus-
tom fares less well on fit because opinio juris often cloaks lex ferenda as lex lata and thus lacks
descriptive accuracy. Modern custom also involves practice that is hard to identify and
evaluate, such as intrastate action and inaction. Consequently, traditional custom is more
likely than modern custom to result in only one eligible interpretation, often removing the
need for substantive considerations and a second sliding scale.
Second, the importance of substantive considerations also varies according to the facili-
tative and moral content of traditional and modern custom. Traditional customs are often
facilitative, so that even when there is a deficiency in fit owing to insufficient opinio juris, the
postinterpretive stage has little application because the rules do not involve substantive
moral issues. Traditional customs should still be procedurally normative, but this aspect does
not provide a strong dimension of substance. By contrast, modern customs often have a
strongly moral content. Thus, modern custom involves strong substantive considerations
about what the law should be, as well as respect for procedural normativity to ensure that
all nations are able to express their views about what substantive aims should be pursued.
For these reasons, modern custom is more likely than traditional custom to give rise to a
strong dimension of substance and a second sliding scale.
IV. CUSTOM AS A REFLECTIVE INTERPRETIVE CONCEPT
Dworkin’s interpretive theory of law incorporates the values of fit (description) and sub-
stance (normativity) but does not provide a mechanism for balancing them. Analyzing cus-
tom on a sliding scale is flawed because it misconceives the nature of fit and substance and
risks creating customs that are apologies for power or utopian and unachievable. This part
presents an alternative vision of Dworkin’s interpretive theory as applied to custom, which
re-conceptualizes the nature of fit and substance and balances them in a Rawlsian reflective
The Theory of Custom as a Reflective Interpretive Concept
Preinterpretation. At the preinterpretation stage, the relevant data for interpretation are
identified. For custom, these are generally assumed to be state practice and opinio juris. How-
ever, the collection of data is theory dependent and our interpretive approach is limited by
the assumptions we make at this stage.183 Focusing exclusively on state practice and opinio
juris is theory dependent because it reflects a statist approach to international law. It is gen-
erally assumed that states, acting individually or collectively in intergovernmental organi-
zations, are the only relevant actors in the formation of custom.184 Nevertheless, nonstate
actors also affect the determination and development of custom in various ways.185
First, judicial decisions and writings by publicists can be used as a subsidiary means of
determining international law.186 Though not formal sources of law, they may provide
DWORKIN, supra note 46, at 246–47, 255–57.
John King Gamble & Charlotte Ku, New Actors and New Technologies: Center Stage for NGOs? 31 LAW & POL’Y
INT’L BUS. 221, 243–44 (2000).
Karsten Nowrot, Legal Consequences of Globalization: The Status of Non-Governmental Organizations Under International
Law, 6 IND. J. GLOBAL LEGAL STUD. 579, 595 (1999); Peter J. Spiro, New Global Potentates: Nongovernmental Organi-
zations and the “Unregulated” Marketplace, 18 CARDOZO L. REV. 957, 959–60 (1996).
ICJ STATUTE Art. 38(1)(d).
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 775
authoritative evidence of the state of the law. In practice, however, judicial decisions can
also have a formative effect on custom by crystallizing emerging rules and thus influencing
state behavior.187 Moreover, academics often cite the work of other writers as evidence of the
existence and content of custom instead of thoroughly analyzing state practice. Yet how
judicial decisions and academic writings could be incorporated into an interpretive analysis
of custom is unclear. For example, Dworkin includes judicial decisions in the dimension of
fit when interpreting domestic law, but this may be inappropriate in international law, which
lacks a doctrine of precedent.188 Custom is also concerned with interpreting past practice,
rather than past case law, and judicial decisions do not always accurately describe practice.189
Second, nonstate actors such as nongovernmental organizations and corporations often
have an impact on custom. They have an indirect effect by influencing state behavior and
statements through actions such as lobbying and calling boycotts.190 They have assisted in
setting the agenda for international conferences and participated in the negotiation and
drafting of treaties and resolutions.191 Nongovernmental organizations help to articulate
emerging customs and monitor state compliance with international law by investigating and
publicizing breaches of the law in areas such as human rights and environmental protec-
tion.192 Isabelle Gunning has also argued that nonstate actors should be granted a direct role
in the formation of custom,193 though this proposal is controversial because it may give these
actors power without accountability.194
The present and potentially increasing role of nonstate actors in the development of
custom is undertheorized and beyond the scope of this paper. Nevertheless, it is an impor-
tant issue and further analysis of it may change our perception of custom in the future.
Identifying the dimension of fit. According to Dworkin, a “successful interpretation must not
only fit but also justify the practice it interprets.”195 An interpretive theory of custom com-
bines descriptive accuracy (fit) with normative considerations (substance).196 Questions of
fit and description are both backward looking, analyzing the extent to which a decision
accords with past practice. Normativity and questions of substance are both forward looking,
involving moral and political considerations relevant to what the law should be. This charac-
terization has important implications for properly identifying fit and substance.
Tasioulas argues that state practice and opinio juris are both relevant to the dimension of
fit. However, the main function of fit is to ensure that interpretations are descriptively accu-
rate. State practice is integral to fit because of its descriptive accuracy, but opinio juris is prob-
lematic because statements are descriptive when they express lex lata and normative when
they express lex ferenda. Ideally, one should be able to distinguish between lex lata (fit) and lex
ferenda (substance). However, as demonstrated above, it is often not possible to distinguish
between the two by looking at the language or travaux préparatoires of treaties and dec-
See examples in BROWNLIE, supra note 6, at 19–25.
ICJ STATUTE Art. 59; see also BROWNLIE, supra note 6, at 20–21.
For example, torture was held to be against customary international law but remains common throughout
the world. Filartiga v. Pena-Irala, 630 F.2d 876 (1980).
For example, consumer boycotts were called on French products over French nuclear testing. Protests Hurt
Wine’s Sales, N.Y. TIMES, Dec. 25, 1995, §1, at 58. Similarly, many U.S. corporations terminated their presence in
South Africa in protest over apartheid well before they were required to do so by federal laws. Kevin P. Lewis,
Dealing with South Africa: The Constitutionality of State and Local Divestment Legislation, 61 TUL. L. REV. 469 (1987).
For example, the International Committee of the Red Cross played a pivotal role in the initiation and negoti-
ation of the Ottawa convention banning land mines. Kenneth Anderson, The Ottawa Convention Banning Landmines,
the Role of International Non-Governmental Organizations and the Idea of International Civil Society, 11 EUR. J. INT’L L.
Nowrot, supra note 185, at 596–98.
Isabelle R. Gunning, Modernizing Customary International Law: The Challenge of Human Rights, 31 VA. J. INT’L
L. 211, 227–34 (1991).
Spiro, supra note 185, at 962–67.
DWORKIN, supra note 46, at 285.
776 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
larations. Thus, strong opinio juris cannot provide an eligible interpretation in the absence
of state practice because that would not be descriptively accurate.
Instead, statements of opinio juris should be relevant to interpretation in two ways. First,
state practice must be accompanied by some articulation of legality so as to distinguish be-
tween legal and social obligations.197 For example, giving aid to developing countries prob-
ably could not form a customary obligation (even though it is supported by state practice
and involves strong substantive considerations) because states have never articulated it as
a legal obligation. Second, strong opinio juris is relevant at the level of substance because
it represents lex ferenda considerations about what the law should be. Thus, a declaration
or treaty adopted by a representative majority of states will provide a strong dimension of
Dworkin’s theory envisages fit as providing a rough threshold for interpretations.198 As we
have seen, there are three possible outcomes at this stage. When there is no eligible inter-
pretation, there will be no custom and the legality of action will be determined according
to the Lotus principle, which states that “[r]estrictions upon the independence of States
cannot therefore be presumed.”199 Second, when there is only one eligible interpretation
because of the uniformity of the practice, there will be no doubt about the existence of a
custom, such as customs on diplomatic immunity. Finally, when there are multiple eligible
interpretations, one must weigh the dimensions of fit and substance to determine the best
I contend that because of its superior descriptive accuracy, traditional custom will fare well
on fit and is more likely than modern custom to result in only one eligible interpretation.
However, modern custom can still pass the threshold of fit by allowing for multiple eligible
interpretations if we adopt a more nuanced approach to state practice in two ways.
First, state practice is not an automatic operation but is open to interpretation.200 The
sliding scale assumes that there must be a trade-off between state practice and opinio juris
because they are fixed and irreconcilable quantities. However, inconsistent state practice
can be interpreted as a breach of an existing custom or a seed for a new custom,201 which
is a particularly important consideration in modern customs such as human rights obli-
gations, where breaches seem reasonably common and widespread.202 Contrary state prac-
tice does not necessarily undermine an eligible interpretation but, rather, may create mul-
tiple eligible interpretations. For example, NATO’s intervention in Kosovo could be inter-
preted as a breach of an existing customary prohibition on intervention or as the seed for
a new custom allowing unilateral humanitarian intervention.203 Thus, modern custom can
produce multiple eligible interpretations of state practice without needing to employ a
sliding scale between state practice and opinio juris.
GÉNY, in D’AMATO, supra note 6, at 49.
DWORKIN, supra note 46, at 246–47, 255–57.
S.S. “Lotus” (Fr. v. Turk.), 1927 PCIJ (ser. A) No. 10, at 18 (Sept. 7). The Court affirmed this principle in
Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion, 1996 ICJ REP. 226, 238, para. 20 ( July 8).
These decisions establish a residual negative principle, which provides that in the case of a non liquet (an absence
of law), whatever is not prohibited in international law is permitted. Daniel Bodansky, Non Liquet and the Incom-
pleteness of International Law, in INTERNATIONAL LAW, THE INTERNATIONAL COURT OF JUSTICE AND NUCLEAR WEAPONS
153 (Laurence Boisson de Chazournes & Philippe Sands eds., 1999); Ole Spiermann, Lotus and the Double Structure
of International Legal Argument, in id. at 131; Julius Stone, Non Liquet and the Function of Law in the International
Community, 1959 BRIT. Y.B. INT’L L. 124, 135.
Martti Koskenniemi, The Pull of the Mainstream, 88 MICH. L. REV. 1946, 1952 (1990); Fastenrath, supra note
49, at 316–17.
D’AMATO, supra note 6, at 97–98.
SCHACHTER, supra note 15, at 335.
See “The Fluid Nature of Custom,” infra p. 784.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 777
Second, modern custom is more likely to yield eligible interpretations if we broaden our
understanding of state practice beyond the traditional forms of interstate interaction and
acquiescence. The atrocities of World War II, the globalization of trade, and improvements
in communications have all diminished the importance of state boundaries.204 International
law now covers many intrastate issues such as the human rights obligations, which largely
protect citizens from their own governments.205 These intrastate issues give rise to little or
no interaction between states.206 In its stead, we need to broaden our understanding of state
practice to include consideration of intrastate action (not just interstate interaction), obliga-
tions being observed (not just obligations being breached), and reasons for a lack of protest
over breaches (other than acquiescence in the legality of those breaches).
State practice should include intrastate practice rather than just interstate interaction
because of the changing subject matter of international law. Patterns of domestic practice,
such as decisions of domestic courts, are recognized as relevant forms of state practice.207
Internal breaches of human rights obligations are often characterized as instances of con-
trary state practice.208 Yet there is little corresponding focus on governments’ respect for
human rights obligations as positive instances of state practice. For one thing, adequate
evidence of internal compliance with human rights obligations may be hard to obtain.209
Breaches are sensational and likely to attract attention, but that should not be allowed to
obscure a general pattern of less publicized observance.210 The observance of many human
rights is also difficult to measure because they are negative rights, which means that they
place limitations on state action rather than impose a positive duty on states to act. Obser-
vance by inaction, in the form of not violating rights, is inherently ambiguous because it may
result from an obligation (prohibitive norm) or discretion (permissive norm);211 or from
domestic or treaty obligations rather than custom.212 However, even though intrastate action
and inaction may be difficult to identify and evaluate, they are still important forms of state
practice in modern international law.
Breaches of intrastate obligations are also likely to result in inaction by other states be-
cause states do not usually protest violations unless they affect their rights or the rights of
their nationals. According to the theory of acquiescence, opinio juris in favor of the legality of
action can be inferred if states fail to protest when faced with an infringement of their
rights.213 Some intrastate obligations, such as fundamental human rights obligations, can
theoretically be enforced by any state because they are owed erga omnes.214 Thus, Karol Wolfke
and Weisburd claim that inaction over intrastate breaches amounts to acquiescence, which
undermines the customary status of these obligations because rules that are not enforced
Richard A. Falk, Kosovo, World Order, and the Future of International Law, 93 AJIL 847, 853 (1999).
Charney, supra note 21, at 543; Henkin, supra note 25, at 36.
LOUIS HENKIN, INTERNATIONAL LAW: POLITICS AND VALUES 180–81 (1995); Bradley & Goldsmith, supra note
35, at 840–41; Henkin, supra note 25, at 38; Simma & Alston, supra note 18, at 99.
E.g., Filartiga v. Pena-Irala, 630 F.2d 876 (1980); The Paquete Habana, 175 U.S. 677 (1900); see also M. O.
Chibundu, Making Customary International Law Through Municipal Adjudication: A Structural Inquiry, 39 VA. J. INT’L
L. 1069 (1999).
Chodosh, supra note 17, at 121; Simma & Alston, supra note 18, at 90–100; Watson, supra note 143, at 609–12;
Arthur A. Weisburd, The Effect of Treaties and Other Formal International Acts on the Customary Law of Human Rights,
25 GA. J. INT’L & COMP. L. 99, 129 (1995/96); Weisburd, Customary IL, supra note 31, at 41.
Schachter, supra note 48, at 10.
MERON, supra note 2, at 61.
S.S. “Lotus” (Fr. v. Turk.), 1927 PCIJ (ser. A) No. 10, at 28 (Sept. 7). The Court affirmed this principle in
North Sea Continental Shelf (FRG/Den.; FRG/Neth.), 1969 ICJ REP. 3, 44 (Feb. 20); see also Simma & Alston, supra
note 18, at 99–100, 104.
D’AMATO, supra note 6, at 62, 63; Akehurst, supra note 7, at 43–47.
Akehurst, supra note 7, at 37; MacGibbon, Acquiescence I, supra note 115, at 145–46; MacGibbon, Acquiescence
II, supra note 115, at 138.
Barcelona Traction, Light & Power Co. (Belg. v. Spain), Second Phase, 1970 ICJ REP. 3, 31 (Feb. 5).
778 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
are not legal rules.215 However, horizontal enforcement of intrastate obligations is unlikely
to work in practice because the right of enforcement (which belongs to states) is divorced
from the interest being protected (which belongs to individuals).216 Silence in the absence
of an obligation to speak should not necessarily imply consent.217 Many plausible explana-
tions can be made for a failure to protest intrastate breaches other than belief in the legality
of the action, including lack of knowledge, political and economic self-interest, and realiza-
tion of the futility of action.218 The lack of protest over intrastate breaches should not neces-
sarily imply acquiescence in the legality of those breaches. Therefore, a lack of protest may
also give rise to multiple interpretations.
In summary, an interpretation is only eligible if it passes the dimension of fit, which exam-
ines whether the interpretation accurately describes the raw data of practice. Although tradi-
tional customs are likely to form eligible interpretations because of their focus on state prac-
tice, modern custom can also pass the threshold of fit if state practice includes intrastate
action and inaction and we recognize that conflicting practice can give rise to multiple
Identifying the dimension of substance. The dimension of substance requires a moral and
political theory of international law, which is problematic because moral and political ideals
vary among states.219 Kirgis refers to strong moral issues but does not provide a theory for
determining morality in international law. Tasioulas limits his discussion of substantive aims
to the coexistence and cooperation of states, which does not account for many modern aims
of international law such as human rights. I contend that the dimension of substance should
include both procedural and substantive normative considerations about whether the con-
tent of custom is substantively moral and whether it is derived by a legitimate process. By
“moral,” I am again referring to commonly held subjective values about right and wrong that
have been adopted by a representative majority of states in treaties and declarations. This
concept includes substantive aims, such as the protection of human rights and the environ-
ment, as well as higher-order convictions about how these substantive aims should be pri-
oritized when they compete.220 For example, how should conflicts between human rights and
state sovereignty be resolved in humanitarian intervention?
Statements of opinio juris will often be relevant at the level of substance because they are
normative statements about what the law is or should be.221 Allowing the dimension of sub-
stance to include procedural and substantive normative considerations helps to justify the
importance of this dimension. Substantive normativity has been criticized for confusing law
and politics,222 and obliterating the distinction between “discovering” and “inventing” the
law.223 While this issue is beyond the scope of this article, it merits a brief comment. First,
substantive considerations become relevant only if the threshold of fit has been passed
and they are weighed against fit.224 Thus, considerations of lex ferenda need not be elimi-
nated, because they are relevant to the development of the law and will not form a custom
in the absence of adequate state practice at the level of fit. Second, the dimension of fit
WOLFKE, supra note 115, at 135; Weisburd, supra note 208, at 107–09; see also Akehurst, supra note 7, at 39.
Henkin, supra note 25, at 41.
DWORKIN, supra note 46, at 192–93; Kelly, supra note 1, at 522.
D’AMATO, supra note 6, at 68–70; Charney, supra note 21, at 536–38.
Kevin T. Jackson, Global Rights and Regional Jurisprudence, 12 LAW & PHIL. 157, 158–59 (1993).
DWORKIN, supra note 46, at 256.
Christine Chinkin, The Challenge of Soft Law: Development and Change in International Law, 38 INT’L & COMP.
L.Q. 850, 861, 865 (1989); Fastenrath, supra note 49, at 323.
Holmes, supra note 52, at 459–62; Martti Koskenniemi, The Police in the Temple—Order, Justice and the UN: A
Dialectical View, 6 EUR. J. INT’L L. 325 (1995).
Kelly, supra note 1, at 458; Kennedy, supra note 102, at 347.
DWORKIN, supra note 46, at 246–47, 255–57.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 779
is also interpretive and political, rather than mechanical.225 The collection and interpre-
tation of data is theory dependent and biased by assumptions.226 Fit also represents a form
of “ideological nostalgia”227 because it demonstrates a conservative political preference for
the past and the status quo.228 Third, the substantive moral principles relied upon result
from statements of opinio juris, rather than abstract moral considerations, so they have been
“given sufficient expression in legal form.”229 They have also been formulated by a majority
of states rather than judges,230 which provides procedural normativity and helps to prevent
criticism of Western ideological bias.231 Finally, self-consciously value-based adjudication can
also increase determinacy and accountability by making the reasons for a decision explicit.232
Balancing Fit and Substance: A Reflective Methodology
Applying Dworkin’s interpretive theory to custom is useful because it incorporates descrip-
tion (fit) and normativity (substance) into a single approach. If there are multiple eligible
interpretations, the best interpretation is the one that most coherently explains the dimen-
sions of fit and substance.233 Coherence, which Dworkin calls integrity, is an important value
in international law.234 Interpretations should aim for coherence by mediating between fit
and substance in the same way as Rawls mediates between intuitions and moral principles
in ethical theories.235 Instead of prioritizing one value above the other, Rawls advocates
revising our interpretation of practice and principles, back and forth, until we have done
everything possible to render our interpretation coherent and justified from both ends.236
Rawls calls this a “reflective equilibrium.”237
To illustrate this process in moral theory, consider the moral principle that killing another
person is wrong. This principle generally coincides with our intuitive judgments, but there
are exceptions. For example, we may intuitively feel that someone who kills in self-defense
is not guilty of murder. Instead of arbitrarily rejecting this intuition or creating an unprin-
cipled exception to the general rule, a reflective equilibrium seeks to provide the most co-
herent explanation of both positions. Killing in self-defense is not as morally culpable as
murder because it lacks a mental fault element. Accordingly, we could amend our general
principle to provide that killing another person is wrong if it is intentional. However, what
if the killing was unintentional but reckless? We could amend our intuition to provide that
killing in self-defense may be morally justifiable when it is a proportionate response to a
perceived threat. If it is a disproportionate response, then the killing is not morally justified,
Id. at 257.
Stanley Fish, Working on the Chain Gang: Interpretation in Law and Literature, 60 TEX. L. REV. 551, 559 (1982);
Stanley Fish, Wrong Again, 62 TEX. L. REV. 299, 306 (1983).
Richard A. Falk, Presentation [on the extent of ideological neutrality of international law and international
lawyers], in CHANGE AND STABILITY, supra note 3, at 137, 137.
BEDJAOUI, supra note 121, at 136–37.
South West Africa (Eth. v. S. Afr.; Liber. v. S. Afr.), Second Phase, 1966 ICJ REP. 6, 34 ( July 18); cf. Christian
Tomuschat, Obligations Arising for States Without or Against Their Will, 241 RECUEIL DES COURS 195, 303 (1993 IV).
Kennedy, supra note 102, at 407.
BEDJAOUI, supra note 121, at 141. While developing states hold the majority of power, these norms may still
be criticized for allowing the majority to bind the minority.
KARL N. LLEWELLYN, THE COMMON LAW TRADITION: DECIDING APPEALS 35–45 (1960).
See also “coherentism” in epistemology. NICHOLAS EVERITT & ALEC FISHER, MODERN EPISTEMOLOGY 102–07 (1995).
THOMAS M. FRANCK, FAIRNESS IN INTERNATIONAL LAW AND INSTITUTIONS 38 (1995); THOMAS M. FRANCK, THE
POWER OF LEGITIMACY AMONG NATIONS 150 (1990); Franck, supra note 56, at 735–36.
RAWLS, supra note 46, at 20; John Rawls, Outline of a Decision Procedure for Ethics, 60 PHIL. REV. 177 (1951); see
also DWORKIN, supra note 46, at 66 n.17. I am not arguing that the international system is akin to Rawls’s concept
of the “original position”; rather, that his notion of a reflective equilibrium can be used more generally to recon-
cile inductive and deductive methodologies.
RAWLS, supra note 46, at 21.
Id. at 19–20.
780 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
though it may be less morally culpable than murder. This circumstance may provide a par-
tial defense to murder, such as provocation, which would reduce a conviction from murder
to manslaughter. Thus, killing another person is wrong if it is intentional or reckless, subject
to a complete defense of self-defense or a partial defense of provocation. This process of
revising both our intuitions and our moral principles and testing them against each other
continues until the two correspond in a reflective equilibrium.
Rawls does not detail a method for determining which end of the spectrum should be
revised to achieve equilibrium.238 Instead of striking an unprincipled compromise, one could
draw an analogy with scientific method, which also seeks to reconcile observations with ab-
stract justifications.239 Scientists develop hypotheses inductively based on observed phenom-
ena and then test these hypotheses deductively against interpretations of experimental
results in a two-way harmonizing process, modifying the theory or practical conditions until
the theory coherently explains observed reality. This process involves mediating between
radical empiricism (which emphasizes the raw facts of experience as primary) and ration-
alism (which gives preference to the importance of scientific theory in interpreting facts).240
If a general principle conflicts with particular experimental results, then either the general
rule should be modified to explain those results or the practical conditions for the experi-
ment should be revised to ensure that accurate results are obtained. In accordance with this
method, a widely supported principle might be modified if there are isolated counter-
examples, whereas a weakly supported theory should be rejected if the contrary data are
overwhelming. However, conflicting data should never be discounted as irrelevant because
what is currently understood as a breach of, or an exception to, an existing theory may later
form the basis of a new theory.
To illustrate this scientific approach, consider the competing explanations of planetary
motion. The Aristotelian geocentric view of the universe (that the sun and planets revolve
around the earth) was discarded in the face of overwhelming contrary observations. This
theory was replaced by the Copernican heliocentric theory of planetary motion (that the
planets revolve around the sun). However, the Copernican theory also fails to explain many
practical observations about the movement of the planets. Instead of modifying the general
theory, these exceptions have been rationalized by examining the many disturbances that
obscure the pure case, such as tidal friction, atmospheric refraction, and instrument error.
Thus, a weakly supported theory was discarded in the face of considerable conflicting data,
whereas a widely supported theory has been upheld despite some contrary data, because
those data could be better explained in other ways. Nevertheless, the conflicting infor-
mation is still relevant for future interpretations because it may in time support a new
theory. For example, Galileo’s theory of the circular planetary orbits was useful to explain
the general movement of the planets but could not explain some observed irregularities in
their orbits. Instead of discarding these irregularities as irrelevant, they were later explained
by embracing Kepler’s model of elliptical orbits. Thus, the best explanation of practice and
principles will vary in light of new state practice, opinio juris, and moral considerations.
While custom and moral theories involve very different considerations, a Rawlsian reflec-
tive approach can be used more generally to reconcile the inductive and deductive method-
ologies that are common to both.241 The dimension of fit and the theory of moral intu-
itionism are both inductive because they seek to infer a general rule from particular
instances of practice. The dimension of substance is similar to moral principles because both
deduce norms about practice from abstract statements of principles. Instead of prioritizing
G. R. Grice, Moral Theories and Received Opinion, 52 ARISTOTELIAN SOC’Y PROC. 1 (Supp. 1978); R. M. Hare,
A Critical Study of John Rawls’ A Theory of Justice, 23 PHIL. Q. 144, 144–55 (1973).
ROBIN ATTFIELD, ENVIRONMENTAL PHILOSOPHY AND PROSPECTS 92 (1994); RAWLS, supra note 46, at 20.
Paul K. Feyerabend, History of Philosophy of Science, in OXFORD COMPANION, supra note 57, at 806, 806–07.
RAWLS, supra note 46, at 20 n.7; see also NELSON GOODMAN, FACT, FICTION, AND FORECAST 65–68 (4th ed. 1983).
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 781
one value above the other, we should revise our interpretation of practice and principles
until the two approaches are coherent and justified from both ends in a reflective equi-
librium. While Dworkin argues that the best interpretation is objectively determinable,242
I believe that subjectivity is inherent in interpreting raw data. However, the reflective pro-
cess provides some guidelines for reconciling practice and principles rather than allowing
one element to override the other. It also explains why the best interpretation of practice
and principles will change over time in light of new data or theories.
Practical Application of the Reflective Interpretive Concept
A reflective equilibrium requires revising our interpretations of practice (fit) and principles
(substance) to find the most coherent explanation of both dimensions in the case of mul-
tiple eligible interpretations. This section explores how the reflective interpretive approach
applies to two practical problems in custom, first, by applying it to the spectrum of facili-
tative and modern customs, and second, by using it to explain the fluid nature of custom.
This analysis demonstrates that the reflective interpretive approach is dynamic, varying over
time and according to the nature of the custom involved.
The spectrum of facilitative and moral customs. The best balance between fit and substance will
depend on the relative strength of the practice and principles in the custom involved. This
is because the best interpretation of a custom usually leans toward the stronger element
(practice or principles) as providing the greatest consistency with both elements. Tradi-
tional facilitative customs will result in a more descriptive equilibrium because they do not
involve strong issues of principle. By contrast, modern customs with a strong moral content
require a more normative equilibrium because they involve important issues of principle.
Other customs, such as environmental protection, may involve strong normative and descrip-
tive considerations, requiring a more even balance of fit and substance. Thus, traditional
and modern customs exist on a facilitative-moral spectrum, which explains the asymmetrical
application of fit and substance to them.
To illustrate this spectrum, let us consider torture. State practice is ambiguous because
some states torture their citizens with minimal protest by other states, while other states do
not engage in torture. Further, most states have accepted declarations against torture (opinio
juris). This situation produces at least two eligible interpretations. First, torture is permitted,
which explains instances of torture and the frequent lack of protests (acquiescence). Other
states refrain from torturing citizens out of choice rather than obligation, while the statements
of opinio juris represent lex ferenda because they do not accord with practice. Second, torture
is prohibited because many states refrain from torturing people out of a sense of obligation
and most states have formally accepted that torture is illegal. Instances of torture breach the
rule and may fail to provoke a response because of diplomatic, economic, and political con-
straints on other states. These two plausible interpretations demonstrate that state practice and
opinio juris are not fixed considerations but are open to interpretation. As there is more than
one eligible interpretation, the dimension of substance becomes relevant to determining
the best interpretation. At this stage, lex ferenda statements about the immorality of torture
carry important weight in representing what states believe the law should be. As a result, the
most coherent explanation of fit and substance would probably be that torture is illegal.
To illustrate the other end of the spectrum, consider a facilitative custom that all ships
should pass on the left. A survey of practice may determine that ships usually pass on the left
but often pass on the right as well. One would need to consider whether passing on the left
or right corresponds to particular regions or states and whether either alternative results in
protests by states. Further, this practice can only form a custom if there are statements that
DWORKIN, supra note 46, at 239–40.
782 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
passing on the left or right is a legal obligation, not just a social practice (opinio juris). If
ships usually pass on the left, this usage may provide the only eligible interpretation without
any consideration of the dimension of substance. Alternatively, if the practice is more equi-
vocal, it may produce several eligible interpretations. For example: all ships must pass on
the left; all ships must pass on the right; all ships must pass on the left (general custom)
except in particular regions where they pass on the right (a special custom); or all ships
should pass on the left, but some states are persistent objectors and may pass on the right.
As there are multiple eligible interpretations, one should consider their merits in light of
the dimension of substance. While having a custom to determine how ships should pass
makes an important contribution to coexistence and cooperation, it does not involve sub-
stantive moral issues. Thus, as the custom is facilitative, it will be determined primarily by
which eligible interpretation best satisfies fit. In this case, the custom would be that all ships
should pass on the left.
The middle of the spectrum is occupied by laws that facilitate coexistence and involve
moral considerations, such as an environmental duty to prevent transboundary harm. To
determine eligible interpretations, we need to consider the following: actions (including
actual examples of transboundary pollution); inaction (including instances of states’ refrain-
ing from, or attempting to limit, transboundary pollution); statements articulating the
legality or illegality of transboundary harm (to determine whether it is a legal, rather than
a merely social, issue); and examples of protests to or acquiescence in instances of trans-
boundary harm by other states. These considerations give rise to at least two eligible inter-
pretations. First, there is a customary norm prohibiting transboundary pollution, which
explains why most states refrain from causing interstate pollution most of the time and why
instances of pollution are often met with protests.243 Second, there is a permissive customary
norm allowing transboundary pollution, which explains why pollutants regularly travel
across international borders, provoking little or no reaction by other states.244 Statements
of opinio juris (asserting that there should be a legal right to pollute or a legal obligation not
to pollute) are mainly relevant at the dimension of substance because they represent what
the practice should be, not necessarily what it has been. These statements generally support
the view that states are, or should be, under a duty to prevent transboundary harm.245 For
example, Principle 21 of the Stockholm Declaration provides that states have “the respon-
sibility to ensure that activities within their jurisdiction or control do not cause damage to
the environment of other States or of areas beyond the limits of national jurisdiction.”246
Thus, the best interpretation of practice and principles probably leans in favor of a pro-
hibition on transboundary harm. However, as this outcome is more finely balanced than the
previous examples, it is more likely to be contentious and open to change.247
PATRICIA W. BIRNIE & ALAN E. BOYLE, INTERNATIONAL LAW AND THE ENVIRONMENT 89 (1992); Pierre-Marie
Dupuy, Overview of the Existing Customary Legal Regime Regarding International Pollution, in INTERNATIONAL LAW AND
POLLUTION 61 (Daniel B. Magraw ed., 1991); Rüdiger Wolfrum, Purposes and Principles of International Environmental
Law, 1990 GER. Y.B. INT’L L. 308, 309.
Bodansky, supra note 41, at 110–13 (noting that transboundary pollution seems to be the rule rather than the
exception and that customary environmental norms seem to be based on statements rather than regularities in behavior);
Oscar Schachter, The Emergence of International Environmental Law, 44 J. INT’L AFF. 457, 462–63 (1991) (noting that there
is only fragmentary evidence of state practice and opinio juris to support international environmental principles).
E.g., Convention on Long-Range Transboundary Air Pollution, Nov. 13, 1979, TIAS No. 10,541, 1302 UNTS
217; Co-operation in the Field of Environment Concerning Natural Resources Shared by Two or More States, GA
Res. 3129, UN GAOR, 28th Sess., Supp. No. 30, at 48, UN Doc. A/9030 (1973); OECD Council Recommendation
C(74)224, Nov. 14, 1974, tit. B(2), reprinted in ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT,
OECD AND THE ENVIRONMENT 142 (1986). See generally Bodansky, supra note 41, at 110–12, 114–15.
Stockholm Declaration on the Human Environment, Princ. 21, REPORT OF THE UNITED NATIONS CONFERENCE
ON THE HUMAN ENVIRONMENT 3, UN Doc. A/CONF.48/14/Rev.1 (1972), reprinted in 11 ILM 1416 (1972). This
principle was reiterated (in slightly modified form) in Rio Declaration on Environment and Development, UN
Doc. A/CONF.151/5/Rev.1 (1992), reprinted in 31 ILM 874 (1992).
BROWNLIE, supra note 6, at 285–86.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 783
These examples illustrate that the best balance between fit and substance varies according
to the relative strength of the practice and principles involved. Strong substantive con-
siderations may compensate for a relatively weak fit, while equivocal substantive consider-
ations will require a finer balance between fit and substance. For example, the moral consid-
erations involved in environmental pollution are strong but not as compelling as the moral
imperative against torture. Consequently, a customary environmental norm will require greater
conformity with state practice, and is more likely to be undermined by conflicting practice,
than a customary prohibition on torture. As Schachter argues: “[I]nternational rules are not
all equal. Some are more important than others because they express deeply-held and widely
shared convictions as to the unacceptability of the prohibited conduct. . . . Contrary and
inconsistent practice would not and should not defeat their claims as customary law.”248
The relative normativity of customs may help to explain the peremptory nature of jus
cogens norms. The Vienna Convention on the Law of Treaties defines jus cogens norms as laws
“accepted and recognised by the international community of states as a whole . . . from
which no derogation is permitted and which can be modified only by a subsequent norm
of general international law having the same character.”249 Examples of jus cogens norms in-
clude the prohibitions on the use of force250 and genocide.251 Scholars disagree over whether
peremptory norms are part of customary international law. Lauri Hannikainen and Ian
Brownlie argue that custom is a suitable source for peremptory norms because it serves as
a vehicle for generally binding international law on important moral issues.252 By contrast,
Weil takes issue with the concept of jus cogens rules, arguing that they are not truly custom-
ary because they can be asserted despite a lack of state practice and consent by states.253
It is unclear whether rules of jus cogens can be analyzed according to a reflective interpre-
tive approach for the same reasons that it is unclear whether they are truly customary rules.
On the one hand, these norms are supported by extremely strong substantive considerations,
which are likely to outweigh contrary practice and deficiencies in fit. The substantive moral-
ity of jus cogens norms may also outweigh the procedural normativity of allowing states to
become persistent objectors to them.254 In addition, jus cogens norms are supported by wide-
spread state practice in the form of inaction. For example, most states do not engage in
officially recognized torture most of the time. Breaches may be sensational and well re-
ported, but that should not be allowed to eclipse a general pattern of less publicized ob-
servance.255 The significance of contrary state practice may also be discounted because
offending states usually base their denials of having breached the law on the facts, rather
than on claims of invalidity of the law itself.256 A state’s resort to factual or legal exceptions
to justify a prima facie breach of a rule has the effect of confirming the general rule, rather
than undermining it or creating an exception to it.257
Schachter, supra note 48, at 11 (citation omitted); see also Schachter, supra note 81, at 733–34.
Vienna Convention on the Law of Treaties, supra note 83, Art. 53.
CASSESE, supra note 83, at 147.
Malcolm N. Shaw, Genocide in International Law, in INTERNATIONAL LAW AT A TIME OF PERPLEXITY, supra note
81, at 797.
BROWNLIE, supra note 6, at 514–17; LAURI HANNIKAINEN, PEREMPTORY NORMS ( JUS COGENS) IN INTERNA-
TIONAL LAW 226–42 (1988).
Weil, supra note 94, at 427.
RESTATEMENT, supra note 6, §102 cmt. k; THIRLWAY, supra note 83, at 110; Charlesworth, supra note 38, at 4;
cf. CASSESE, supra note 83, at 179.
MERON, supra note 2, at 61.
E.g., Human Rights Commission, Report of the Special Rapporteur on Torture and Other Cruel, Inhuman
and Degrading Treatment or Punishment, UN Doc. E/CN.4/1988/17, at 23 (1988); see also SCHACHTER, supra
note 15, at 338; D’Amato, Response, supra note 31, at 469.
Nicaragua, 1986 ICJ REP. at 98, para. 186; see also ANTONIO CASSESE, VIOLENCE AND LAW IN THE MODERN AGE
35–39 (S. J. K. Greenleaves trans., 1988); HENKIN, supra note 118, at 70; MERON, supra note 2, at 60 (citing
Jonathan I. Charney, The Power of the Executive Branch of the United States Government to Violate Customary International
Law, 80 AJIL 913, 916 (1986)); Charlesworth, supra note 38, at 21.
784 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
On the other hand, jus cogens norms do not seem to require any conformity of practice,
which means they could exist even without any eligible interpretations of practice for sup-
port. Weil has thus criticized the theory of jus cogens for forcing states “to accept the super-
normativity of rules they were perhaps not even prepared to recognize as ordinary norms.”258
Rules of jus cogens also do not admit of persistent objectors, which undermines the con-
sensual basis of custom. For example, South Africa was not recognized as a persistent ob-
jector to the prohibition on apartheid.259 Finally, the concept of jus cogens seems to discount
the importance of contrary practice, treaties, and declarations unless they also form a rule
of jus cogens. This factor denies the possibility that conflicting practice could form a new cus-
tom, no matter how widespread the practice, and seems to assume an answer to the inter-
pretive process before it is applied, rather than using the process to determine whether a
custom exists. Thus, jus cogens norms are more static and less open to change than other
types of custom.
The Fluid Nature of Custom
Customs can generally change and harden over time because custom is a fluid source of
law. The content of custom is not fixed; it can develop and change in light of new circum-
stances. The formation and modification of custom is an uncertain process because inter-
national law lacks an authoritative guide as to the amount, duration, frequency, and repe-
tition of state practice required to develop or change a custom.260 Using Hart’s terminology,
custom would be a primitive source of law because it lacks clear rules of change. Instead,
custom develops through a “slow process of growth, whereby courses of conduct once thought
optional become first habitual or usual, and then obligatory, and the converse process of
decay, when deviations, once severely dealt with, are first tolerated and then pass unnoticed.”261
One reason for the difficulty of identifying the formation and change of custom is the
radical decentralization of the international system. States are both legislators and subjects
of international law, which explains why D’Amato argues that every breach of a customary
law contains the seed for a new legality.262 In one sense, the action is a breach because the
state is judged as a subject of international law; in another sense, the action is a seed for a
new law because the state acts as a legislator of international law.263 However, whether a
custom develops or changes depends not only on the actions of some states but also on the
reactions of other states. This is because states are also both legislators and enforcers of
international law. Thus, a breach will effectively repeal or modify an existing custom only
if other states emulate the breach or acquiesce in its legality.264
The reflective interpretive approach is useful in explaining the fluid nature of custom.
Just as customs can develop and change in light of new circumstances, so a reflective equi-
librium is not stable because the best explanation of practice and principles must be reevalu-
ated in light of new state practice, opinio juris, and moral considerations.265 The best balance
between practice and principles must be regularly reassessed. Consequently, conflicting data
should never be discounted as irrelevant because they may significantly affect future inter-
pretations. What is now an exception to, or a breach of, an accepted rule may later become
integral to the explanation of a new general rule. According to D’Amato:
Weil, supra note 94, at 427.
BYERS, supra note 6, at 183; Jiménez de Aréchaga, supra note 3, at 27.
BYERS, supra note 6, at 156–62; D’AMATO, supra note 6, at 56–66.
HART, supra note 47, at 90.
D’AMATO, supra note 6, at 97–98.
Weisburd, Customary IL, supra note 31, at 30–31.
North Sea Continental Shelf (FRG/Den.; FRG/Neth.), 1969 ICJ REP. 3, 44 (Feb. 20); id. at 230–31 (Lachs,
J., dissenting); Akehurst, supra note 7, at 37; Weisburd, supra note 208, at 107.
RAWLS, supra note 46, at 20.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 785
When a state violates an existing rule of customary international law, it undoubtedly is
“guilty” of an illegal act, but the illegal act itself becomes a disconfirmatory instance of
the underlying rule. The next state will find it somewhat easier to disobey the rule, until
eventually a new line of conduct will replace the original rule by a new rule.266
The number of disconfirmatory acts that are required before the breach of an old rule will
constitute the basis for a new rule depends on the extent of previous practice and the
importance of the moral principles involved. Moral customs, and in particular jus cogens
norms, are unlikely to be undermined by contrary practice. Furthermore, well-established
customs will demonstrate relative resistance to change because new state practice or opinio
juris must be weighed against a wealth of previous contrary practice.267 However, a custom
can change quickly in the face of very strong state practice or opinio juris, particularly if the
rule was uncertain or still developing.268 Recent practice may also carry proportionately
greater weight than past practice in determining the present or future state of custom.
Customs can develop or change in light of the recognition of new moral considerations in
international law. For example, the customary prohibition against genocide stemmed from
the recognition of human rights as a substantive aim of international law after World War
II. Likewise, currently nonbinding aspirations may harden into legally binding custom in the
future. For example, D’Amato and Sudhir Chopra have argued that whales may have an
emerging right to life under customary international law.269 Thus, the content of custom can
change in view of new practice and principles in international law.
The fluidity of custom is demonstrated by the present debate over whether NATO’s
intervention in Kosovo has formed the basis for an emerging customary right to unilateral
humanitarian intervention. In the Nicaragua case, the Court found a general customary pro-
hibition on intervention in other states but held that “[r]eliance by a State on a novel right
or an unprecedented exception to the principle might, if shared in principle by other States,
tend towards a modification of customary international law.”270 Whether states have suc-
cessfully created an exception to the general custom of nonintervention will depend in part
on whether they “justified their conduct by reference to a new right of intervention or a new
exception to the principle of its prohibition.”271 It will also depend on whether the action
provokes protest by other states or is emulated or met with acquiescence. However, if a state
prima facie breaches a custom but “defends its conduct by appealing to exceptions or justi-
fication contained within the rule itself, then whether or not the State’s conduct is in fact
justifiable on that basis, the significance of that attitude is to confirm rather than to weaken
Most commentators have concluded that NATO’s intervention in Kosovo was illegal under
existing international law.273 For example, Charney finds it indisputable that NATO’s inter-
vention violated the United Nations Charter and international law.274 However, the prece-
dential status of the intervention is more contentious. Simma argues that NATO’s inter-
vention should not become a precedent because of the exceptional factual circumstances
in Kosovo, the insistence by participating states such as France and Germany that the
intervention did not form a precedent, and the detrimental impact of allowing an isolated
D’AMATO, supra note 6, at 97.
BYERS, supra note 6, at 157–59.
Bin Cheng, in INTERNATIONAL LAW, supra note 19, at 249.
Anthony A. D’Amato & Sudhir K. Chopra, Whales: Their Emerging Right to Life, 85 AJIL 21 (1991).
Nicaragua, 1986 ICJ REP. at 109, para. 207.
Id. at 98, para. 186.
E.g., Charney, supra note 124, at 834; Louis Henkin, Kosovo and the Law of “Humanitarian Intervention,” 93 AJIL
824, 824–25 (1999).
Charney, supra note 124, at 834.
786 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
breach of collective security to become a general rule for unilateral action.275 By contrast,
Antonio Cassese argues that NATO’s action may support an emerging custom allowing the
use of forcible countermeasures to impede a state from committing large-scale atrocities
within its own territory, in circumstances where the Security Council is incapable of respond-
ing to the crisis.276 I submit that these conflicting arguments can be understood by employ-
ing a reflective interpretive approach.
To determine eligible interpretations of fit, one must primarily consider the state practice
respecting unilateral humanitarian intervention. Opponents of the norm focus on the lack
of explicit authorization of NATO’s action by the Security Council and the likelihood that
China or Russia would have vetoed any resolution supporting intervention. The intervention
also provoked significant protests by China, India, Namibia, Belarus, and the Russian Feder-
ation,277 which undermines its ability to form a custom. Prominent members of NATO,
including France and Germany, insisted that the intervention did not constitute a precedent
for a general right to unilateral humanitarian intervention.278 In pleadings for the case
brought by the Federal Republic of Yugoslavia against ten NATO members in 1999, only
Belgium expressly mentioned humanitarian intervention as a possible legal justification for
action.279 Thus, the intervening states did not attempt to justify their intervention on the
basis of a new exception to the general prohibition on intervention.280
States have also repeatedly abstained from invoking the right of humanitarian interven-
tion where there were grounds for doing so, including India’s intervention in East Pakistan
in 1971, Vietnam’s invasion of Kampuchea in 1978, and Tanzania’s intervention in Uganda
in 1979.281 These states justified their interventions on other grounds, such as self-defense,
which suggests that they did not believe that unilateral humanitarian intervention was ac-
cepted in international law.282 However, states will generally resort to factual or legal excep-
tions rather than openly admit they have breached a law, even if they wish to change it.283
The absence of intervention in potentially analogous situations involving, for example, the
Kurds in Turkey, the Chechens in Russia, and the Tibetans in China may also suggest that
states did not believe that unilateral humanitarian intervention was permitted under inter-
By contrast, proponents of an emerging norm argue that NATO had authority to act
because the Security Council had defined the situation in Kosovo as a “threat to peace and
security in the region.”284 The lack of condemnation of NATO’s intervention by the Security
Council and most other states also indicates a general acquiescence in its legality. For ex-
ample, a draft resolution condemning NATO’s use of force was rejected by a vote of 12 to
Bruno Simma, NATO, the UN and the Use of Force: Legal Aspects, 10 EUR. J. INT’L L. 1 (1999), at <http://
www.ejil.org/journal/index.html> (visited Oct. 1, 2001).
Cassese, supra note 124.
The draft resolution condemning NATO’s use of force, UN Doc. S/1999/328, was sponsored by Belarus, India,
and the Russian Federation and supported by China, Namibia, and the Russian Federation in the Security Council.
For example, German Foreign Minister Kinkel stated, “The decision of NATO [on air strikes against the Fed-
eral Republic of Yugoslavia] must not become a precedent.” Deutscher Bundestag, Plenarprotokoll 13/248, at
23,129 (Oct. 16, 1998), quoted in Simma, supra note 275.
Some states focused exclusively on the preliminary issue of jurisdiction; others, including Germany, argued
that the intervention represented a justifiable exception to the normal rules. The United States focused on the hu-
manitarian catastrophe, the acute threat to security of neighboring states, the serious violation of humanitarian law,
and the resolutions of the Security Council but did not expressly argue for a right to unilateral humanitarian inter-
vention. Oral pleadings (Yugo. v. U.S. et al.), 1999 ICJ Pleadings (Legality of Use of Force), at <http://www.icj-cij.org>.
Charney, supra note 124, at 836–37.
C. Gray, After the Ceasefire: Iraq, the Security Council and the Use of Force, 1994 BRIT. Y.B. INT’L L. 135, 162; Dino
Kritsiotis, Reappraising Policy Objections to Humanitarian Intervention, 19 MICH. J. INT’L L. 1005, 1014 (1998).
Charney, supra note 124, at 836–39.
CASSESE, supra note 257, at 35–39; MERON, supra note 2, at 60 (citing Charney, supra note 257); HENKIN, supra
note 118, at 70; Charlesworth, supra note 38, at 21.
SC Res. 1203 (Oct. 24, 1998); see also SC Res. 1199 (Sept. 23, 1998); SC Res. 1160 (Mar. 31, 1998).
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 787
3 in the Security Council,285 and no state requested an immediate meeting of the General
Assembly to condemn the intervention. Precedents can also be found for humanitarian
intervention without explicit authorization by the Security Council, such as the action by the
United States, France, and Great Britain in Iraq in 1991. Moreover, the interventions in Iraq
and Kosovo provide more modern, and thus potentially more weighty, precedents than the
counterexamples from the 1970s. Contrary examples of states seeking Security Council ap-
proval before intervening in other situations could represent a politically prudent course
of action, rather than a necessary element of legality. Further, the failure of states to inter-
vene in previous humanitarian crises represents discretionary nonaction, rather than obliga-
tory negative practice or acquiescence, because a custom allowing intervention is a permis-
sive right rather than an obligatory duty.286
The range of eligible interpretations brings the dimension of substance into play to help
determine the best interpretation in light of the substantive aims of international law. Uni-
lateral humanitarian intervention involves the conflicting substantive considerations of state
sovereignty (which would support a duty of nonintervention) and humanitarian aims (which
would support a right to intervention). Prohibitions on intervention and the use of force
appear to be peremptory norms of international law.287 Some humanitarian considerations,
such as the prohibition on genocide, have also achieved jus cogens status.288 The dimension
of substance must include consideration of how the substantive aims of international law
should be prioritized when they conflict. While state sovereignty has traditionally been ac-
corded primary importance in international law, arguably the international community is
increasingly recognizing exceptions to this principle in extreme humanitarian crises.289 Thus,
which principle will prevail when the two conflict depends on the extremity of the humani-
tarian crisis in question.
The dimension of substance also requires consideration of the procedural normativity of
unilateral intervention. Critics of unilateral intervention argue that it is open to abuse and
likely to result in selective and self-interested action.290 Proponents argue that selective
application is inherent in the concept of rights rather than duties, that selective intervention
may be better than no intervention, and that it is unrealistic to expect states to intervene
only when they are completely disinterested.291 Some of the objections may also be met by
clearly identifying the conditions for intervention so as to provide safeguards against abuse,
rather than creating a blanket prohibition on intervention. Further, the alternatives to uni-
lateral intervention, which are no intervention and collective intervention,292 should be
taken into account. Collective intervention, approved by the Security Council, is arguably
less partisan and less open to abuse than unilateral intervention. However, the legitimacy
of the Security Council itself is questionable because it comprises only fifteen states and the
five permanent members all have the power of veto. Thus, collective action may not enjoy
the advantages of procedural normativity. Collective action may also not be possible even
in the face of extreme violations of human rights such as genocide. In such cases, unilateral
intervention may be justified on the basis that the importance of the substantive moral issues
involved outweighs deficiencies in procedural normativity.293
See supra note 277.
This distinction is made by D’AMATO, supra note 6, at 61–63, though not with respect to unilateral humani-
BROWNLIE, supra note 6, at 515; CASSESE, supra note 83, at 147.
BROWNLIE, supra note 6, at 515; Shaw, supra note 251.
Kritsiotis, supra note 281, at 1040–46.
Christine M. Chinkin, Kosovo: A “Good” or “Bad” War? 93 AJIL 841, 847 (1999).
Kritsiotis, supra note 281, at 1020–39.
Henkin, supra note 273, at 824–25.
Ruth Wedgwood, NATO’s Campaign in Yugoslavia, 93 AJIL 828, 833 (1999).
788 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
In applying the reflective equilibrium, we need to consider which explanation most coher-
ently explains the competing conceptions of practice and principles outlined above. There
are two eligible interpretations of custom, one prohibiting unilateral humanitarian inter-
vention and one permitting it. Since the first interpretation more adequately explains the
raw data of practice to date, the second interpretation would only be preferred if supported
by a strong dimension of substance. Substantively, the principles of state sovereignty and
nonintervention generally prevail, except in cases of extreme human rights violations such
as genocide. From a procedural perspective, collective intervention is usually preferable to
unilateral intervention, though the latter may be justified if the Security Council is paralyzed
by the veto power. Thus, the dimension of substance provides equivocal support for both
interpretations, which varies according to the circumstances of the case.
The present state of practice and principles probably leads to the conclusion that uni-
lateral humanitarian intervention is not currently recognized as an exception to the prin-
ciple of nonintervention. However, a narrow right to unilateral humanitarian intervention
may be emerging in exceptional circumstances where there are gross violations of human
rights, peaceful avenues for settling the dispute have been exhausted, the veto power has
rendered the Security Council incapable of taking coercive action, the action is undertaken
by a group of states with the support (or at least the nonopposition) of the majority of states,
and the armed force is used exclusively to stop the atrocities.294 Whether this emerging
norm is transformed into a binding custom will depend on future developments in the prac-
tice and principles of international law. Thus, the reflective interpretive approach can be
used to demonstrate the fluid nature of custom as a source of law.
V. THE ADVANTAGES OF A REFLECTIVE INTERPRETIVE APPROACH
The contemporary interest in custom as a source of international law poses a challenge
to articulate a coherent theory of custom. This is a difficult exercise because the traditional
and modern approaches to custom appear to be opposed, with traditional custom empha-
sizing state practice and modern custom emphasizing opinio juris. The divergence between
the descriptive and normative approaches of traditional and modern custom causes prob-
lems because the tests and justifications for traditional custom do not apply to modern
custom and vice versa. Instead of advocating the rival merits and legitimacy of either ap-
proach, this article has sought to build an enriched theoretical account of custom that
analyzes the competing justifications for traditional and modern custom and accommodates
both approaches in a consistent interpretive theory.
This reconciliation provides a methodology for assessing asserted customs. After gathering
evidence of state practice and opinio juris, one must apply the threshold criterion of fit to
determine if there are any eligible interpretations that adequately explain the raw data of
practice. Fit provides continuity and descriptive accuracy, so that state practice will assume
primary importance at this stage. However, state practice is open to interpretation and should
include intrastate action and inaction, not just interstate interaction and acquiescence.
Some articulation of legality is needed to differentiate between legal custom and social
practice. If there is no eligible interpretation, then there is no custom. If there is one eli-
gible interpretation, then the custom is clear. If there are multiple eligible interpretations,
then one must weigh the dimensions of fit and substance to determine the best interpretation.
Strong statements of opinio juris become relevant at this third stage because they represent
normative considerations about what the law should be. The best interpretation is the one
that most coherently explains fit and substance, which varies according to the facilitative or
Cassese, supra note 124; Charney, supra note 124, at 836–39; Wedgwood, supra note 293, at 828.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 789
moral content of the custom involved. Primarily facilitative customs do not involve strong
substantive considerations and thus will be determined principally by fit. Primarily moral
customs give rise to strong procedural and substantive normative considerations, which must
be balanced against deficiencies in fit. Custom is also a fluid source of law, which causes the
point of equilibrium to vary over time in light of new state practice, opinio juris, and moral
considerations. Consequently, the reflective interpretive approach results in a more sophis-
ticated understanding of fit and substance and constitutes a more nuanced method for
reconciling them than the sliding scale in various ways.
First, the reflective interpretive approach recognizes that the role of the dimension of fit
is to provide descriptive accuracy. Questions of fit and descriptive accuracy are both back-
ward looking because they focus on whether a custom is supported by past practice. Tasioulas
claims that strong statements of opinio juris can form the basis of eligible interpretations at
the dimension of fit. Similarly, Michael Akehurst has argued that state practice should in-
clude paper practice in the form of statements, declarations, and resolutions.295 I have dem-
onstrated that this approach is not feasible because statements often fuse lex lata and lex
ferenda and thus lack descriptive accuracy. However, one can still form eligible interpreta-
tions of traditional and modern custom by considering the open-textured nature of practice.
State practice should also include consideration of intrastate action (not just interstate
interaction), obligations being observed (not just obligations being breached), and reasons
for a lack of protest over breaches (other than acquiescence in the legality of those breaches).
These forms of practice reflect the changing subject matter of international law to include
intrastate issues and are descriptively accurate because they focus on action and inaction.
Second, I have outlined a dimension of substance that embodies both substantive and
procedural normativity. Kirgis refers to strong moral issues but does not provide a theory
for determining them. Tasioulas deliberately limits his discussion of the substantive aims of
international law to coexistence and cooperation. By contrast, my approach provides a more
expansive understanding of the substantive aims of international law, which includes recog-
nized moral aims (such as the protection of human rights) that do not necessarily affect
coexistence and cooperation. Substantive aims are frequently criticized as being subjective
value judgments that serve as a vehicle for normative chauvinism. For this reason, I have
defined moral issues as commonly held subjective values about right and wrong that have
been adopted by a representative majority of states in treaties and declarations. This ap-
proach has several advantages. Focusing on commonly held, or intersubjective, values avoids
the need to consider whether moral values can be objectively determined and it explains
why these values can change over time. It also denotes an agreed set of values rather than
requiring interpreters to determine what they believe the substantive aims of international
law should be. It builds the concept of procedural normativity into the dimension of sub-
stance because these values have been accepted by a majority of states, which helps prevent
accusations of Western ideological bias. Finally, while these statements may include lex lata
and lex ferenda, the dimension of fit already provides a threshold test to determine if a custom
is adequately supported by practice.
Third, I have suggested a more nuanced approach to balancing fit and substance than the
crude sliding scale. Finding traditional custom on the strength of state practice and fit alone
allows it to become an apology for state power.296 Similarly, deducing modern custom purely
from opinio juris and substance can create utopian laws that cannot regulate reality. Thus,
the sliding scale can allow one element completely to outweigh the other. While a reflective
equilibrium will lean toward the stronger value, it avoids extremes of apology and utopia.
Akehurst, supra note 7, at 53.
KOSKENNIEMI, supra note 45, at 2.
790 THE AMERICAN JOURNAL OF INTERNATIONAL LAW [Vol. 95:757
The strength of fit and state practice in traditional customs must still be balanced against
their substantive deficiencies, such as a lack of procedural normativity. Similarly, modern
customs must be supported by state practice because they must pass the threshold of fit and
deficiencies in their fit may still outweigh their moral content.297 A lower standard of prac-
tice may be tolerated for customs with a strong moral content because violations of ideal
standards are expected.298 However, while occasional breaches may not nullify their legal
character, massive, grave, and persistent violations will.299 The only exceptions are jus cogens
norms, which by definition cannot be undermined by contrary practice unless that practice
creates another rule of jus cogens.300
Fourth, Kirgis and Tasioulas argue that the more morally distasteful an activity, the more
readily the Court will substitute opinio juris for state practice and vice versa. However, I have
explained that substantive considerations apply asymmetrically to traditional and modern
customs because of their facilitative and moral content. This reasoning results in a better
explanation of when, and on what basis, traditional and modern customs will be formed. It
also means that, to the extent that the reflective equilibrium can still be criticized for apology
and utopia, these criticisms are less compelling because they apply to facilitative and moral
customs, respectively. Criticizing a facilitative custom for being an exercise of power is not
problematic because it does not concern substantive moral issues. Instead, these customs
are akin to domestic traffic rules. If developing states wish to challenge a traditional custom,
they can enact declarations in international fora, as they did on whether there should be
an international minimum standard of compensation for expropriation301 or a national treat-
ment standard.302 Similarly, modern customs that set up ideal standards about moral issues
are expected to be somewhat utopian. Meron argues that the international community is
willing to accept gradual or partial compliance as fulfilling the requirements for forming
moral customs.303 Alasdair MacIntyre argues that the charge of utopianism is made by “the
deliberately shortsighted who congratulate themselves upon the limits of their vision.”304
While hard law that is always enforced may be preferable to soft law, the choice in areas such
as human rights is often between soft law and no law.305 Giving these aspirations some legal
force may be preferable to giving them no legal status, because they can be enforced in
extreme situations such as apartheid in South Africa306 and their legal status may harden
Finally, the reflective equilibrium can be used to explain the fluid nature of customary
international law. The sliding scale assumes that state practice and opinio juris are fixed and
irreconcilable quantities that must be traded off against each other to form eligible inter-
pretations of custom. However, I have demonstrated that state practice is open textured and
capable of being interpreted in various ways. For example, contrary state practice can be
analyzed as a breach of an old rule or as the seed of a new rule. Finding the best inter-
MERON, supra note 2, at 44–45.
Id. at 44; Schachter, supra note 110, at 539; Schachter, supra note 81, at 735.
MERON, supra note 2, at 58.
Vienna Convention on the Law of Treaties, supra note 83, Art. 53.
Supported by Resolution on Permanent Sovereignty over Natural Resources, GA Res. 1803, UN GAOR, 17th
Sess., Supp. No. 17, at 15, UN Doc. A/5217 (1962). See also BROWNLIE, supra note 6, at 527–29, 535–38.
Supported by Charter of Economic Rights and Duties of States, GA Res. 3281, UN GAOR, 29th Sess., Supp.
No. 30, at 50, UN Doc. A/9030 (1974). See also BROWNLIE, supra note 6, at 526–27, 538.
MERON, supra note 2, at 44.
ALASDAIR MACINTYRE, THREE RIVAL VERSIONS OF MORAL ENQUIRY: ENCYCLOPAEDIA, GENEALOGY, AND TRADI-
TION 234 (1990); see also Tasioulas, supra note 42, at 127.
Condorelli, supra note 66, at 81; Pellet, supra note 68, at 47.
D’AMATO, supra note 6, at 89; Henkin, supra note 25, at 39, 42.
Chinkin, supra note 221, at 857–58.
2001] TRADITIONAL AND MODERN APPROACHES TO CUSTOMARY INTERNATIONAL LAW 791
pretation of practice and principles requires one to determine the most coherent expla-
nation of state practice and opinio juris, rather than simply giving preference to one and dis-
counting the other. For this reason, conflicting state practice should never be discounted
as irrelevant to interpretation, because it may contain the seed for a new custom. It also
clarifies how customs change over time in light of new state practice, opinio juris, and moral
The reflective interpretive approach rearticulates the theoretical foundations of custom
in a more principled and flexible fashion. Instead of debating the relative merits and legiti-
macy of traditional and modern custom, this interpretive theory seeks to justify and recon-
cile the two approaches and, in so doing, offers a coherent theory of custom that helps to
defend its integrity as a source of international law.