VOLUME 6 EJLS ISSUE 1
126
European Journal of Legal Studies
Title: Evolutive Treaty Interpretation: Legality, Semantics and Distinctions
Author(s): Sondre Torp Helmersen
Source: European Journal of Legal Studies, Volume 6, Issue 1 (Spring/Summer 2013), p
127-148
Abstract:
According to the ICJ, ‘generic’ terms in long-term treaties were presumably intended to be interpreted
evolutively. This ‘general rule’ on evolutive interpretation appears simple, but leaves unanswered
questions. Moreover, linguistic analyses show that the ICJ is inconsistent in its definition of ‘generic’,
and that evolutive interpretations are unsuited to solving ambiguity (as opposed to vagueness). There
is, moreover, a tendency in the literature to confuse or conflate evolutive interpretation with the
doctrine of intertemporality or the VCLT Article 31.3.c—these are three distinct concepts.
VOLUME 6 EJLS ISSUE 1
127
EVOLUTIVE TREATY INTERPRETATION: LEGALITY, SEMANTICS AND
DISTINCTIONS
Sondre Torp Helmersen*
TABLE OF CONTENTS
1. INTRODUCTION ..................................................................................................................128
2. EVOLUTIVE INTERPRETATION IN CUSTOMARY INTERNATIONAL
LAW ................................................................................................................................130
2.1 Generally ...........................................................................................................................130
2.2 Party Intention.................................................................................................................134
2.3 A ‘General Rule’...............................................................................................................135
2.4 A Taxonomy of Terms...................................................................................................139
3. EVOLUTIVE INTERPRETATION AND SEMANTICS...........................................139
3.1 ‘Generic Terms’................................................................................................................139
3.2 Ambiguity and Vagueness.............................................................................................142
4. EVOLUTIVE INTERPRETATION DISTINGUISHED............................................144
4.1 The Doctrine of Intertemporality................................................................................144
4.2 The VCLT Article 31.3.c ...............................................................................................146
5. CONCLUSION.........................................................................................................................148
* University lecturer, University of Oslo. E-mail: acousticbandits@gmail.com. I thank Geir Ulfstein, Ivar
Alvik and Cecilie Christin Kverme for useful comments.
VOLUME 6 EJLS ISSUE 1
128
1. INTRODUCTION
The topic for this article is evolutive interpretation of treaties.1 The article aims both to
clear up some confusion regarding the concept and its place in customary international law,
and to contribute to our analytical understanding of it.
Apart from this introductory section, the article has four sections. The next section (2) aims
to clarify the place of evolutive interpretation in customary international law, primarily
through a fresh look at practice from the Permanent Court of International Justice and
International Court of Justice (hereinafter the ICJ). The analysis ends in a taxonomy of
three types of terms that must be treated differently. Section 3 explores the semantics of
evolutive interpretation. The penultimate section (4) seeks to distinguish evolutive
interpretation from similar or related concepts, by showing that it is neither the part of the
‘doctrine of intertemporality’ nor of the VCLT2 Article 31.3.c. Section 5 is a conclusion.
First, a definition: An evolutive interpretation is an interpretation where a term is given a
meaning that changes over time.3 As with all interpretations, the evolutive interpretation of a
term is distinct from its application.4 A term that is applied to new circumstances while its
meaning remains constant is not being interpreted evolutively. This also means that a
change of mind is not an evolutive interpretation. In interpreting a term that is open to
multiple interpretations, a court may choose one interpretation in one case, and then change
its mind and prefer another in a later case. This way, the term’s meaning can be said to have
‘changed’ over time. However, if the change is not prompted by an evolution intended by
the parties, the interpretation is not evolutive. The term has not evolved; only the opinion of
the court.
Normally, evolutive interpretations are made possible by evolution in the linguistic meaning
of the interpreted term itself, independent of the interpretation.5 However, a term does not
have to evolve linguistically to be interpreted evolutively. Treaty interpretation is an
1 ‘Evolutive interpretation’ is synonymous with ‘dynamic interpretation’ (Malgosia Fitzmaurice, ‘The Practical
Working of the Law of Treaties’ in Malcolm D Evans (ed), International Law (3rd edn, OUP 2010) 188) and
with the European Court of Human Rights’ (ECtHR) ‘living instrument’ doctrine (George Letsas, A Theory of
Interpretation of the European Convention on Human Rights (OUP 2007) 65). Pierre-Marie Dupuy, 'Evolutionary
Interpretation of Treaties: Beyond Memory and Prophecy' in Enzo Cannizzaro (ed), The Law of Treaties
Beyond the Vienna Convention (OUP 2011) 123 uses the term ‘evolutionary interpretation’, while Paul
Tavernier, ‘Relevance of the Inter-temporal Law’, in James Crawford, Alain Pellet and Simon Olleson (eds),
The Law of International Responsibility (OUP 2006) 400 also calls it ‘progressive’ interpretation. ‘Evolutive’ and
‘evolutively’ will be used in this article, the former as an adjective, the latter as an adverb.
2 Vienna Convention on the Law of Treaties (adopted 23 May 1969, entered into force 27 January 1980) 1155
UNTS 331 (VCLT).
3 See eg the Joint Dissenting Opinion in Feldbrugge v. The Netherlands, ECHR (1986) Series A, no 99, 266, para
24.
4 The two are about ‘determining the meaning of a text’, and the consequences of that determination,
respectively, see Richard Gardiner, Treaty Interpretation (OUP 2008) 76, quoting the commentary of the
‘Harvard Draft Convention on the Law of Treaties’ (in (1935) 29 Supplement to the AJIL 653, at 938). The
distinction is perhaps sharper in theory than in practice; see Richard Gardiner, Treaty Interpretation (OUP
2008) 29 and Separate Opinion of Judge Shahabuddeen in Applicability of the Obligation to Arbitrate under Section
21 of the United Nations Headquarters Agreement of 26 June 1947 (Advisory Opinion) [1988] ICJ Rep 12, 59.
5 This can be called ‘evolving terms’.
VOLUME 6 EJLS ISSUE 1
129
inherently subjective process; if the parties intend a term to evolve, it is irrelevant whether
it evolves linguistically as well.6
The opposite of an evolutive interpretation can be called a ‘static’ interpretation (ie an
interpretation where terms do not change their meaning over time).
When analysing the process of treaty interpretation, it is pertinent to distinguish between
factors that may be invoked when interpreting treaties, methods of treaty interpretation,
and the potential results of treaty interpretation. Factors are arguments used in the
interpretive process.7 They include the elements mentioned in the VCLT Article 31-33, eg
ordinary meaning, context, object and purpose, subsequenct practice, and so on.8 Method is a
catch-all term for the approach used when interpreting treaties.9 Customary international
law prescribes a single, unified methodology,10 of which directions on when and how to
interpret evolutively is part. When a treaty term is interpreted evolutively, its content will
evolve over time. This is a result of the interpretation. Other interpretive results include
static interpretations, extensive or restrictive interpretations, and effective interpretations.
Evolutive interpretations can be extensive, restrictive and/or effective, but are not
inherently so.
Beyond the fact that all legal texts require interpretation,11 the evolutive variety has
considerable practical importance. Treaties, once concluded, tend to remain (formally)
static. Amendment is always possible, but can be difficult in practice.12 At the same time, the
reality that treaties operate in is in constant flux. Economic, political, cultural, and
6 Georg Ress, ‘The Interpretation of the Charter’ in Bruno Simma (ed), The Charter of the United Nations: A
Commentary (2nd edn, OUP 2002) 23 defines evolutive interpretation as ‘based on the linguistic usage of the
term at the time of interpretation’. That is only fitting for terms that evolve linguistically.
7 Called ‘interpretive arguments’ by Ole Kristian Fauchald, ‘The Legal Reasoning of ICSID Tribunals – An
Empirical Analysis’ (2008) 19 EJIL301, 301-302 and 308, and ‘means’ by Ulf Linderfalk, 'Is the Hierarchical
Structure of Articles 31 and 32 of the Vienna Convention Real or Not? Interpreting the Rules of
Interpretation', in (2007) 54 Netherlands Intl L Rev 133, 135, and Mark E. Villiger, Commentary on the 1969
Vienna Convention on the Law of Treaties (Brill 2009) 435. The VCLT uses the word ‘means’ in art 32.
8 As noted in note 24 below, the VCLT art 31-33 prescribes ‘principles’ (as opposed to rules) of treaty
interpretation. The articles thus mention interpretive factors, and lay down principles regarding whether,
when, and how these factors should be employed.
9 Alexander Orakhelashvili, The Interpretation of Acts and Rules in Public International Law (OUP 2008) 309
apparently uses the term ‘methods’ about ‘text, context, preparatory work, subsequent practice’, ie what was
defined as ‘factors’ above. However, he also calls ‘[t]ext, context and object-and-purpose’ ‘factors’ at 310-311.
10 A traditional view has been that there are three ‘methods’ or ‘schools’ of interpretation; the ‘textual’, the
‘teleological’, and the ‘purposive’; see eg ILC, Yearbook of the International Law Commission 1966 Volume II
(United Nations 1966) 218 para 2; Jeff Waincymer, WTO Litigation: Procedural Aspects of Formal Dispute
Settlement (Cameron May 2002) 397-398; Fauchald (n 7) 315. Martin Dixon, Textbook on International Law (6th
edn, OUP 2007) 71-72 adds ‘the principle of effectiveness’, while Villiger (n 7) 421-422 adds the ‘historical’ and
‘logical’ methods. The VCLT arts 31-33 nonetheless prescribe a single, unified methodology where text
(‘textual’), good faith, and object and purpose (part of ‘teleology’) are relevant factors when ascertaining the
parties’ intentions (‘purposive’ interpretation); see Robert Jennings and Arthur Watts (eds), Oppenheim’s
International Law – Volume 1: Peace, Parts 2 to 4 (Longman 1992) 1272; Orakhelashvili (n 9) 310; Gardiner (n 4)
9-10; Villiger (n 7) 435; Isabelle Van Damme, Treaty Interpretation by the WTO Appellate Body (OUP 2009) 35-
36.
11 Due to the inherent limits in language and the unpredictability and multitude of reality; see HLA Hart, The
Concept of Law (2nd edn, OUP 1994) 126.
12 Anthony Aust, Modern Treaty Law and Practice (2nd edn, CUP 2007) 262.
VOLUME 6 EJLS ISSUE 1
130
technological realities change. In many (if not most) fields, law must be flexible if it is to
remain relevant and effective. Flexibility, in turn, has to be constantly balanced against
stability, which is an important aspect of the rule of law.13
The problems that evolutive interpretation may alleviate are not restricted to international
law; they apply universally to all legal systems. The concept is thus well known in domestic
law.14 There is nonetheless a difference between domestic and international law in that the
legislative branches of most domestic governments are considerably more flexible and
responsive in their legislating than States are in drafting and revising treaties. Since one
function of evolutive interpretations is to allieviate the need for new rules to address present
concerns, the need for (but not necessarily the prevalence of) evolutive interpretations is
comparatively greater in the international sphere.
2. EVOLUTIVE INTERPRETATION IN CUSTOMARY INTERNATIONAL LAW
2.1 Generally
The goal is this section is to determine the place of evolutive interpretation in customary
international law. This necessitates answering two questions: First, when (ie on what
conditions) should terms be interpreted evolutively? Second, how (ie by what benchmarks)
should the terms evolve?
Treaty interpretation is regulated by customary international law and (for its parties) the
VCLT. As per the ICJ Statute15 Article 38.1.b,16 customary international law is found by
examining state practice and establishing opinio juris.17 In practice, though, it is often
difficult to pin down the exact content of customary international law,18 especially in the
indeterminate,19 contested,20 and loosely regulated21 field of treaty interpretation. Therefore,
proxies are useful to ascertaining its content.
13 Markus Kotzur, ‘Intertemporal Law’, The Max Planck Encyclopedia of Public International Law (January 2013
edn) <www.mpepil.com> accessed 28 April 2013, para 4.
14 Søren C Prebensen, ‘Evolutive Interpretation of the European Convention of Human Rights’, in Paul
Mahoney and others (eds), Protection des droits de l’homme: la perspective européenne (Carl Heymanns Verlag
2000) 1126. William N Eskridge Jr, Dynamic Statutory Interpretation (Harvard Universty Press 1994) writes
about it from a US perspective. David Souter, ‘Harvard University’s 359th Commencement Address’ (2010)
124 Harvard L Rev 429 gives the example of the United States Constitution’s Equal Protection Clause being
interpreted evolutively by the US Supreme Court between Plessy v. Ferguson, 163 U.S. 537 (1896) and Brown v.
Board of Education, 347 U.S. 483 (1954).
15 Statute of the International Court of Justice (adopted 26 June 1945, entered into force 24 October 1945) 33
UNTS 993 (ICJ Statute).
16 Which, according to eg Jennings and Watts (eds) (n 10) 24; Ian Brownlie, Principles of International Law (7th
edn, OUP 2008) 5; Malcolm N Shaw, International Law (6th edn, OUP 2008) 70, reflects customary
international law.
17 North Sea Continental Shelf [1969] ICJ Reports 3, para 77.
18 Martti Koskenniemi, From Apology to Utopia: The Structure of International Legal Argument (2nd edn, CUP
2005) 396.
19 For example, Richard A Falk, The Status of Law in International Society (Princeton UP 1970) 372 has written
that ‘the interpretation of broad international agreement is operating in a largely indeterminate setting’.
20 Vaughan Lowe, International Law (OUP 2007) 73 notes that there are ‘debates over every step in the
reasoning process that leads from a treaty text to the conclusion concerning its effects in a concrete case’.
21 Reflected in the fact that interpretation is sometimes said to be ‘to some extent an art, not an exact science’,
VOLUME 6 EJLS ISSUE 1
131
With regards to treaty interpretation, the VCLT Articles 31-33 is a widely accepted
proxy.22 Its status as a proxy is not relevant to States that are parties to it, even though the
underlying customary law still binds them.23 Two other proxies are mentioned in the ICJ
Statute Article 38.1.d: ‘judicial decisions’ and ‘teachings of the most highly qualified
publicists’.
The VCLT Articles 31-33 prescribe principles of treaty interpretation,24 which permit
evolutive interpretation. Terms’ ‘ordinary meaning’ (Article 31.1) may change over time,
and the VCLT does not determine whether it is the ‘ordinary meaning’ at the time of a
treaty’s conclusion or at the time of its interpretation that shall prevail. ‘Good faith’ and
‘object and purpose’ (Article 31.1) may require that a term is interpreted evolutively, and
may affect how it evolves. A ‘subsequenct agreement’ (Article 31.3.a) may determine both
whether a term should evolve as well as how it should evolve. The same is the true for
‘subsequent practice’ (Article 31.3.b)25 and ‘relevant rules of international law’ (Article
31.3.c)26. Non-evolving terms may be given a ‘special meaning’ (Article 31.4) that
nonetheless evolves. ‘Preparatory works’ and circumstances of a treaty's conclusion (Article
32) may support an evolutive interpretation. Beyond permitting evolutive interpretations,
however, the VCLT Articles 31-33 provide limited guidance.
Evolutive interpretations are found in ‘decisions’ from various international tribunals. Since
first appearing in the Tyrer27 judgement, it has become a ‘key theme’ in the jurisprudence of
originally put forward in the ILC Draft Articles on the Law of Treaties with Commentaries, reproduced in
ILC (n 10) 218, and discussed critically in eg Koskenniemi (n 18) 340-341 and Panos Merkouris, ‘Introduction:
Interpretation Is A Science, Is An Art, Is A Science’ in Malgosia Fitzmaurice, Olufemi Elias and Panos
Merkouris (eds), Treaty Interpretation and the Vienna Convention on the Law of Treaties: 30 Years on (Martinus
Nijhoff 2010) 8-13.
22 See eg Territorial Dispute (Libyan Arab Jamahiriya/Chad) [1994] ICJ Reports 6, para 41; Oil Platforms
(Islamic Republic of Iran v. United States of America) [1996] ICJ Reports 803, para 23.
23 See Military and Paramilitary Activities in and against Nicaragua (Nicaragua v. United States of America) [1984]
ICJ Reports 392, para 73 for the general point that customary international law can still bind the parties to a
treaty codifying it.
24 As opposed to ‘rules’. Principles ‘do not set out legal consequences that follow automatically’, whereas rules
are ‘applicable in an all-or-nothing fashion’ (Ronald Dworkin, Taking Rights Seriously (Duckworth 1978) 24-
25). Even though the VCLT art 31 itself uses the word ‘rule’, the VCLT arts are sufficiently flexible to make
‘principles’ (or even ‘means’) a more appropriate word (see Gardiner (n 4) 36-38; Van Damme (n 10) 35). More
generally, the notion of strict ‘rules’ of interpretation is theoretically problematic; as George Letsas,
‘Strasbourg’s Interpretive Ethic: Lessons for the International Lawyer’ (2010) 21 EJIL 509, 534 observes, ‘no
treaty can tell us how to interpret treaties’.
25 The potential interaction between evolutive interpretation and subsequent practice has been noted in ILC,
‘Report on the Work of its Sixty-Third Session’ (26 April to 3 June and 4 July to 12 August 2011), UN Doc
A/66/10, at 283, which states that ‘[e]volutionary interpretation is a form of purpose-oriented interpretation
that is given direction by subsequent practice in a narrower and a wider sense (specific practice of states
parties, as well as other developments in international relations or society)’. This does not explicitly
distinguish between subsequent practice establishing whether or how terms shall interpreted evolutively. The
relationship between the two concepts is explained in detail by Julian Arato, ‘Subsequent Practice and
Evolutive Interpretation: Techniques of Treaty Interpretation over Time and Their Diverse Consequences’
(2010) 9 L and Practice of Intl Courts and Tribunals 443.
26 See s 4.2 below for a more detailed description of the relationship between evolutive interpretation and the
VCLT art 31.3.c.
27 Tyrer v. The United Kingdom (1978) Series A no 26, para 31.
VOLUME 6 EJLS ISSUE 1
132
the European Court of Human Rights (interpreting the ECHR28).29 It has also been used by
the European Court of Justice,30 the Inter-American Court of Human Rights,31 the UN
Human Rights Committee,32 the International Tribunal for the Law of the Sea,33 in at least
one arbitration,34 and in two reports from the WTO Appellate Body.35
The ICJ has used evolutive interpretations in the 1970 Namibia advisory opinion,36 the 1978
Aegean Sea judgement,37 and the 2009 Dispute Regarding Navigational and Related Rights
judgement.38 The concept may also have featured in three other cases: The first is
Nationality Decrees Issued in Tunis and Morocco39, where the Court held that whether a matter
is ‘solely within the domestic jurisdiction’ of a party is ‘essentially relative’ and depends
‘upon the development of international relations’.40 The second is Aegean Sea, in which the
Court called the term ‘rights’ a ‘generic term’, noting that it should ‘evolve in meaning’ in
accordance with ‘the development of international relations’.41 Finally, in GabcikovoNagymaros42,
the Court labelled certain treaty provisions ‘evolving’, found that ‘the Treaty
is not static, and is open to adapt to emerging norms of international law’, and that ‘current
standards’ of environmental protection should be taken into account.43
28 Convention for the Protection of Human Rights and Fundamental Freedoms (adopted 4 November 1950,
entered into force 3 September 1953) 213 UNTS 222 (ECHR).
29 Robin CA White and Clare Ovey, Jacobs, White & Ovey. The European Convention on Human Rights (5th edn,
OUP 2010) 64.
30 Case 283/81, Srl CILFIT and Lanificio di Gavardo SpA v Ministry of Health [1982] ECR 3415, para 20.
31 The Right to Information on Consular Assistance in the Framework of the Guarantees of the due Process of Law,
Advisory Opinion, OC-16, Inter-American Court of Human Rights Series A No 16 ( 1 October 1999), para
114.
32 Roger Judge v. Canada, Human Rights Committee, Communication No. 829/1998, U.N. Doc.
CCPR/C/78/D/829/1998 (2003), para 10.3.
33 Responsibilities and obligations of States sponsoring persons and entities with respect to activities in the Area, Seabed
Disputes Chamber of the International Tribunal for the Law of the Sea, Advisory Opinion (1 February 2011)
ITLOS/Case 17, para 117.
34 Iron Rhine, Arbitration (Belgium v. The Netherlands) (2005) para 79.
35 WTO, United States – Import Prohibition of Certain Shrimp and Shrimp Products – Report of the Appellate Body
(22 October 2001) WT/DS58/AB/R, para 130 and WTO, China – Measures Affecting Trading Rights and
Distribution Services for Certain Publications and Audiovisual Entertainment Products – Report of the Appellate Body
(21 September 2009) WT/DS363/AB/R, para 397.
36 Legal Consequences for States of the Continued Presence of South Africa in Namibia (South West Africa)
notwithstanding Security Council Resolution 276 (1970), [1971] ICJ Reports 16, para 53.
37 Aegean Sea Continental Shelf [1978] ICJ Reports 3, para 77.
38 Dispute regarding Navigational and Related Rights (Costa Rica v. Nicaragua) [2009] ICJ Reports 213, para 64-
66.
39 Nationality Decrees Issued in Tunis and Morocco on Nov. 8th, 1921, PCIJ Rep Series B, No 4
40 Nationality Decrees Issued in Tunis and Morocco (n 39) 24. However, ‘domestic jurisdiction’ is not prone to
change over time in the way that ‘sacred trust’, ‘territorial status’, and ‘comercio’ have; ‘domestic jurisdiction’
means the same today as it did in 1923.
41
Aegean Sea (n 37) para 78. Similarly to ‘domestic jurisdiction’, though, ‘rights’ is a term that cannot be said to
change its meaning over time. New rights are created and old rights cease to exist, but they are all ‘rights’ in
the original meaning of the term. (Hugh Thirlway, 'The Law and Procedure of the International Court of
Justice 1960–1989 Part One' (1989) 60 BYBIntlL 1, 141 makes a similar point.)
42 Gabcikovo-Nagymaros Project (Hungary/Slovakia) [1997] ICJ Reports 7.
43 Gabcikovo-Nagymaros (n 42) paras 112 and 140. This was (perhaps arguably) not an instance of evolutive
interpretation by the Court; it merely recommended the parties to take current environmental standards into
account when renegotiating the treaty (Dupuy (n 1) 129-130). Judge Bedjaoui interpreted the three articles
evolutively; see para 17 of his Dissenting Opinion.
VOLUME 6 EJLS ISSUE 1
133
There are also examples of tribunals using explicitly static interpretations; see eg the ICJ's
Land and Maritime Boundary between Cameroon and Nigeria44
case, and the Laguna del desierto45
and Decision regarding delimitation of the border between Eritrea and Ethiopia46
arbitrations.
Evolutive interpretation is also recognized by ‘publicists’, who tend to emphasize the
concepts,47 terms,48 objects and purposes,49 or intentions50 of a treaty as the basis for
evolution. There have also been dissenting voices; evolutive interpretation has been called
‘one of the most contentious, disputed and discussed issues in treaty interpretation’,51 and its
compatibility with the VCLT Articles 31-33 (and thus with customary international law)
has been questioned.52 The methodologies of tribunals have been criticized,53 as has the
normative soundness of the concept.54
The rest of this section will focus on ICJ decisions. These are (purportedly) the best proxy
for the content of customary international law.55 Moreover, unlike other tribunals, the ICJ
44 Land and Maritime Boundary between Cameroon and Nigeria (Cameroon v. Nigeria: Equatorial Guinea
intervening) [2002] ICJ Reports 303, para 159.
45 Case concerning a boundary dispute between Argentina and Chile concerning the delimitaion [sic] of the frontier line
between boundary post 62 and Mount Fitzroy, Reports of International Arbitral Awards volume XXII 3-149, para
130.
46 Decision regarding delimitation of the border between Eritrea and Ethiopia, Reports of International Arbitral
Awards volume XXV 83, para 3.5.
47 Jennings and Watts (eds) (n 10), at 1282.
48 Eduardo Jiménez de Aréchaga, ‘International Law in the Past Third of a Century’ (1979) 159 Recueil des
Cours 1, 49; Ian Sinclair, The Vienna Convention on the Law of Treaties (2nd edn, Manchester UP 1984) 140.
49 Rudolf Bernhardt, ‘Evolutive Treaty Interpretation, Especially of the European Convention on Human
Rights’, in (1999) 42 German YB Intl L 11, 16-17.
50 Frank Engelen, Interpretation of Tax Treaties under International Iaw: A Study of Articles 31, 32 and 33 of the
Vienna Convention on the Law of Treaties and Their Application to Tax Treaties (IBFD 2004) 285-286; Bruno
Simma and Theodore Kill, ‘Harmonizing Investment Protection and International Human Rights: First Steps
Towards a Methodology’ in Binder and others (eds), International Investment Law for the 21st Century: Essays in
Honour of Cristoph Schreuer (OUP 2009) 694; Dupuy (n 1) 126.
51 M Fitzmaurice (n 1), at 188.
52 John H Jackson, Sovereignty, the WTO, and Changing Fundamentals of International Law (CUP 2006) 187;
Petros C Mavroidis, ‘No Outsourcing of Law? WTO Law as Practiced by WTO Courts’ (2008) 102 AJIL 421,
445. Jackson, however, seems to be caught in a false dichotomy between ‘originalism’ and ‘living document’
views, not taking into account that an original intention may be for a text to evolve.
53 See eg Martin Dawidowicz, ‘The Effect of the Passage of Time on the Interpretation of Treaties: Some
Reflections on Costa Rica V. Nicaragua’ (2011) 24 Leiden J Intl L 201, 221-222; Thirlway (n 41) 137 and 142;
Duncan French, ‘Treaty Interpretation and the Incorporation of Extraneous Legal Rules’ (2006) 55 ICLQ 281,
296-300.
54 Antonin Scalia, ‘Common-Law Courts in a Civil-Law System: The Role of United Stated Federal Courts in
Interpreting the Constitution and Laws’ in Amy Gutmann (ed) A matter of Interpretation: Federal Courts and the
Law (Princeton UP 1997) 44-45 argues that what he labels ‘evolutionism’ (in the context of United States
constitutional law) is ‘not a practicable constitutional philosophy’ since there is ‘no chance of agreement, upon
what is to be the guiding principle of the evolution’. However, the chance of agreement on ‘present meaning’
as the guiding principle of evolution should be no less than the chance of agreement of ‘historic meaning’ as
the guiding principle of static interpretation. What a term meant in the past is no more objectively
ascertainable than what it means now.
55 See eg Hersch Lauterpacht, The Development of International Law by the International Court (Stevens & Sons
1958) 22; Rosalyn Higgins, Problems & Process: International Law and How We Use It (OUP 1994) 202; Alain
Pellet, ‘Article 38’ in Andreas Zimmermann, Christian Tomuschat, and Karin Oellers-Frahm (eds), The Statute
VOLUME 6 EJLS ISSUE 1
134
has offered both generalized and relatively detailed instructions on when to use evolutive
interpretations.
2.2 Party Intention
All evolutive interpretations by the ICJ have – at least prima facie – been prompted by the
intentions of the treaty parties.
In Namibia, the Court was asked to clarify the legal consequences of South Africa’s
continued presence in Namibia, after South Africa's mandate to administer the territory was
terminated in 1966. To do this, the Court had to interpret Article 22 paragraph 1 of the
Covenant of the League of Nations56. The Court held that the terms ‘the strenuous
conditions of the modern world’, ‘the well-being and development of such peoples’, and
‘sacred trust’ ‘were not static, but were by definition evolutionary’.57 The Court noted ‘the
primary necessity of interpreting an instrument according to the intentions of the parties at
the time of its conclusion’, and found that the parties must ‘be deemed to have accepted’ the
evolution,58 in the absence of decisive evidence to the contrary. This meant that South
Africa’s obligations towards the Namibian people (under the ‘sacred trust’) were affected by
‘changes which have occurred’ since the drafting of the Covenant.59 These led the Court to
conclude that ‘the ultimate objective of the sacred trust was the self-determination and
independence of the peoples concerned’,60 even though this right of independence was not a
reality – and perhaps not even contemplated – when the Covenant was drafted.
The Aegean Sea case sprang out of the Aegean Dispute between Greece and Turkey. Greece
had requested the Court to rule in a dispute over the continental shelf boundary between
the two States. The Court eventually found that it was without jurisdiction to decide the
matter. The result hinged on the interpretation of a reservation in Greece’s instrument of
accession to the General Act61. The reservation excluded ‘disputes concerning questions
which by international law are solely within the domestic jurisdiction of States, and in
particular disputes relating to the territorial status of Greece, including disputes relating to
its rights of sovereignty over its ports and lines of communication’.62 One of Greece’s
arguments was that at the time when the General Act was drafted (1928), and when Greece
acceded to it (1931), ‘the very idea of the continental shelf was wholly unknown’.63 The
argument implies that the meaning of the provision was frozen in time, either in 1928 or
1931. The Court did not agree, but rather established a ‘presumption’ that the meaning of
of the International Court of Justice: A Commentary (OUP 2006) 789-790. There are also dissenting voices; see eg
Jörg Kammerhofer, ‘Alexander Orakhelashvili. The Interpretation of Acts and Rules in Public International
Law’ (2009) 20 EJIL 1282.
56 The Covenant of the League of Nations (adopted 28 June 1919, entered into force 10 January 1920) 225
CTS 195.
57 Namibia (n 36) para 53.
58 ibid para 53.
59 ibid para 53.
60 ibid para 53.
61 General Act for the Pacific Settlement of International Disputes (adopted 26 September 1928, entered into
force 16 August 1929) 93 LNTS 344 (General Act).
62 Cited in Aegean Sea (n 37) para 48.
63 ibid para 77.
VOLUME 6 EJLS ISSUE 1
135
‘territorial status’ was ‘intended’ to ‘follow the evolution of the law’. As in Namibia, the
Court did not examine whether this presumed intention had been explicitly acknowledged
at the time of drafting or at the time of accession.
Dispute regarding Navigational and Related Rights concerned the interpretation of an 1858
treaty64 between Costa Rica and Nicaragua. The treaty, which was drafted only in Spanish,
gave Costa Rica rights of navigation ‘con objetos de comercio’ on the San Juan River, which
runs on the border between the two countries. Nicaragua wanted ‘con objetos de comercio’ to
be interpreted as ‘with articles of trade’, ie only transportation of physical goods. Costa Rica
argued that the correct interpretation was ‘for the purposes of commerce’, which would
extend the freedom of navigation to a much wider range of activities. The latter
interpretation was accepted by the Court.65 Regarding the term ‘comercio’, Nicaragua argued
that in 1858, it included only transportation of physical goods, and that this original
meaning should prevail. The Court’s starting point was that ‘a treaty must be interpreted in
light of what is determined to have been the parties’ common intention’, and that the
intention is ‘contemporaneous with the treaty’s conclusion’.66 The intention may, however,
have been ‘to give the terms used […] a meaning or content capable of evolving’. Such an
intention does not have to be explicit; it ‘may be presumed’.67 The Court supported its
argument with a reference to Aegean Sea.68
These three cases show that, according to the ICJ, terms must be interpreted evolutively if
(and, apparently, only if) the parties intended it. While this only says when, and not how,
terms shall evolve, the latter question is presumably also controlled by party intention.
That a treaty's drafters intended terms to evolve does not presuppose that they could have
foreseen the exact interpretive results reached by a future interpreter, or that they intended
a specific future interpretation to prevail.69 There is an important distinction between
intention, control, and prediction. Evolutive treaty provisions may evolve as intended, even
though they do so in ways the drafters cannot control and could not predict.70
2.3 A ‘General Rule’
In Dispute regarding Navigational and Related Rights, the ICJ formulated a ‘general rule’ to
determine when an evolutive intention ‘must’ be presumed:71
64 Treaty of Limits (Costa Rica-Nicaragua, 15 April 1858).
65 Dispute regarding Navigational and Related Rights (n 38) para 56.
66 ibid para 63.
67 ibid para 64.
68 ibid para 65.
69 See Matthews v. The United Kingdom, 30 EHRR (1999) 361, para 39: ‘The mere fact that a body was not
envisaged by the drafters of the [ECHR] cannot prevent that body from falling within the scope of the
Convention’.
70 Bernhardt (n 49) 17 is at risk of confounding this when presenting a dichotomy between ‘the original
intentions of the drafters’ of the ECHR and ‘the relevance of changing conditions and opinions in State and
society’. In this sense, evolutive interpretations do not have to be ‘removed from’ the intentions of the parties,
as seems to be suggested by Catherine M Brölmann, 'Law-Making Treaties: Form and Function in
International Law' (2005) 74 Nordic J Intl L 383, 394.
71 Dispute regarding Navigational and Related Rights (n 38) para 66.
VOLUME 6 EJLS ISSUE 1
136
(1) First, the parties have used ‘generic terms’ (in which case the parties have ‘necessarily
[…] been aware that the meaning of the terms was likely to evolve over time’), and
(2) the treaty ‘has been entered into for a very long time or is 'of continuing duration'‘.
Applied on the treaty at issue, the ‘general rule’ allowed the Court to presume an intention
to let the term ‘comercio’ evolve; the term was ‘generic’, and the treaty’s duration was
‘unlimited’.72 The conclusion was that Costa Rica’s right now covered activities that in 1858
(when the treaty was concluded) were not considered ‘comercio’.
The rule built on the ICJ’s approach in Aegean Sea, where the fact that ‘territorial status’
was a ‘generic’ term gave rise to the presumption that ‘its meaning was intended to follow
the evolution of the law’, an argument that was supported by the fact that the General Act
was ‘designed to be of the most general kind’ and of ‘continuing duration’.73
The ‘general rule’ was not used in Namibia. There, the Court found that the terms it
interpreted were ‘by definition, evolutionary’ and that the parties ‘must be deemed to have
accepted’ this.74 The Court instead supported its reasoning with the observation that ‘an
international instrument has to be interpreted and applied within the framework of the
entire legal system prevailing at the time of the interpretation’.75 The quote seems to refer
to the principle in Article 31.3.c,76 but has alternatively read been read as a reference to a
‘principle of harmonization’,77 and an obiter on evolutive interpretation.78
In formulating the ‘general rule’, the ICJ used the word ‘must’. As a consequence, presuming
evolutive intent is an obligation on, not just an option for, the interpreter.
72 ibid para 67.
73 Aegean Sea (n 37) para 77.
74 Namibia (n 36) para 53.
75 ibid para 53.
76 Ulf Linderfalk, On the Interpretation of Treaties: The Modern International Law as Expressed in the 1969 Vienna
Convention on the Law of Treaties (Springer 2007) 83; Oliver Dörr, ‘Article 31. General Rule of Interpretation’
in Oliver Dörr and Kirsten Schmalenbach (eds), Vienna Convention on the Law of Treaties: A Commentary
(Springer 2011) 560. Villiger (n 7) 433 and Gardiner (n 4) 255-256 seem to agree. The statement is universal,
in the sense that it applies to all ‘international instruments’. That makes it similar to the principle in art 31.3.c.
It also indicates that the statement is not a reference to the concept of evolutive interpretation, since that only
applies to instruments intended to evolve. On the other hand, both the principle in art 31.3.c and the concept
of evolutive interpretation concern only the ‘interpretation’ of treaties, while the quoted passage includes both
‘interpreted’ and ‘applied’. The quote does not seem to be a reference to the doctrine of intertemporality (see s
4.1 below), since the ‘continued manifestation’ of a ‘right’ is something else than the ‘interpretation and
application’ of a ‘legal instrument’.
77 Isabelle Van Damme, 'Jurisdiction, Applicable Law, and Interpretation' in Daniel Betlehem and other (eds),
The Oxford Handbook of International Trade Law (OUP 2009) 330. This principle is, according to ILC,
‘Fragmentation of International Law: Difficulties arising from the Diversification and Expansion of
International Law’, Report of the Study Group of the International Law Commission, 13 April 2006,
A/CN.4/L.682, at para 415, something more than the principle in art 31.3.c. Van Damme backs up her
argument by referring to para 38 of the same ILC report. However, that part of the report and the phrase in
Namibia concern different things: The report refers to ‘previous treaty obligations’, while Namibia’s phrase
concerns rules ‘prevailing at the time of the interpretation’.
78 Dupuy (n 1) 129.
VOLUME 6 EJLS ISSUE 1
137
The ‘general rule’ is retroactive, in the sense that it applies to older treaties (for example,
the treaty in Dispute regarding Navigational and Related Rights was concluded in 1858).
With regards to treaties, the VCLT restricts retroactive application (Article 28), but not
retroactive interpretation. The issue has practical relevance since customary law of treaty
interpretation is continuously developing.79
According to the Court, presuming evolutive intent in cases where this would expand a
State’s obligations does not violate the principle that ‘[r]estrictions upon the independence
of States cannot […] be presumed’.80 The principle stems from the PCIJ’s Lotus
judgement,81 and could generally be said to have lost traction (if not disappeared
completely) in recent times.82
Even though the ‘general rule’ applies generally, it is no universal solution to all questions
regarding evolutive interpretation.83 It only says that when two conditions are fulfilled, an
evolutive intention shall be presumed. This does not exclude establishing evolutive
intentions where the conditions are not fulfilled, based on other arguments. And even when
the two conditions are fulfilled, the presumption can be refuted by other arguments.
Moreover, the rule seems suited only for terms that evolve (linguistically), since it offers no
guidance on when to interpret non-evolving terms evolutively. Finally, it only concerns
when an intention to let terms evolve shall be presumed; it says nothing about how terms
shall evolve.
The rule has been criticized for an inherent risk of producing fictional intentions,84 and it
has been warned that evolutive interpretations detached from the intention of the parties
‘may provide tribunals too much latitude, with too few safeguards, for discretionary
decision-making’85. Regarding the latter, that is true of all interpretations detached from the
parties’ intentions. A static interpretation where the parties intended evolution can be just
as harmful as an evolutive interpretation the parties did not intend. As for the risk of
fictional intentions, this is somewhat mitigated by the fact that the ICJ has on several
occasions concluded that treaty parties did not intend evolution despite using evolving
terms. Two examples are cited in Dispute regarding Navigational and Related Rights.86
The first is Rights of nationals of the United States of America in Morocco87. The Court found that
the term ‘dispute’ in an 1836 treaty88 between the US and Morocco was intended to cover
both civil and criminal cases, since this was how the term was used in the Moroccan legal
79 See Gardiner (n 4) 51-69 for a historical overview.
80 The point is discussed briefly in Dispute regarding Navigational and Related Rights (n 39) para 47.
81 The Case of the S.S. Lotus, [1927] PCIJ Series A No 10, 18.
82 Luigi Crema, ‘Disappearance and New Sightings of Restrictive Interpretation(s)’ (2010) 21 EJIL 681, 686-
688.
83 As Malgosia Fitzmaurice, ‘Dynamic (Evolutive) Interpretation of Treaties, Part I’, in (2008) 21 Hague YB
Intl L 101, 153 notes, the concept of evolutive interpretation still awaits ‘some general and definite
conclusions’.
84 Thirlway (n 41) 142; French (n 53) 296-297.
85 French (n 53) 300.
86 Dispute regarding Navigational and Related Rights (n 38) para 63.
87 Case concerning rights of nationals of the United States of America in Morocco [1952] ICJ Reports 176.
88 Treaty of Peace and Friendship (United States of America-Shereefian Empire) (16 September 1836).
VOLUME 6 EJLS ISSUE 1
138
system when the treaty was drafted.89 This was despite the argument that the term, in its
‘ordinary and natural sense’ at the time the case was decided, referred only to civil cases.
The second example, the Kasikili/Sedudu Island90 case, concerned a border dispute. It was
settled on the basis of an 1890 treaty91 between the former empires of Germany and Britain,
drafted in both a German and an English version. In interpreting the phrase ‘centre of the
main channel’, which corresponded to ‘Thalweg des Hauptlaufes’ in the German version, the
Court noted that the terms ‘centre’ and ‘Thalweg’ did not have the same meaning at the
time the case was decided,92 but also that they were ‘used interchangeably’ in 1890.93
Therefore the parties had intended them to mean the same, and the Court solved the dispute
on that basis. Put differently, the meaning of the terms had evolved since the treaty’s
conclusion, but the parties had not intended any evolving meaning to prevail.
A third example of generic terms not being interpreted evolutively is found in the Petroleum
Development Ltd v. Sheikh of Abu Dhabi94 arbitration. While not an ICJ case itself, it is
notable because the ICJ explicitly distinguished it from its own reasoning in Aegean Sea.95
The case concerned the interpretation of a 1939 contract that gave the company Petroleum
Development the right to extract oil from the ‘lands which belong to the Ruler of Abu
Dhabi and its dependencies’ and from ‘all the islands and sea waters which belong to that
area’.96 The umpire presumed that by 1939, the modern concept of ‘continental shelf’ was
unknown, and ‘sea waters’ thus had to be limited to the ‘territorial maritime belt and its
subsoil’ of three miles from the coast.97 An alternative approach would have been to
consider ‘sea waters’ an evolving term, which would include whatever ‘sea waters’ (and their
corresponding shelf) that at any time was under the Sheik’s sovereignty. In distinguishing
its own reasoning in Aegean Sea from the umpire’s statement, the ICJ noted that there was
‘an essential difference’ between the two cases: It may be presumed that someone parting
with valuable property rights ‘intends only to transfer the rights which he possesses at that
time’, while a State, ‘in agreeing to subject itself to compulsory procedures of pacific
settlement, excepts from that agreement’ a ‘generic’ category of disputes, can be presumed
to have intended to make a reservation against anything falling within the ambit of the
reservation in the future.98
In sum, these cases indicate that even where terms evolve, the ICJ is not willing to
construct an evolutive intention in cases where a non-evolving intention is evident, or
where there are specific circumstances that make the presumption of evolving intent
implausible.
89 Rights of Nationals of the United States of America in Morocco (n 87) 189.
90 Kasikili/Sedudu Island (Botswana/Namibia) [1999] ICJ Reports 1045.
91 Anglo-German Agreement of 1 July 1890 (Great Britan-Germany) (1 July 1980).
92 Kasikili/Sedudu Island (n 90) para 24.
93 ibid para 25.
94 Petroleum Development Ltd v. Sheikh of Abu Dhabi, (1951) 18 ILR 144.
95 Aegean Sea (n 37) para 77.
96 Petroleum Development Ltd v. Sheikh of Abu Dhabi (n 94) 151.
97 ibid 152.
98 Aegean Sea (n 37) para 77.
VOLUME 6 EJLS ISSUE 1
139
2.4 A Taxonomy of Terms
For purposes of evolutive interpretation, terms can be divided by two important
distinctions. One is between terms that cannot be interpreted without value judgements,
and terms whose meaning does not depend on values. These can be called ‘value driven’ and
‘non-value driven’ terms, respectively. Examples of the former include ‘inhuman
punishment’, ‘fair trial’, and ‘the well-being and development’ of peoples. Examples of the
latter include ‘territorial status’ and ‘comercio’. The other distinction is between terms that
do and do not evolve linguistically, as outlined in the introduction above. The two
categories can be called ‘evolving’ and ‘non-evolving’.
When value driven terms evolve, tribunals seem to accept that the evolution was intended
by the parties, without demanding further justification. That is presumably because values
inevitably change over time, as new generations will have their own views on what is (for
example) ‘inhuman’ or ‘fair’. The parties are simply assumed to have been aware of this. The
evolution of a non-value driven term is, on the other hand, not inevitable, and is thus less
easily anticipated. The practice of the ICJ illustrates the point. Of the ICJ’s three evolutive
interpretations, only those in Namibia required value judgements. Namibia is also the only
case where the ICJ did not see the need for its ‘general rule’, but established evolutive intent
solely on the basis of the nature of the terms themselves. A further example comes from
human rights tribunals, who interpret many value driven terms, and frequently use
evolutive interpretations.99
As noted in section 2, the only way a non-evolving term can be interpreted evolutively is to
give it an evolving ‘special meaning’, as per the VCLT Article 31.4.
A taxonomy of terms could thus look like this:
TABLE I
Category Approach Illustrations
Value driven
evolving terms
Evolutive intent can be assumed by default Namibia, human rights
tribunals
Non-value driven
evolving terms
Evolutive intent can be established after a more
comprehensive evaluation, eg the ICJ’s ‘general
rule’
Aegean Sea, Dispute
regarding Navigational
and Related Rights
Non-evolving terms Evolution must be based on a ‘special meaning’ The VCLT Article 31.4
3. EVOLUTIVE INTERPRETATION AND SEMANTICS
3.1 ‘Generic Terms’
The ICJ has made ‘generic terms’ one of two conditions in its ‘general rule’ on evolutive
interpretation. The closest thing to a definition of ‘generic terms’ it has given is that they
99 See s 2.1 above, on how evolutive interpretation is a ‘key theme’ of the ECtHR.
VOLUME 6 EJLS ISSUE 1
140
‘[refer] to a class of [something]’.100 The terms that the ICJ has acknowledged to be
generic are ‘continental shelf’101 and ‘comercio’.102
This section will try to establish what the ICJ means by ‘generic’ terms.
‘Generic reference’ is a concept in the philosophy of language.103 It can be defined as a
designation for references that may be used to assert a ‘generic proposition’.104 A generic
proposition is one whose referent is not a specified group or individual, but an indeterminate
class of referents.105 The generic reference is distinct from the ‘singular reference’ and the
‘general reference’. A singular reference refers to an individual entity, while a general
reference refers to a specific set of entities.106 Both can be either ‘definite’ or ‘indefinite’. The
former type refers to some specific individual entity or group, while the latter does not.107
Since generic references refer to ‘classes’ of referents, this philosophical definition looks
similar to the one given by the ICJ.
The interpretations of ‘territorial status’ and ‘comercio’ seem to conform to the philosophical
definition of generic. ‘Territorial status’ refers to a class of issues, and ‘comercio’ to a class of
activities. Neither of them refers to specified entities, but to whatever entities that happen to
share some particular trait(s).
In both cases, the interpreted terms had evolved (linguistically), in the manner described in
section 1. Such evolution seems to be a prerequisite for using the ICJ’s ‘general rule’, as
noted in section 2. Not all generic references evolve, however. A basic example of a generic
reference is the proposition ‘lions are friendly beasts’.108 ‘Lions’ does not refer to any set
group of lions, but to lions as such. Yet the term is not likely to change in the way that
‘sacred trust’, ‘territorial status’, and ‘comercio’ have. Lions will always be lions. While all
lions die, and new lions will be borne, they are all ‘lions’ in the original meaning of the
proposition. The ‘ordinary meaning’ of the term neither has changed nor is likely to change;
the term ‘lion’ does not have to ‘evolve’ to encompass a lion that will be born tomorrow.
This inevitably leads to the conclusion that genericness, in the philosophical sense, does not
necessitate evolution, in the legal sense. This conclusion must be reconciled with the ICJ’s
statements in Dispute regarding Navigational and Related Rights. The Court noted that ‘where
the parties have used generic terms in a treaty, the parties necessarily having been aware
100 Dispute regarding Navigational and Related Rights (n 38) para 67.
101 Aegean Sea (n 37).
102 Dispute regarding Navigational and Related Rights (n 38).
103 Lyons, Semantics: 1 (1977) 193-197.
104 ibid 194.
105 ibid 194.
106 ibid 178.
107 ibid 178; Linderfalk (n 76) 75-76. For example, the ‘the parties’ in the VCLT art 31.3.c is a general definite
reference. It is ambiguous (see note 144 below), but not generic. Van Damme (n 77) 334-335 notes that the
WTO panel in the European Communities – Measures Affecting the Approval and Marketing of Biotech Products –
Report of the Panel (29 September 2006) WT/DS291/R, WT/DS292/R, and WT/DS293/R rejected an
‘evolutionary and effective interpretation of the phrase’. The Panel was correct in doing so, since the phrase is
not generic, and the interpretive issue is about ambiguity rather than vagueness (see s 3.2 below).
108 Taken from Lyons (n 103) 194.
VOLUME 6 EJLS ISSUE 1
141
that the meaning of the terms was likely to evolve over time […]’,109 but also that ‘a generic
term’ is one that refers ‘to a class’ of (in this case) activities.110 As has been shown above, not
all generic references are ‘likely’ (ie have more than a 50 % chance) to evolve. The Court
could not infer from the fact that ‘comercio’ refers to a class of activities that the term ‘was
likely to’ change meaning over time.111
There seems to be three ways of reconciling the ICJ’s statements.
The first is to interpret all generic (in the philosophical sense) terms evolutively, regardless
of whether they have evolved linguistically. This is not feasible; a term cannot be
interpreted evolutively without either having evolved or having been assigned an evolving
‘special meaning’.
Another option to apply the rule to all evolving terms, leaving non-evolving terms
unaffected (despite the fact these too can be generic). This solution is somewhat
unsatisfactory from a lex ferenda point of view, since the rule would still apply to all generic
terms that had evolved; even those where evolution was unlikely, or nigh impossible to
predict. In the latter cases, it is hardly fair to presume an intention to let evolved meanings
prevail.
The final possibility is to introduce a third condition for presuming evolutive intent. In
addition to the evolving term being generic and the treaty being long-term or indefinite,
some evolution in the term’s meaning must have been more likely than not at the time of
the treaty’s conclusion.112 This seems preferable from a lex ferenda point of view, but cannot
be said to be reflected in the ICJ’s doctrine.113
One thing that is clear from the ICJ’s decisions is that presumptions in favour of evolutive
intent are restricted to generic terms. This is sensible; when a reference is not generic, it is
singular or general. When treaty parties use singular or general references, they specify
what entities they refer to. They usually do not intend later changes in meaning to affect
that. An example is the ICJ's Land and Maritime Boundary between Cameroon and Nigeria case.
The parties had referred to ‘the mouth of the [river] Ebeji’, which is a singular reference, in
a treaty. The ICJ found that at the time of the treaty’s conclusion, ‘the parties only
109 Dispute regarding Navigational and Related Rights (n 38) para 66.
110 ibid para 67.
111 Just as the fact that the proposition ‘lions are friendly beasts’ is generic does not imply that ‘lions’ will
change its meaning over time.
112 Thirlway (n 41) 137 criticizes the ICJ’s reasoning in Namibia, arguing that it was never proved that at the
time of the treaty’s conclusion, the concepts being interpreted evolutively were in fact regarded as such.
Dawidowicz (n 53) 221-222 criticises Dispute regarding Navigational and Related Rights on the same grounds,
endorsing the Separate Opinion of Judge Skotnikov. The complaints are fair, and the introduction of this third
condition would make the requirement of such proof unequivocal.
113 The closest approximation is Judge Higgins’ Declaration in Kasikili/Sedudu Island (n 90), which in para 2
defines ‘generic term’ as ‘a known legal term, whose content the parties expected would change through time’.
Her definition is echoed by Dörr (n 76) 534. The synthesis of ICJ doctrine up to 2007 (ie excluding Dispute
regarding Navigational and Related Rights) contained in Linderfalk (n 76) 95 seems to include the condition as
well, in that evolutive interpretation is only permissible if ‘it can be shown that the thing interpreted is a
generic referring expression with a referent assumed by the parties to be alterable’ (emphasis added).
VOLUME 6 EJLS ISSUE 1
142
envisaged one mouth’,114 and it let that understanding prevail without further discussion.
This view also explains the static interpretations reached in the Laguna del desierto and
Decision regarding delimitation of the border between Eritrea and Ethiopia arbitrations
mentioned above. In the first, the arbitration panel found that the (singular) reference to a
‘water-parting’ was ‘not susceptible of any subsequent change through usage’ or ‘evolution
of the language’.115 The commission deciding the Decision regarding delimitation of the border
between Eritrea and Ethiopia arbitration held that it would interpret treaties ‘by reference to
the circumstances prevailing when the treaty was concluded’, which involved ‘giving
expressions (including names) used in the treaty the meaning that they would have
possessed at that time’.116 References to names will usually be singular references, which
means that static interpretations are the most sensible.
3.2 Ambiguity and Vagueness
Questions of treaty interpretation can exist on two different levels. On one level are
questions of resolving ambiguity, on another, questions of resolving vagueness.
The distinction is relatively clear-cut: ‘A vague word has one meaning (and its application is
unclear in some cases); an ambiguous word has more than one meaning (and it may be
unclear, in some cases, which is in use)’.117
To illustrate the distinction, Ogden and Richards’ ‘triangle of reference’ could be a useful
tool. It distinguishes between ‘symbol’, ‘reference’, and ‘referent’. ‘Symbols’ are words,
‘references’ are the thoughts symbolized by the symbols, and ‘referents’ are the phenomena
referred to by thoughts.118 One symbol may symbolize several distinct references, but each
reference has only one (more or less clear-cut) referent, which may or may not be
fictional.119
Ogden and Richards visualized the triangle as follows.120
114 Land and Maritime Boundary between Cameroon and Nigeria (n 44) para 59.
115 Laguna del desierto (n 45) para 130.
116 Decision regarding delimitation of the border between Eritrea and Ethiopia (n 46) para 3.5.
117 Timothy AO Endicott, Vagueness in Law (OUP 2000) 54. Jeremy Waldron, ‘Vagueness in Law and
Language: Some Philosophical Issues’ (1994) 82 California L Rev 509, 512-513 has a more detailed
explanation.
118 Ogden and Richards, The Meaning of Meaning (10th edn, Routledge & Kegan Paul 1949) 9-11.
119 For example, the symbol ‘Napoleon’ symbolizes, among other things, the first Emperor of the French, and a
character in George Orwell’s 1945 novella Animal Farm. Their referents are a real man and a fictional pig,
respectively.
120 Ogden (n 118), at 11.
VOLUME 6 EJLS ISSUE 1
143
Combining the two theories shows, firstly, that only symbols can be ambiguous, by
symbolizing more than one reference, and secondly, that only references can be vague,
which is the case when the scope of a reference is unclear.
The distinction between ambiguity and vagueness has implications for the concept of
evolutive interpretation.
Treaties (and all other sources of law) consist of symbols. Symbols may be ambiguous, but
they are, presumably, always intended to symbolize a single reference. The reference may or
may not be vague, and vagueness may or may not be intentional.
Certain references are vague in the sense that their scope varies over time. In such cases, it
could be plausible to presume that the text is intended to evolve in line with the changing
reference, in which case an evolutive interpretation is appropriate. For example, ‘cruel and
unusual punishment’ (in the Eighth Amendment to the United States Constitution) will
always symbolize the same reference, but the scope of the reference will change in line what
is considered ‘cruel’ or ‘unusual’ in any given era.
Symbols may also change over time, in the sense that the symbol may come to symbolize
new references, and no longer symbolize old references. An example is the word ‘gay’. It
used to primarily symbolize the attribute of being light-hearted and carefree. Nowadays, the
most common symbolization is homosexuality.
Evolutive interpretation is an inappropriate tool for resolving ambiguity, for two reasons.
First, it is generally easier to predict whether a reference will change than to predict whether
symbolizations will change. For example, using a value driven term makes the reference
bound to change over time. Non-value driven terms can also be predicted to change their
reference; for example in that the term ‘comercio’ can come to include new activities in
future. The same cannot be said about changing symbolizations.
Second, it is easier to predict how a reference will change than to predict how a
symbolization will change. When a reference changes, it is usually as a variation on what it
was before (such as when a form of punishment that used to be considered human is
considered ‘inhuman’, or when ‘comercio’ comes to include a new activity). When a
VOLUME 6 EJLS ISSUE 1
144
symbolization changes, however, the new symbolization may bear little resemblance to the
old. The changing symbolizations of ‘gay’ is a case in point.
These two reasons make it much less plausible to presume that treaty drafters intended new
symbolizations to prevail than it is to presume evolutive intent for changing references.
Thus, if an interpretive issue is on the level of resolving ambiguity, evolutive interpretations
are of little use. Their main function lies in resolving vagueness.
All the ICJ’s evolutive interpretations have concerned vagueness. Dispute regarding
Navigational and Related Rights illustrates the point especially well. The symbol ‘con objetos
de comercio’ was ambiguous, and the Court resolved the ambiguity without evolutive
interpretations. The reference ‘comercio’, which was part of the symbolization the Court
chose, was vague, and the vagueness was resolved by an evolutive interpretation.
The distinction also explains an interesting difference between the two evolutive
interpretations by the WTO Appellate Body. In US – Shrimp, it had to interpret the phrase
(ie symbol) ‘natural resources’ in the GATT 1994121 Article XX(d). The parties disagreed
on whether living resources were covered by the provision.122 This was about resolving
vagueness; the parties agreed that ‘natural resources’ were resources found in nature, but
not on the exact contours of the concept. In China – Publications, the interpretation of the
phrase (ie symbol) ‘Sound Recording Distribution Services’ in China’s GATS Schedule123
was contested. The parties offered two rivalling interpretations of ‘sound recording’: It
could refer either to the physical medium on which sound was recorded, or to the intangible
‘sound recording’ itself.124 This was a question of ambiguity, since these two are
fundamentally distinct references. The question could not be solved by evolutive
interpretation. This distinction is reflected by the role the evolutive interpretations played
in the two cases. The evolutive interpretation in US – Shrimp was part of the report's ratio
decidendi, and was used to resolve the interpretive issue.125 In China – Publications, by
contrast, the interpretive result was reached on the basis of ‘ordinary meaning’, ‘context’,
and ‘object and purpose’;126 the evolutive interpretation was (and had to be) an obiter. The
obiter served to outline the (vague) reference that was chosen by resolving the (ambiguous)
interpretive issue of the case.
4. EVOLUTIVE INTERPRETATION DISTINGUISHED
4.1 The Doctrine of Intertemporality
The so-called ‘doctrine of intertemporality’ is conceptually distinct from evolutive
interpretation, despite certain similarities between the two.
121 General Agreement on Tariffs and Trade (adopted 15 April 1994, entered into force 1 January 1995) 1869
UNTS 187 (GATT 1994).
122 US – Shrimp (n 35) paras 125-127.
123 Schedule CLII – The People's Republic of China (1 October 2001) WT/ACC/CHN/49/Add.2 (China’s
GATS Schedule).
124 China – Publications (n 35) paras 349-350.
125 US – Shrimp (n 35) paras 130-131.
126 China – Publications (n 35) para 398.
VOLUME 6 EJLS ISSUE 1
145
Giving a precise definition of the doctrine has proven difficult.127 It most famously featured
in the Island of Palmas128 arbitration. It was formulated as a ‘principle’, designed to answer
‘the question which of different legal systems prevailing at successive periods is to be
applied in a particular case’.
The umpire presented the doctrine as made up of two ‘elements’. The first was that ‘a
juridical fact must be appreciated in light of the law contemporaneous with it, and not of the
law in force at the time when a dispute in regards to it arises or falls to be settled’. Secondly,
‘the existence of [a] right, in other words its continued manifestation, shall follow the
conditions required by the evolution of law’.129
The first element in the doctrine concerns which system of law that should be applied on a
given ‘juridical fact’.130 Treaty interpretation, including questions of evolutive
interpretation, concerns how a treaty (which may be part of the law applied on a juridical
fact) is to be interpreted. These are two different matters.131
The doctrine’s second element is harder to pin down. Prima facie, it only says that a right
can be curbed or extinguished because of later developments in international law.132 The
question of how a right must be maintained is clearly distinct from the question of how
treaties should be interpreted (eg questions of evolutive interpretation).133
The distinction between the doctrine and evolutive interpretation does not seem to be
uniformly observed. Evolutive interpretation has been presented as a ‘qualification’ to the
doctrine’s first element,134 and the ICJ’s evolutive interpretation in Aegean Sea has been
called an ‘application of the doctrine of intertemporal law to the interpretation of a
treaty’.135 This confuses the distinct processes of deciding what law to apply and
interpreting terms. The doctrine can determine what law of treaty interpretation that
applies in a given situation, but only that law itself can determine whether terms in a treaty
should be interpreted evolutively.136
127 Koskenniemi (n 18) 455.
128 Island of Palmas (The Netherlands v. United States), 2 Reports of International Arbitral Awards (1928) 829.
129 ibid 845.
130 A ‘juridical fact’ can be defined as a ‘fact with juridical relevance’; Rosalyn Higgins, ‘Some Observations on
the Inter-Temporal Rule in International Law’ in Jerzy Makarczyk (ed), Theory of International Law at the
Threshold of the 21st Century: Essays in honour of Krysztof Skubiszewski (Springer 1996)173.
131 See Draft articles on Responsibility of States for Internationally Wrongful Acts, with Commentaries, reproduced in
ILC, Yearbook of the International Law Commission 2001 Volume II, Part Two (United Nations 2007), para 9 of
the commentary to art 13. Iron Rhine (n 32) para 79 apparently attempts to link the two concepts, by treating
the ‘intertemporal rule’ as a ‘relevant rule of international law’ under the VCLT art 31.3.c. The approach is
confusing; what the tribunal calls the ‘intertemporal rule’ is applicable to all treaties by default, there is no
need to use art 31.3.c.
132 Rosalyn Higgins, ‘Time and the Law: International Perspectives on an Old Problem’ (1997) 46 ICLQ 501,
516; Gardiner (n 4) 253.
133 Higgins (n 132) 178; Ulf Linderfalk, ‘Doing the Right Thing for the Right Reason – Why Dynamic or
Static Approaches Should be Taken in the Interpretation of Treaties’ (2008) 10 Intl Community L Rev 109,
118.
134 John H Currie, Public International Law (2nd edn, Irvin Law 2008) 166.
135 Taslim O Elias, ‘The Doctrine of Intertemporal Law’ (1980) 74 AJIL 285, 301; Kotzur (n 13) para 7.
136 Ulf Linderfalk, ‘The Application of International Legal Norms over Time: The Second Branch of
VOLUME 6 EJLS ISSUE 1
146
In addition to the ‘doctrine of intertemporality’, the term ‘intertemporal law’ is used in
various contexts. Textually, ‘intertemporal’ law means any law concerned with the passage
of time. Under that definition, the ‘doctrine of intertemporality’ and evolutive interpretation
are two examples of ‘intertemporal law’,137 with other examples being rules concerning
retroactivity (such as the VCLT Article 28 on the retroactivity of treaties) and norms of lex
posterior. The usage of the terms is, however, not uniform.138 The doctrine of
intertemporality has been given various names, including ‘the rule of intertemporal law’,139
‘the intertemporal rule’, and ‘the intertemporal principle’.140 More problematic is the fact
that the doctrine is not always distinguished from intertemporal law in general.141
Another related term is the ‘principle of contemporaneity’, under which ‘the terms of a
treaty must be interpreted according to the meaning they possessed, or which would have
been attributed to them, and in the light of current linguistic usage, at the time when the
treaty was originally concluded’.142 With the recognition of evolutive interpretation as part
of international law, the ‘principle’ is now only applicable to terms that the parties did not
intend to be interpreted evolutively.143
4.2 The VCLT Article 31.3.c
The principle in the VCLT Article 31.3.c allows ‘any relevant rules of international law
applicable in the relations between the parties’ to be taken into account when interpreting
treaties. There are debates over the interpretation of the provision’s various elements,144 but
those will not be pursued here.
Intertemporal Law’ (2011) 58 Netherlands Intl L Rev 147, note 51 makes a similar distinction.
137 See eg Tavernier (n 1) 397; Kotzur (n 13) para 1-3. According to Campbell McLachlan, ‘The Principle of
Systemic Integration and Article 31(3)(c) of the Vienna Convention’ (2005) 54 ICLQ 279, 316, evolutive
interpretation is one response to the ‘problem of intertemporality as it applies to treaties’. M Fitzmaurice (n
83) 113 calls the broader concept ‘the theory of intertemporal law’, which includes both the doctrine and
evolutive interpretation. Higgins (n 55) 797 calls evolutive interpretation ‘the temporal issue in treaty
interpretation’, which is presumably one out of several ‘temporal issues’ to be addressed by different rules of
intertemporal law.
138 While linguistic differences do not necessarily entail legal disagreement, there is a risk that Higgins (n 132)
516 is right in that the doctrine of intertemporality has ‘been read in the most remarkably extensive fashion, as
providing obligatory rules in circumstances that it never addressed, with consequences that it never intended’.
139 ibid 515.
140 Gardiner (n 4) 252 uses ‘intertemporal law’, ‘the intertemporal rule’ and ‘the intertemporal principle’ as
synonyms.
141 ibid 25 equates the doctrine with ‘intertemporal law’. Joost Pauwelyn, Conflict of Norms in Public
International Law: How WTO Law Relates to other Rules of International Law (CUP 2003) 266 writes: ‘This
reflects the so-called “evolutionary approach” to treaty interpretation. It is the second part of the
intertemporal law’. The statement equates, first, the doctrine of intertemporality with ‘intertemporal law’ in
general, and second, the doctrine’s second element with evolutive interpretation.
142 Gerald Fitzmaurice, ‘The Law and Procedure of the International Court of Justice 1951-4: Treaty
Interpretation and Other Treaty Points’ (1958) 34 BYBIL 203, 212. The definition is repeated by Dörr (n 76)
533 and Carlos Fernández de Casadevante y Romani, Sovereignty and Interpretation of International Norms (2007)
153 (who calls it the ‘principle of contemporariness’).
143 Dörr (n 76) 533 calls static interpretation a ‘basic rule’, Romani (n 142) 153 calls static interpretation a
‘general rule’ and evolutive interpretation an ‘exception’. That is imprecise; the ‘basic rule’ is that treaties shall
be interpreted according to their drafters’ intentions, be it evolutively or statically.
144 See eg Gardiner (n 4) 259-265; the biggest debate seems to be over the ambiguous phrase ‘the parties’.
VOLUME 6 EJLS ISSUE 1
147
Instead, the focus will be on the tendency of some sources to confound Article 31.3.c with
evolutive interpretation. One common assumption seems to be that the concept of evolutive
interpretation is limited to determining whether ‘relevant rules’ in Article 31.3.c must exist
at the time of a treaty’s conclusion or if subsequent rules are relevant as well.145 Whereas
Article 31.3.c does not specify whether subsequent rules can be ‘relevant’, evolutive
interpretation is not necessary to solving the question: Article 31.3.c is located in the same
subparagraph as 31.3.a and 31.3.b, both concerning interpretive elements subsequent to the
treaty being interpreted, and it may therefore be a plausible conclusion that any rule,
regardless of the time of its creation, can be ‘relevant’.146 The problem with the assumption
noted above is not that it is superfluous, however, but that it is incorrect: Evolutive
interpretation is conceptually independent from Article 31.3.c.147
The difference is simply that Article 31.3.c is about interpretation in light of other law,
while evolutive interpretation is about interpretation in light of some current meaning. This
means that the range of relevant arguments to determine the evolution of an evolving term
will often be much broader than just the ‘rules of international law’ that Article 31.3.c
mentions.148 Moreover, since evolutive interpretations are based primarily on the parties’
original intentions, and thus rooted in other parts of the VCLT Article 31 than 31.3.c, they
are permissible regardless of whether the conditions in Article 31.3.c are fulfilled. Evolution
145 ILC (n 77) para 478 seems to do this, by referring to Namibia (n 36) and Aegean Sea (n 37) when
interpreting ‘relevant rules’ in art 31.3.c, not accounting for the fact that art 31.3.c was not invoked in Aegean
Sea, and only as a supporting argument in Namibia. Similar reasoning is found in other sources, including
Sinclair (n 48) 139-140; Gabrielle Marceau, ‘A Call for Coherence in International Law: Praises for the
Prohibition Against 'Clinical Isolation' in WTO Dispute Settlement’ (1999) 33(5) J World Trade 87, 120-122;
Pauwelyn (n 141) 265-266; Aust (n 12) 243-244; Stefan Zleptnig, Non-Economic Objectives in WTO Law:
Justification Provisions of GATT, GATS, SPS and TBT Agreements (Martinus Nijhoff Publishers 2010) 75-77;
Vassilis P Tzevelekos, ‘The Use of Article 31(3)(C) of the VCLT in the Case Law of the ECtHR: An Effective
Anti-Fragmentation Tool or a selective Loophole for the Reinforcement of Human Rights Teleology?
Between Evolution and Systemic Integration’ (2010) 31 Michigan J Intl L 621, 660; Matthias Herdegen,
‘Interpretation in International Law’, The Max Planck Encyclopedia of Public International Law (January
2013 edn) <www.mpepil.com> accessed 28 April 2013, para 22. Bugge Thorbjørn Daniel, ‘Chapter 3:
Interpretation, sources of law and precedent’ in Birgitte Egelund Olsen, Michael Steinicke and Karsten Engsig
Sørensen (eds), WTO Law – from a European perspective (Kluwer Law International 2006) 83 writes that art
31.3.c ‘includes evolutionary interpretation’.
146 Gardiner (n 4) 251 and 259, referring to DW Greig, Intertemporality and the Law of Treaties (2001) 46, and
Villiger (n 7) 433 support this. Sinclair (n 48) 139; Marceau (n 145) 120-122; Pauwelyn (n 141) 265; Zleptnig
(n 145) 75 explicitly disagree.
147 Donald H Regan, ‘International Adjudication: A Response to Paulus–Courts, Custom, Treaties, Regimes,
and the WTO’ in Samantha Besson and John Tasioulas (eds), The Philosophy of International Law (OUP 2010)
235 notes that art 31.3.c gives a rule ‘normative significance’ to the interpretation, as distinct from treating it
as ‘empirical evidence’. As Benn McGrady, ‘Fragmentation of International Law or 'Systemic Integration' of
Treaty Regimes: EC – Biotech Products and the Proper Interpretation of Article 31(3)(C) of the Vienna
Convention on the Law of Treaties’ (2008) 42 J World Trade 589, 593 observes, ‘the question of when a
decision maker may take an extraneous treaty into account in treaty interpretation is distinct from the
question of when Article 31(3)(c) binds a decision maker to do so’. Dörr (n 76) 566 makes a similar point, in
that ‘rules extrinsic to the treaty’ may become relevant without the use of art 31.3.c.
148 Dispute regarding Navigational and Related Rights (n 38) para 64 illustrates this point: ‘[…] a meaning or
content capable of evolving, not one fixed once and for all, so as to make allowance for, among other things,
developments in international law’ (emphasis added). Ress (n 6) 25 makes a similar point: ‘the theory of
evolutionary treaty interpretation does not provide for any particular limitation to certain types of legal acts,
declarations, or circumstances’.
VOLUME 6 EJLS ISSUE 1
148
is thus possible even though the rule being invoked is not a formal ‘rule’ in Article 31.3.c’s
sense, and even though it is not binding on ‘the parties’. Similarly, Article 31.3.c may be
invoked in cases where interpreting evolutively is not permissible, notably when the term
being interpreted was not intended to be evolutive.149
While the two concepts must be kept apart at the conceptual level, there is nonetheless
room for interplay at the practical level.150 ‘Rules of international law’ may help determine
whether a term should be interpreted evolutively; if ‘relevant rules of international law’ are
taken to be evolutive, perhaps the term being interpreted should be evolutive as well.
Relevant rules can also be used to determine how an evolving term shall evolve; an evolving
term can be influenced by ‘relevant rules’ to the same extent as static terms.
5. CONCLUSION
This article’s introduction presented two goals: to clear up confusion regarding the concept
of evolutive interpretation, and to deepen our understanding of it.
Confusion is both expressed in and generated by the debate, part lex lata and part lex
ferenda, over whether the concept has a place in international law at all. The lex lata part of
this debate could have been settled by a clear and general statement from the ICJ. The
Court has delivered a statement, which is commendably general and apparently quite clear:
‘[G]eneric’ terms in long-term or indefinite treaties were presumably intented to be
interpreted evolutively. However, the Court is not sufficiently consistent when defining
‘generic’, which means that the debate is not yet completely settled.
Another apparent source of confusion is the tendency to conflate evolutive interpretation
with the (itself somewhat unwieldy) ‘doctrine of intertemporal law’, and with the VCLT
Article 31.3.c. These are distinct from the concept of evolutive interpretation, even though
they may interact with it on a practical level.
In an attempt to deepen our understanding of evolutive interpretation, the article has shown
that the approach to evolutive interpretation seems to vary between three distinct
categories of terms: value driven evolving terms, non-value driven evolving terms, and nonevolving
terms. The article has also shown that evolutive interpretations may help solve
issues of vagueness, but not those of ambiguity.
149 Lorand Bartels, ‘Article XX of the GATT and the Problem of Extraterritorial Jurisdiction: The Case of
Trade Measures for the Protection of Human Rights’ (2002) 36 J World Trade 353, note 8.
150 Namibia (n 36) is an example; art 31.3.c was used as a supporting argument in interpreting an evolving
term (see n 76 above).