international law as a system “international law as an open system”, in j. crawford,...

PaRT II

INTERNATIONAL LAW AS A SYSTEM

CHaPTeR vI

PeRSOnaLITY anD PaRTICIPaTIOn

“It is the world of words that creates the worldof things.”

Jacques Lacan, The Function and Field of Speech and Language

in Psychoanalysis (1966) 386.

A. A Relational Process between Participants

218. In Part I we explored various difficulties with the idea ofinternational law, focusing on the identification of particular rules oflaw — how legal rules could exist in a decentralized society ofsovereigns, how new international law rules could be made by cus-tom and treaty and whether international law rules are sufficientlydeterminate to provide guidance in particular cases. 219. next we will examine a number of questions concerned

with the premise, or assumption, that international law is a systemand not just a set of rules. after all, for all the difficulties relating to their identification and application, no one in their right mindwould deny that there are rules of international law. For example the rule of diplomatic precedence articulated at the Congress ofvienna in 1815 387 and now incorporated in article 14 of the viennaConvention on Diplomatic Relations 388 is a legal rule. But interna-tional law might only be an assortment of such rules — just like the

137

386. In J. Lacan, Ecrits : A Selection (London, Routledge Classics, 2001), p. 72.387. See Final act of the Congress of vienna, vienna, 19 March 1815,

annex XvII, Regulation concerning the Relative Ranks of Diplomatic agents,extracted in “Diplomatic Intercourse and Immunities Memorandum” preparedby the Secretariat, un doc. a/Cn.4/98, 21 February 1956, ILC Ybk 1956/II, p. 133, p. 129.388. vienna Convention on Diplomatic Relations, vienna, 18 april 1961 (in

force, 24 april 1964), 500 UNTS 95.

inhabitants of a village might have a rule that you can pasture cowson the village green in summer, and that you must wear a suit onSunday. Is international law only an assortment or set of rules thathappen to be generally accepted as such, but which have no otherlink ? The rules I have mentioned have nothing in common exceptthat they are accepted in the village. Is international law like this village ? 220. This first chapter of Part II is divided into two parts. In the

first part we investigate further the assumption that international lawis a system and not merely a miscellany of primary rules. The dis-tinction has implications also for how legal rules may be applied.For example, reasoning by analogy is employed in legal systems,premised on an assumption of commonality or coherence. we mustask to what degree such commonality or coherence exists, or canexist, at the international level. 221. The challenge to international law’s systemic character is

that, with the rise of positivism, the making of international law isrepresented as a wholly amorphous process, unconstrained by theaccepted set of natural law beliefs that the founding writers such asGrotius and Pufendorf held and which we no longer hold in com-mon. But I will suggest that international law is a legal systembecause it is a function of a social process between States and otherpersons — a key aspect of the structuring of human relations beyondthe State 389. In these and other ways, it is more than just discon-nected rules. 222. If international law is to be assumed a system, however,

there are five difficulties that need to be addressed, and will beaddressed in the five chapters in Part II. First is the issue of partici-pation, encapsulated in the problem of personality ; second, the dual-ity of international law and national law, expressed in the well-known but problematic formula of the dédoublement fonctionnel(how, paid by one system, can judges loyally serve another ?) ; third,the impossibility of multilateralism when multilateral bonds canapparently always be reduced to disconnected bilateral relations ;fourth, the connected concepts of proliferation, fragmentation andpurportedly self-contained regimes ; and fifth, the extent to which

138 J. Crawford

389. J. Crawford, “International Law as an Open System”, in J. Crawford,International Law as an Open System : Selected Essays (London, Cameron May,2002), preface, p. 13.

international law is truly universal and not merely a function of apassing european predominance.223. In the second part of this chapter, we examine the first of

these difficulties : personality and participation in international law.The core problem may be identified immediately. International lawaffects everyone but participation in the international law system isvery unevenly distributed. Traditionally, only States were consideredsubjects of international law 390, but this is no longer true. The prob-lem is that we have no equivalent simple criterion to replace that ofstatehood. In effect, the criterion or test for whether one participatesin international affairs now is whether one participates in interna-tional affairs — an obvious circularity 391. But first something moreshould be said about the character of international law as a system.

1. Conceptions of a “system” of law

224. International lawyers commonly speak and think of theirsubject as a system. Yet according to one of the most influentialtwentieth-century accounts, they are wrong. H. L. a. Hart saw thefoundations of a legal system in the union of primary rules — thoseconferring rights or imposing obligations — and secondary rules —those determining how the primary rules are made, changed, adju-dicated and enforced 392. while Hart supported the character of international law as “law” (since the primary rules of a simple or“primitive” society can still be counted as law), he doubted itsdescription as a legal system. In his view :

“there is no basic rule providing general criteria of validity forthe rules of international law. . . The rules which are in factoperative constitute not a system but a set of rules, among whichare the rules providing for the binding force of treaties.” 393

Hart owed this idea of a “basic rule” to Kelsen’s Grundnorm. ButKelsen never doubted that international law had a Grundnorm : his

General Course on Public International Law 139

390. L. Oppenheim, International Law : A Treatise (London, Longmans,Green & Co., 1st ed., 1905), p. 18.391. Reparation for Injuries Suffered in the Service of the United Nations,

Advisory Opinion, ICJ Reports 1949, pp. 8-9.392. H. L. a. Hart, The Concept of Law (Oxford, Clarendon Press, 2nd ed.,

1994), pp. 214, 232. 393. Ibid., p. 236. Other grounds include the absence of secondary rules and

an “international legislature, courts with compulsory jurisdiction and centrallyorganized sanctions” : ibid., p. 214.

doubt rather concerned whether the international law Grundnormwas or was not by a form of delegation also the Grundnorm of alllegal systems — the grundest of all norms, so to speak 394. 225. even so, Hart has identified a problem, which Kelsen’s

abstract hypothesis of the Grundnorm fails to resolve. Consider thebreach of a bilateral agreement. The rules of international law,according to article 33 (1) of the un Charter, call for pacific settle-ment of disputes. Depending on the dispute settlement provisions ofthe treaty, the injured State has a right to notify the responsible Stateof its injury and elect a remedy, among other things 395. If an arrange-ment is reached without arbitration or judicial dispute settlement, theproblem has been resolved by the entities themselves, without resortto an external authority 396. But there are few mechanisms for ensur-ing that this self-regulating arrangement is consistent with the frame-work of international law : it might not be, and the articles of thevienna Convention on the Law of Treaties regarding the relationsbetween treaties do not begin to solve that problem 397. 226. But perhaps to require that they should do so a priori is to

ask too much. Joseph Raz does not consider “legal system” a tech-nical term, preferring instead to use it as a way of informing thinkingabout how law works 398. He more loosely characterizes it as “intri-cate webs of interconnected laws” 399 — which international law nodoubt satisfies at least at an elementary level by its large volume ofconventional and customary law, moderated by common rules ofinterpretation.

140 J. Crawford

394. H. L. a. Hart, “Kelsen’s Doctrine of the unity of Law”, in H. L. a. Hart(ed.), Essays in Jurisprudence and Philosophy (Oxford, Clarendon Press, 1983),pp. 309-343 ; J. von Bernstorff, The Public International Law Theory of HansKelsen : Believing in Universal Law (Cambridge, CuP, 2010), pp. 81, 160-165.395. See Responsibility of States for Internationally wrongful acts, ILC Ybk

2001/II (2), art. 43.396. Cf. J. Raz, Concept of a Legal System (Oxford, Clarendon Press, 1970),

pp. 192-194 on primary law-applying organs.397. vienna Convention on the Law of Treaties, vienna, 23 May 1969 (in

force, 27 January 1980), 1155 UNTS 331 (vCLT), arts. 40, 41, 58, 59. In rela-tion to treaty conflicts, for instance, the Convention can ultimately provide onlya “principle of political decision”. See J. Klabbers, Treaty Conflict and theEuropean Union (Cambridge, CuP, 2009), p. 98. This may steer States towardsoutcomes that are consistent with fundamental principles of international law butneither guarantees nor requires it. See S. Ranganathan, Strategically-createdTreaty Conflicts and the Politics of International Law (Cambridge, CuP, forthcoming), Chap. 1.398. J. Raz, The Authority of Law (Oxford, Clarendon Press, 1979), pp. 78-79.399. Raz, Concept of a Legal System, op. cit., p. 183.

227. according to Raz, continuity of a legal system is not neces-sarily disrupted by the creation of new original laws — laws notauthorized by another law — even though emanating from a differ-ent sub-system 400. This is an important point for international law,because treaties with varying membership and different subject matter establish new rules and institutions on the international plane.The wTO did not authorize the establishment of the ICC, nor did theSecurity Council — so in this sense the Rome Statute could be seen as an original law — yet all these institutions are congeners ofinternational law.

2. The coherence challenge

228. although international law lacks institutional infrastructure,the inference that this renders it unable to work in a systematic waycannot be sustained. The International Court, in its advisory Opinionon the WHO Regional Headquarters, rightly observed that a rule ofinternational law “does not operate in a vacuum” but operates with“relation to facts and in the context of a wider framework of legalrules of which it forms only a part” 401. Likewise, Judge Greenwoodasserted in Diallo (Compensation) that international law “is not aseries of fragmented specialist and self-contained bodies of law, eachof which functions in isolation from the others ; it is a single, unifiedsystem of law” 402. 229. This view was also favoured by the International Law

Commission (ILC), which extensively studied challenges of frag-mentation and proliferation raised by the emergence of “technicallyspecialized cooperation networks with a global scope” such as trade, human rights, environment, diplomacy and communications.In examining methods of resolving conflicts between special andgeneral international law, conflicts between successive norms andsystemic integration, Special Rapporteur Koskenniemi provided acomprehensive analysis of topical issues that, according to some tra-ditional views, threaten the notion that international law is a system.


400. Raz, Concept of a Legal System, op. cit., pp. 187-188.401. Interpretation of the Agreement of 25 March 1951 between the WHO

and Egypt, Advisory Opinion, ICJ Reports 1980, p. 76 (para. 10).402. Ahmadou Sadio Diallo (Compensation Owed by the Democratic

Republic of the Congo to the Republic of Guinea) (Republic of Guinea v.Democratic Republic of the Congo), Judgment, ICJ Reports 2012, p. 3 (para. 8).

But the ILC’s conclusion was the opposite : “International law is alegal system” 403, and specialized topics that emerge within interna-tional law do not erode the systematization of international law 404.230. The ILC suggested that the rules and principles of interna-

tional law “act in relation to and should be interpreted against thebackground of other rules and principles”, emphasizing that “inter-national law is not a random collection of such norms” 405. Thecapacity of States to modify the system is also in practice curtailed.as Pauwelyn explains :

“[I]n their treaty relations states can ‘contract out’ of one,more or, in theory, all rules of international law (other thanthose of jus cogens), but they cannot contract out of the systemof international law.” 406

The reason States cannot contract out of the system of internationallaw is not that the system itself is peremptory ; it is that it is neces-sary, if anything lasting is to be achieved at the international level.Imagine for a moment that a State attempts to contract out of thepacta sunt servanda principle 407. It will still have to make promisesin order to achieve its purposes, but other States have no reason totake its promises seriously. Perhaps the State may assure its “co-contracting” party that it intends that particular promise to be bind-ing. But its general disavowal of the binding character of itspromises would apply equally to this assurance. Pacta sunt servandais not a peremptory norm, but it is a necessary one for States’promises to be taken seriously. States are serial promisors and theinstitution of formal promising — treaty making — entails the normpacta sunt servanda. States’ fondness for international law may ebband flow, but the system remains.

142 J. Crawford

403. “Conclusions of the work of the Study Group on the Fragmentation ofInternational Law : Difficulties arising from the Diversification and expansionof International Law”, Report of the International Law Commission on theworks of its Fifty-eighth Session (1 May-9 June and 3 July-11 august 2006),un doc. a/61/10, Supp. no. 10, p. 407 (para. 251). 404. Ibid., p. 405 (para. 245).405. Ibid., p. 407 (para. 251). 406. J. Pauwelyn, Conflict of Norms in Public International Law : How WTO

Law Relates to Other Rules of International Law (Cambridge, CuP, 2003), p. 37.407. M. Koskenniemi, “Fragmentation of International Law : Difficulties

arising from the Diversification and expansion of International Law”, Report ofthe Study Group of the International Law Commission, Fifty-eighth Session (1 May-9 June and 3 July-11 august 2006), un doc. a/Cn.4/L.682, p. 93 (para. 176).

3. The relation between international law and international society :transnational rules lacking societies

231. Raz makes the vital point that legal systems are not distinctand separate from the societies that create and apply them. Legalsystems are complex forms of life and involve a process of socialinteraction 408, through which shared legal norms are created andmaintained over time. at the international level, patterns of socialinteraction are held out to be legitimate or proper, and are treated asoperating within a system of international law 409.232. It is true that attempts have been made to “constitute”

systems of law divorced from social systems. a notable example is the idea of lex mercatoria postulated by French comparatistBerthold Goldman 410 and developed further by institutions such asunIDROIT 411. The problem with it is that it is a pure confection,unrelated to any real source of authority or any existing praxis. It is a law of and for professors, a Buchrecht reduced to a singlebook, based on the assumption that comparative law techniques candistil a true or real underlying common law — a sort of natural lawwithout the benefit of divinity. The assumption is demonstrablyuntrue. 233. Perhaps the most comprehensive attempt to distil an interna-

tional common core of contract law, applicable to private actors, wasthe 10-year long comparative assessment of the law relating to formation of contracts conducted at Cornell university by Rudolf


408. See, e.g., Raz, Concept of a Legal System, op. cit., p. 188. 409. J. Bruneé and S. J. Toope, Legitimacy and Legality in International Law :

An International Account (Cambridge, CuP, 2010), pp. 350-352. 410. B. Goldman, “Frontières du droit et lex mercatoria” (1964) 9 Archives

de philosophie du droit 177. See also P. Kahn, Vers la quête de la LexMercatoria : l’apport de l’école de Dijon, 1957-1964. In 2009 the Centre forTransnational Law at the university of Bologna established an online resourcecataloguing lex mercatoria principles and accompanying primary materials,called “TransLex Principles”. The database currently contains a non-exhaustivelist of 135 principles of transnational commercial law. See Transnational LawDatabase, “Purpose” and “Concept” <http ://trans-lex.org/content.php ?what=8>.411. unIDROIT Principles of International Commercial Contracts, 1994

<http ://www.unidroit.org/english/principles/contracts/main.htm>. unCITRaLhas endorsed the use of the Principles. See Report of the united nationsCommission on International Trade Law on the work of its Forty-fifth Session,new York, 25 June-6 July, un doc. a/67/17, pp. 33-34 (paras. 137-140).another attempt at codification is the Trento Project : u. Mattei and M. Bussani(eds.), The Common Core of European Private Law (London, Kluwer LawInternational, 2003).

Schlesinger 412, with the aim to “enhance professional knowledge . . .by finding and formulating the common ground as well as the differences among legal systems” and to test the feasibility of itsresearch method 413. The study concluded with “haphazard” if notrather abstract observations, among which was the “expected” find-ing that “the areas of agreement are larger than those of disagree-ment” but that areas of disagreement were more complex than anti-cipated 414. Despite the intelligent choice of topic — after all, forma-tion of contract is prior to contract — this outcome was disappoint-ing ; an “insular study of a peripheral subject”, in the words of onereviewer 415. Schlesinger’s own reflective question after 10 years’work, “did we merely demonstrate the obvious ?”, supports the sceptics of a synoptic universal contract law.234. More recently, Ole Lando has also pursued the quest for

european systemic legal integration through the Principles of euro-pean Contract Law project, born of the european Parliament’s desireto establish a common european civil law 416. The Lando Commis-sion seek to forge a new european jus commune and elucidate amodel set of “general rules” for a uniform regional contract law system to be applied in european business-to-business and business-to-consumer contracts 417. The project has grown in ambition, leadingto the proposed Draft Common Frame of Reference, which purportsto codify the whole of european contract law 418. 235. There are enormous difficulties with the idea that such

“codes” can identify an “essence” distilled from the different legal

144 J. Crawford

412. R. B. Schlesinger et al., Formation of Contracts. A Study of the CommonCore of Legal Systems (London, Stevens & Sons, 1968), vols. 1, 2. 413. Ibid., pp. 1-4, 20-30. 414. Schlesinger, Formation of Contracts. A Study of the Common Core of

Legal Systems, op. cit., pp. 41-42. 415. a. Rubin, “Book Review” (1970-1971) 50 Oregon L. Rev. 99.416. Private Law of the Member States, Resolution on action to Bring into

Line the Private Law of the Member States, Doc. a2-157/89, 26 May 1989, OJ1989 C 158/400. 417. O. Lando and H. Beale (eds.), Principles of European Contract Law :

Parts I and II (The Hague, Kluwer, 2000), art. 1 :101 (1). See also O. Lando andH. Beale (eds.), Principles of European Contract Law : Part III (The Hague,Kluwer, 2003). 418. C. von Bar et al. (eds.), Principles, Definitions and Model Rules of

European Private Law : Draft Common Frame of Reference (Study Group on aeuropean Civil Code, 2009) <http ://ec.europa.eu/justice/contract/files/european-private-law_en.pdf>. See also Decision 2010/233/eu of 26 april 2010 onSetting up the expert Group on a Common Frame of Reference in the area ofeuropean Contract Law, OJ 2010 L 105/109.

systems which could be applied in transnational cases 419. For onething, the assumption is that the detailed examination of legal systems will produce a common core. In practice, the more compara-tive lawyers search for a common core, the more elusive it seems tobe. as Martin Shapiro notes : “I do not know whether ‘commoncore’ is a lesser or greater acknowledgement of defeat than ‘prin-ciples’ but surely it is in the same neighbourhood.” 420 In fact it is inno neighbourhood at all. Lex mercatoria is not the law of any indi-vidual human society, still less a society of merchants or even ofprofessors, it is an abstraction that still fails to remain insulated fromthe legal and policy preferences of its proponents.

4. International law is a system, even if imperfect

236. In short, legal systems are social systems. while it is nodoubt good for international and comparative lawyers to socialize,even a willing society is not only composed of cocktail parties.There is an international social process, defective no doubt, but existent, made up of the actions, interactions and programmes ofGovernments and significant others.237. The working of international law reveals a “complete” sys-

tem of laws 421, albeit one that cannot be uncritically analogized todomestic legal systems 422. It is a legal system because it is a func-tion of the social process between States and other persons regardingmatters of common concern. as Ian Brownlie notes, internationallaw provides both the vocabulary and underlying grammar of inter-state relations 423. Indeed, it is impossible to imagine international


419. Scelle comments “for the rule of law and for legislation to exist, theremust first be society ; this social reality is anterior and superior to any legalright”, in G. Scelle, Précis du droit de gens (Paris, Sirey, 1932), vol. 2, p. 16.See also “The new Lex Mercatoria Does not Have the Quality of an‘autonomous Legal System’ ”, in K. P. Berger, The Creeping Codification of theNew Lex Mercatoria (alphen aan den Rijn, Kluwer Law International, 2nd ed.,2010), Chap. G. See also G. R. Delaume, “Comparative analysis as a Basis ofLaw in State Contracts : The Myth of the Lex Mercatoria” (1988-1989) 63Tulane L. Rev. 575.420. M. Shapiro, “The Common Core : Some Outside Comments”, in

u. Mattei and M. Bussani (eds.), The Common Core of European Private Law(London, Kluwer Law International, 2003), p. 223. 421. R. Jennings and a. watts, Oppenheim’s International Law (Harlow,

Longman, 9th ed., 1992), vol. 1, p. 12.422. J. Crawford, Brownlie’s Principles of Public International Law (Oxford,

OuP, 8th ed., 2012), p. 15.423. See I. Brownlie, “The Reality and efficacy of International Law” (1981)

52 (1) BYIL 1-2.

diplomacy and relations without international law 424. Internationallaw has the characteristics of a system, not just a random collectionof rules : notably, the basic constructs of personality, sources,treaties, interpretation and responsibility 425. The system is a super-structure, if you like, but only in the Marxist sense that all law is asuperstructure 426. It is made up of conventional and customary rulesthat interact, interlink, reinforce each other and sufficiently cohere.Those who worry about its institutional deficiencies may recallRichard Gardiner’s proposition that the institutional framework forsecuring observance of most international obligations is diplomacyand then international organizations 427 — with an increasinglyimportant role for international courts and tribunals.238. But if international law is a social system, who are its partici-

pants ? This brings us to the second part of this chapter and the problem of international personality.

B. The Problem of International Personality

239. It is now generally accepted that under international law nat-ural persons, and certain legal persons, may be beneficiaries of rightsand holders of obligations. But this has not always been so. In 1904,international law was canonically seen as a system designed exclu-sively by States exclusively for States — as a system with 73 par-ticipants, 74 including the Pope 428. The expansion of the personalscope of international law since that time is remarkable for manyreasons, not least because it acknowledges that natural persons areentitled to protection and that participation in the system is no longer

146 J. Crawford

424. G. Simpson, “International Law in Diplomatic History”, in J. Crawfordand M. Koskenniemi (eds.), The Cambridge Companion to International Law(Cambridge, CuP, 2012), p. 45.425. Crawford, Brownlie’s Principles of Public International Law, op. cit.,

p. 16.426. K. Marx, A Contribution to the Critique of Political Economy (trans.

n. I. Stone, Chicago, Charles H. Kerr, 1911), preface ; a. Stone, “The Place ofLaw in the Marxian Structure-Superstructure archetype” (1985) 19 Law & Soc.Rev. 39-68. 427. R. K. Gardiner, International Law (Harlow, Longman, 2003), p. 4. 428. See Oppenheim, op. cit., for a list of States at that time. The territorial

status of the Papacy was at the time unresolved : the vatican City was not estab-lished until the Lateran Treaty (Italy-Holy See), Rome, 11 February 1929 (inforce, 7 June 1929), (1929) 23 AJIL Supp. 187 ; 130 BFSP 791. See J. Crawford,The Creation of States in International Law (Oxford, OuP, 2nd ed., 2006), pp. 221-233.

limited to States. But it also brings new challenges, a selection ofwhich include : ascertaining the criteria for statehood in internationallaw and their connections to the institution of recognition ; managingthe horizontal effects of actors other than States in the internationalsystem ; and establishing mechanisms to hold non-State participantsto account.

1. Recognition and subjectivity of legal personality

240. In Chapter III we discussed the relationship betweensovereignty and statehood. Statehood is essentially a product ofinternational law, but there is enduring debate about the legal criteriafor statehood. The standardly-cited Montevideo formula lists as criteria defined territory, permanent population, stable governmentand the capacity to enter into relations with other States 429. This hasmultiple defects : it is inaccurate (territory may not be defined), ten-dentious (whether a population is permanent depends partly onwhether the people have their own State) and question-begging (thephrase “other States” assumes what it seeks to demonstrate, i.e. thatwe already know what a State is). But above all it is incomplete inthree crucial respects. First, it leaves out independence from otherStates, which is nonetheless (once securely established) the “deci-sive criterion of statehood” 430. Second it ignores the role of inter-national law in determining or precluding statehood in certain impor-tant cases 431 : these are the exceptional situations where internationallaw either declines to accept a for-the-time-being effective entity asa State for reasons of its substantive illegality — as was the casewith Manchukuo 432, Southern Rhodesia 433 and the “bantustans” 434

— or accepts as continuing in existence entities which are for-the-time-being suppressed — as with ethiopia in 1935 435, the Baltic


429. Convention on the Rights and Duties of States adopted by the SeventhInternational Conference of american States, Montevideo, 26 December 1933(in force, 26 December 1934), 165 LNTS 19, art. 1 (Montevideo Convention). 430. C. Rousseau, Droit international public (Paris, Sirey, 1974), vol. 2,

pp. 67-68. For an analysis of external sovereignty (as independence) seeCustoms Regime between Germany and Austria, Advisory Opinion, (1930) PCIJ,Ser. A/B, No. 41, p. 57 (Judge anzilotti) ; Crawford, The Creation of States inInternational Law, op. cit., pp. 62-88. 431. Crawford, The Creation of States in International Law, op. cit., pp. 196-

252.432. Ibid., pp. 75, 78, 132-133.433. Ibid., pp. 128-131.434. Ibid., pp. 47, 341-345.435. Ibid., pp. 97, 311 ff., 402, 626.

States after 1940 436, Kuwait under Iraqi occupation in 1990 437, andsome others. Third, it says nothing whatever about recognition byother States ; but such recognition is undeniably relevant to state-hood, whatever “theory” of recognition one may choose to adopt.241. In truth the best theory of recognition may be none at all.

The declaratory theory (the theory that statehood is independent ofrecognition) overlooks the great significance attributed to recogni-tion in the practice of States. If recognition did not matter, Taiwanwould be a State, Somaliland would be a State ; yet according to thestandard view neither is. On the other hand, as under the constitutivetheory, if recognition were decisive then Kosovo (97 recognitions)and Palestine (132 recognitions) would be States — whereas thereare still questions about both. Indeed if recognition even by oneState were decisive, so would the Turkish Republic of northernCyprus (1 recognition), South Ossetia (2 recognitions) and abkhazia(2 recognitions). Ian Brownlie compares such theories to a “bank offog on a still day” : they obscure rather than illuminate 438. In thesetheories the complexity of legal issues in international relations “hasbeen compacted into a doctrinal dispute between the ‘declaratory’and ‘constitutive’ views” 439, without adding to the sum of humanknowledge. Plainly recognition is a political process interposed byunilateral assessment by a State deciding whether or not it will recognize another as a State and thereby engage in normal politicaland legal relations with that other State. 242. But recognition should not be construed as a magical spell

that gives life to an unborn State 440. unlike Hamlet’s proclamationto Rosencrantz, it cannot be said that whether a State does or doesnot exist, individual recognition makes it so 441. Here again theMontevideo Convention is defectively formulated : article 3 declares

148 J. Crawford

436. Crawford, op. cit. footnote 431, pp. 80, 97, 177, 393-395, 689, 703-705.See also J. Crawford, “Preface” (2001) 1 Baltic YB Int’l. L.437. Crawford, The Creation of States in International Law, op. cit., pp. 105,

162, 439, 688-689.438. I. Brownlie, “Recognition in Theory and Practice”, in R. St. J. Mac-

donald and D. M. Johnston (eds.), The Structure and Process of InternationalLaw : Essays in Legal Philosophy, Doctrine and Theory (The Hague, nijhoff,1983), p. 107. 439. Crawford, Brownlie’s Principles of Public International Law, op. cit.,

pp. 144-145 ; S. Talmon, “The Constitutive versus the Declaratory Theory ofRecognition : Tertium Non Datur?” (2004) 75 BYIL 101-102 ; a. Clapham,Brierly’s Law of Nations (Oxford, OuP, 7th ed., 2012), p. 151.440. Brownlie, “Recognition in Theory and Practice”, op. cit., pp. 107-108. 441. with apologies to Shakespeare. Hamlet, act 2, Scene 2, ll. 249-251.

that “[t]he political existence of the state is independent of recog-nition by the other states” 442 : it had better said, independent ofrecognition by any individual third State.243. what of legal barriers to recognition, or conversely, the duty

to recognize 443 ? In the Kosovo Opinion the International Court wascareful not to say more about controversial issues of contested inde-pendence than was necessary for the purposes of answering thequestion put by the General assembly, as to whether the KosovarDeclaration of Independence was unlawful. In particular it declined(as the Canadian Supreme Court had earlier declined 444) to saywhether the idea of “remedial secession” has any purchase in inter-national law as an extension of the principle of self-determination.Some Governments (notably Germany) had supported the idea,which was prima facie more applicable to Kosovo in view of itsrecent history than it would have been to Quebec 445. The Court said :

“During the eighteenth, nineteenth and early twentieth cen-turies, there were numerous instances of declarations of inde-pendence . . . Sometimes a declaration resulted in the creationof a new State, at others it did not. In no case, however, doesthe practice of States as a whole suggest that the act of promul-gating the declaration was regarded as contrary to internationallaw. On the contrary, State practice during this period pointsclearly to the conclusion that international law contained noprohibition of declarations of independence. During the secondhalf of the twentieth century, the international law of self-deter-mination developed in such a way as to create a right to inde-pendence for the peoples of non-self-governing territories andpeoples subject to alien subjugation, domination and exploita-tion . . . a great many new States have come into existence asa result of the exercise of this right. There were, however, alsoinstances of declarations of independence outside this context.


442. Montevideo Convention, art. 3 (emphasis added). See also Crawford,The Creation of States in International Law, op. cit., p. ix ; J. Crawford, “Intro-duction to the Paperback edition”, in H. Lauterpacht, Recognition in Inter-national Law (Cambridge, CuP, 2013), p. xxxii. 443. On which see Lauterpacht, Recognition in International Law, op. cit. ;

P. Capps, “Lauterpacht’s Method” (2012) 82 BYIL 248-280.444. Reference re Secession of Quebec, [1998] 2 SCR 217, p. 70 (para. 124).445. Accordance with International Law of the Unilateral Declaration of

Independence in Respect of Kosovo, Advisory Opinion, ICJ Reports 2010, p. 438(paras. 82-83).

The practice of States in these latter cases does not point to theemergence in international law of a new rule prohibiting themaking of a declaration of independence in such cases.” 446

after referring to cases of Security Council-mandated non-recogni-tion (Southern Rhodesia, northern Cyprus), the Court continued :

“in all of those instances the Security Council was making adetermination as regards the concrete situation existing at thetime that those declarations of independence were made ; theillegality attached to the declarations of independence thusstemmed not from the unilateral character of these declarationsas such, but from the fact that they were, or would have been,connected with the unlawful use of force or other egregiousviolations of norms of general international law, in particularthose of a peremptory character (jus cogens). In the context ofKosovo, the Security Council has never taken this position.” 447

For the purposes of the answering the question this short essay ongeneral international law was sufficient.244. Beyond the class of well-established entities with legal per-

sonality are entities that are proximate to States or are treated likeStates for certain purposes. Take Taiwan as an example. while it isnot recognized as a State by anyone 448, it has proto-State inter-national legal identity for various purposes, including as a fishingentity under the Law of the Sea Convention 1982 449 and as a sepa-

150 J. Crawford

446. ICJ Reports 2010, p. 436 (para. 79).447. Ibid., p. 438 (para. 81).448. Crawford, Brownlie’s Principles of Public International Law, op. cit.,

p. 125.449. a. Serdy, “Bringing Taiwan into the International Fisheries Fold : the

Legal Personality of a Fishing entity” (2004) 75 BYIL 183-184, 189. whetherTaiwan was a “State” eligible for membership to various regional fisheries man-agement organizations was side-stepped by innovative legal mechanics thatenabled “any entity or fishing entity” to apply for membership to the extendedCommission for the Conservation of Southern Bluefin Tuna. See, e.g., agree-ment for the Implementation of the Provisions of the united nations Conventionon the Law of the Sea of 10 December Relating to the Conservation andManagement of Straddling Fish Stocks and Highly Migratory Fish Stocks, undoc. a/COnF.164/37, 8 September 1995, arts. 1 (2) (b), 1 (3), 8, 17 ; Conven-tion for the Conservation of Southern Bluefin Tuna, Canberra, 10 May 1993 (inforce, 20 May 1994), 189 UNTS 360, art. 18 ; Resolution to establish anextended Commission and an extended Scientific Committee and Rules of Pro-cedure of the extended Commission for the Conservation of Southern BluefinTuna, paras. 1, 6, adopted at the Seventh annual Meeting of the Convention forthe Conservation of Southern Bluefin Tuna, Sydney, 18-21 april 2001.

rate customs territory for the purposes of wTO membership 450.Somaliland may be another example. even though it is a self-declared, yet contested, secessionist region, it has maintained de factoindependence for a considerable period of time and participates onthe international plane, for example as a refugee-receiving entity 451. 245. Palestine is a more controversial case 452. The status of

Palestine (west Bank and Gaza Strip) remains subject to vehementdebate. even though the people of Palestine are undoubtedly entitledto full self-determination, Israel has long denied its statehood pend-ing resolution of a range of issues through “permanent status nego-tiations”, for a long time stalled 453. Palestine’s request for full un membership is undetermined but it has recently been accorded “non-member observer State” status by the General assembly 454 —the same status as the Holy See. uneSCO’s earlier admission ofPalestine resulted in the withdrawal of uS funding to that organi-zation 455. But with 132 States now recognizing Palestine as a State,it seems to be eking its way toward statehood456.

2. Participation and effect at the horizontal level

246. The international legal system is further enriched — or com-plicated — by the participation of entities without any territorial


450. Marrakesh agreement establishing the world Trade Organization,Marrakesh, 15 april 1994 (in force, 1 January 1995), 1867 UNTS 3, art. 12. 451. See Judgment of 30 October 2003, 3 ue 4952/96.a (Hesse adminis-

trative Court, Kassel, Germany). The Court determined that Somaliland is a“State” for the purposes of asylum law ; Report of the Secretary-General on thesituation in Somalia, un doc. S/2004//469, 9 June 2004, p. 8 (para. 47). Seegenerally unHCR Global appeal 2012-2013, “Somalia” <http ://www.unhcr.org/4ec230fc0.pdf> accessed 5 March 2013. See also Crawford, The Creation ofStates in International Law, op. cit., pp. 375, 412-414, 417. 452. See J. Crawford, “Israel (1948-49) and Palestine (1989-1999) : Two

Studies in the Creation of States”, in G. S. Goodwin-Gill and S. Talmon (eds.),The Reality of International Law : Essays in Honour of Ian Brownlie (Oxford,OuP, 1999), pp. 95-124.453. Declaration of Principles of Interim Self-Government arrangements,

13 September 1993, 32 ILM 1527. 454. un doc. a/67/L.28, 25 november 2012, para. 2 ; Ga res. 67/19, 29 nov-

ember 2012, p. 3 (para. 2). 455. Public Law 103-236, § 410 (1), 30 april 1994, 108 Stat. 454 (1994) pro-

hibits the united States making any voluntary or assessed contribution to “anyaffiliated organisation of the united nations which grants full membership as astate to any organization or group that does not have the internationally recog-nized attributes of statehood”. 456. united nations educational, Scientific and Cultural Organization,

“General Conference admits Palestine as uneSCO Member” (Press Release, 31 October 2011).

base of their own 457. The International Court first faced up to thisquestion in the Reparation case, holding that :

“The subjects of law in any legal system are not necessarilyidentical in their nature or in the extent of their rights . . .Throughout its history, the development of international lawhas been influenced by the requirements of international life,and the progressive increase in the collective activities of Stateshas already given rise to instances of action upon the interna-tional plane by certain entities which are not States . . . In theopinion of the Court, the Organization was intended to exerciseand enjoy, and is in fact exercising and enjoying, functions andrights which can only be explained on the basis of the posses-sion of a large measure of international personality and thecapacity to operate upon an international plane . . . theOrganization is an international person . . . a subject of interna-tional law and capable of possessing international rights andduties, and that it has capacity to maintain its rights by bringinginternational claims.” 458

It is a mark of international law’s conservatism that it took until1949 for the separate legal personality of an international legal orga-nization like the united nations to be recognized. The status of itspredecessor, the League of nations, for example, remained unsettled,such that the Secretary-General of the League apparently had to signcontracts of employment in his personal capacity. 247. The question following Reparation was whether this idea

of the legal personality of international organizations was limited to organizations having universal aspirations such as the unitednations or whether it applied more widely. The International Courtleft this question open, perhaps implying the narrower view, but in

152 J. Crawford

457. See, e.g., C. Brölmann, “Deterritorialization of International Law”, in J. nijman and a. nollkaemper (eds.), New Perspectives on the Divide betweenNational and International Law (Oxford, OuP, 2007).458. Reparation for Injuries Suffered in the Service of the United Nations, op.

cit., pp. 178-179. In Jurisdiction of the European Commission of the Danubebetween Galatz and Braila, Advisory Opinion, (1927) PCIJ, Ser. B, No. 14, pp. 63-64 the Permanent Court “hinted at the possibility of international institu-tions being bestowed with functionally limited international personality”, butthis was not quoted in Reparation : R. Portmann, Legal Personality inInternational Law (Cambridge, CuP, 2010), p. 108 ; K. Parlett, The Individual inthe International Legal System : Continuity and Change in International Law(Cambridge, CuP, 2011), pp. 31-32.

practice a very large number of organizations have been establishedby States, sometimes even on a bilateral basis, that are said to haveand seem to be accepted as having international legal personality 459. 248. So just as has occurred with the corporate form at the

national level, the separate corporate personality of internationalorganizations has become a basic tool, widely available to States andto organizations themselves (some international organizations havecreated separate international organizations 460). International organi-zations are a manifestation of collective State action to addressshared concerns, including public health, environment, trade and climate change. To advance common interests international organi-zations establish international standards (e.g., International Telecom-munication union, wHO, wTO), impose mandatory action on States(e.g., the Security Council) and implement and enforce rules (e.g.,international courts and tribunals) 461.249. The next question is whether the principle of international

legal personality declared by the International Court extends furtherthan international organizations of an inter-State character. InBarcelona Traction the International Court expressed the position inthe following way : “international law has had to recognize the corporate entity as an institution created by States” that has becomea “powerful factor in the economic life of nations” 462.


459. There is no definitive list of international organizations : C. F. amera-singhe, Principles of the Institutional Law of International Organisations(Cambridge, CuP, 2nd ed., 2005), p. 5. The International Yearbook of Interna-tional Organisations (2009/2010) states that there were at that point 241 “con-ventional” intergovernmental organizations : Figure 2.9. 460. For example, the member States of the Organisation for economic

Co-operation and Development established the International energy agency asan autonomous body within the framework of the Organisation, in response tothe oil crisis at that time, pursuant to its Decision of the Council establishing theInternational energy agency of the Organisation, OeCD Council, 373rd meet-ing, 15 november 1974 (in force, 15 november 1974), art. 1. 461. That international organizations may enter into treaties is now widely

accepted : vienna Convention on the Law of Treaties between States and Inter-national Organizations or between International Organizations, vienna, 21 March1986 (not in force), 25 ILM 543. not all organizations that operate on the inter-national plane unequivocally possess international legal personality, and thosethat do may have restricted competence : e.g., the earlier Commonwealth ofnations ; un Conference on Trade and Development ; the High Commissionerfor Refugees. See, e.g., J. e. S. Fawcett, British Commonwealth in InternationalLaw (London, Stevens & Sons, 1963) ; w. Dale, “Is the Commonwealth anOrganisation” (1982) 31 ICLQ ; F. Morgenstern, Legal Problems of InternationalOrganizations (Cambridge, Grotius, 1986), pp. 23-26. 462. Barcelona Traction, Light and Power Company, Limited (Belgium v.

Spain), Judgment, ICJ Reports 1970, pp. 33-34 (paras. 38-39) (emphasis added).

250. It is true that there are examples of domestic associationsthat came to be accepted as international persons. Thus theInternational Committee of the Red Cross (ICRC) was formed as aprivate humanitarian initiative, sanctioned by a diplomatic confer-ence in 1863. In 1915 it was formally established as a Swiss organi-zation 463. It has now developed a dual character. From an organiza-tional point of view it remains a Swiss entity, yet it is increasinglyrecognized as performing public international functions and it isclear that the ICRC is as a separate international legal person inrespect of the exercise of those functions — a status recognized bythe Red Cross Conventions of 1949, by the Security Council and (inthe context of non-disclosure of information) by the ICTY 464. 251. another example, also Swiss, is the Bank for International

Settlements (BIS), established in 1930 at a time when it was notclear that a bank could be established by international law andbecome an international person 465. In 1987 the BIS was recon-stituted and recognized as an international legal person 466 ; it has

154 J. Crawford

463. G. Barile, “Caractère du Comité international de la Croix-Rouge” (1979)62 Riv diritto internazionale 112 ; D. Bindschedler-Robert, “Red Cross” (1983)5 EPIL 250 ; Y. Beigbeder, The Role and Status of International HumanitarianVolunteers and Organizations (Dordrecht, Martinus nijhoff, 1991), p. 66 ; C. Koenig, “Observer Status for the International Committee of the Red Cross atthe united nations : a Legal viewpoint” (1991) 280 Int. Rev. Red Cross ; a. Schlögel, “International Red Cross”, in R. wolfrum (ed.), United Nations :Law, Policies and Practice (Dordrecht, Martinus nijhoff, rev. ed., 1995), vol. 2, pp. 814-815 ; H-P. Gasser, “International Committee of the Red Cross (ICRC)”,in R. wolfrum (ed.), The Max Planck Encyclopedia of Public International Law(OuP, 2009, online). 464. agreement between the International Committee of the Red Cross and

the Swiss Federal Council to Determine the Legal Status of the Committee inSwitzerland, 19 March 1993 ; Ga res. 45/6, 16 October 1990 ; Statutes of theInternational Red Cross and Red Crescent Movement (as amended), art. 5 ;Rules of Procedure and evidence for the application of the Rome Statute of theInternational Criminal Court, art. 73 ; Prosecutor v. Simic (Decision on theProsecution Motion under Rule 73 for a Ruling concerning the Testimony of aWitness), IT-95-9, (1999) § 46 ; SC res. 770, 13 august 1992, para. 2 ; SCres. 771 (1992), 13 august 1992, paras. 4, 5 ; SC res. 775 (1992), 28 august1992 ; SC res. 776 (1992), 14 September 1992, para. 2 ; SC res. 814 (1993), 26 March 1993, para. 9 ; SC res. 1010 (1995), 10 august 1995, para 1 ; SC res.1019 (1995), 9 november 1995, para. 2 ; SC res. 1034 (1995), 21 December 1995,paras. 4, 5 ; SC res. 1199 (1998), 23 September 1998, paras. 4 (c), 5 (d), 5 (e) ;SC res. 1398 (2002), 15 March 2002, Preamble (para. 4), para. 13 ; SC res. 1430(2002), 14 august 2002, Preamble (para. 6) ; SC res. 1434 (2002), 6 Septem-ber 2002, Preamble (para. 3) ; SC res. 1466 (2003), 14 March 2003, Preamble(para. 3). 465. See M. Jacob, “Bank for International Settlements”, in R. wolfrum (ed.),

The Max Planck Encyclopedia of Public International Law (OuP, 2008, online). 466. Ibid.

since concluded treaties and been sued at the international level467.Dozens of international and regional banks now exist. 252. But the category of corporate actors remains an ad hoc one.

So far, several other corporate entities have been treated differently.The International Court has always insisted that corporations are created and regulated by States even if they operate well beyond thelimits of the State of incorporation ; it said so in Anglo Iranian Oil 468, and maintained that position in Barcelona Traction 469 and latterly Diallo 470. 253. elsewhere, there is some tendency to elevate the status of

transnational corporations, with the apparent aim of increasing theiraccountability. numerous treaties impose obligations on States toregulate the conduct of corporations ; and corporations have invokedhuman rights under regional human rights treaties 471. But thereappear to be no express obligations in international human rights lawimposed directly on corporations : the development of a general lawof international responsibility has focused almost exclusively onStates and, by derivation, international organizations. Insofar asinternational law has recognized direct criminal responsibility, it hasdone so with respect to individuals, not corporations. none of theconstituent instruments of the international criminal tribunals pro-vide for corporate criminal responsibility : “no international criminaltribunal has had jurisdiction to try a company as a legal entity forcrimes under international law” 472. 254. This may appear to be unsatisfactory. Some corporate

entities are immensely powerful : they generate revenue larger than the GDP of many States, operate on a global basis and exercise


467. Reineccius et al. v. Bank for International Settlements, Final Award, 19 September 2003, 140 ILR 1. 468. Anglo Iranian Oil Co. (United Kingdom v. Iran), Preliminary

Objections, Judgment, ICJ Reports 1952, pp. 102, 112. 469. Barcelona Traction, Light and Power Company, Limited, op. cit., pp. 33-

34 (paras. 38-39).470. Ahmadou Sadio Diallo (Republic of Guinea v. Democratic Republic of

the Congo), Preliminary Objections, Judgment, ICJ Reports 2007, pp. 605-606(paras. 61-65) ; Ahmadou Sadio Diallo (Republic of Guinea v. DemocraticRepublic of the Congo), Merits, Judgment, ICJ Reports 2010, pp. 673, 675-676(paras. 99, 104-105). 471. See, e.g., The Sunday Times v. United Kingdom (no. 2), app. no.

13166/87, 26 november 1991 ; Observer and Guardian v. United Kingdom, app.no. 13585/88, 26 november 1991.472. International Commission of Jurists, Corporate Complicity & Legal

accountability Report, vol. 2 : Criminal Law and International Crimes (2008),p. 6.

considerable influence over domestic affairs. Influence, however, isnot the same as legal capacity on the international plane, and theInternational Court’s insistence that corporations are creatures ofdomestic law is underwritten by powerful policy concerns to avoidthe creation of wholly unaccountable entities. John Ruggie, SpecialRapporteur of the Secretary-General on the issue of human rightsand transnational corporations, has noted the emerging “governancegaps created by globalisation”, stating :

“The international community is still in the early stages ofadapting the human rights regime to provide more effectiveprotection to individuals and communities against corporate-related human rights harm.” 473

In 2011 the un Human Rights Council endorsed Ruggie’s “imple-menting” framework on business and human rights, which serves toimprove international accountability and responsibility through threepillars :

“The first is the State duty to protect against human rightsabuses by third parties, including business enterprises, throughappropriate policies, regulation, and adjudication. The secondis the corporate responsibility to respect human rights, whichmeans that business enterprises should act with due diligence toavoid infringing on the rights of others and to address adverseimpacts with which they are involved. The third is the need forgreater access by victims to effective remedy, both judicial andnon-judicial.” 474

255. Other attempts to improve accountability of corporationsand international actors are found in soft law forms, including volun-tary codes, agreements and declarations. Such instruments eithercontain broadly defined standards applicable to transnational acti-vities of corporations (or businesses more broadly) 475 or they areaimed at particular industries 476. while these instruments have the

156 J. Crawford

473. un doc. a/HRC/8/5, 7 april 2008, p. 3 (paras. 1, 3). 474. un doc. a/HRC/17/31, 21 March 2011, p. 4 (para. 6) ; un doc.

a/HRC/ReS/17/4, 16 June 2011, p. 2 (para. 1).475. For example, the OeCD Guidelines for Multinational enterprises,

21 June 1976, 15 ILM 967 ; un Global Compact, 26 July 2000 ; un norms on theResponsibilities of Transnational Corporations and Other Business enterprises withRegard to Human Rights, 13 august 2003, un doc. e/Cn.4/Sub.2/2003/12 (2003).476. For example, the extractive Industries Transparency Initiative <http ://

www.eiti.org/eiti/principles>.

potential to increase public awareness of human rights issues, theydo not treat international actors as legal entities. 256. at the domestic level, there has been no generalized develop-

ment of a concept of corporate responsibility for violations of, orcomplicity in State violations of, international human rights law. Theprincipal exception has been the uS alien Tort Claims act 477. Butthe concern here is with legal personality, not powers or obligations.no one denies that States may subject corporations properly withintheir jurisdiction to regulation that might — if thought useful —extend to liability for breach of international norms. It is anotherthing to treat corporations as addressees of those norms as such. TheSupreme Court in Kiobel sidestepped these issues and curtailed theact’s presumed extraterritorial application, thereby largely barringpetitioner action for breaches of the law of nations occasionedagainst them when outside uS territory 478.257. a similar issue arises with respect to non-corporate indivi-

duals like you and me. The primary expansion of legal subjectsunder international law is the recognition and protection of indivi-duals and peoples through human rights law — a sub-category of indi-vidual rights 479. It is now incumbent upon States to satisfy a range ofpositive and negative obligations, such as to provide a fair trial andto refrain from unlawful killing and torture 480. But human rights law


477. 28 uSC, § 1350. There is also a possibility of such claims being broughtby uS nationals under the Torture victims Protection act for more limited violations of international law : Public Law 102-256, 106 Stat. 73 (1992), codi-fied as a note to 28 uSC, § 1350. See generally S. Joseph, Corporations andTransnational Human Rights Litigation (Oxford, Hart Publishing, 2004).478. Kiobel, Individually and on Behalf of Her Late Husband et al. v. Royal

Dutch Petroleum Co. et al., 569 uS (2013), pp. 1-3, 14. earlier case-law, starting with Filártiga v. Peña-Irala, 630 F. 2d 876 (2nd Cir. 1980), hadexpanded the scope of operation of the act, permitting considerable humanrights litigation. Cf. Sosa v. Alvarez-Machain, 542 uS 692 (2004), pp. 724-725which constrained the permitted causes of action to some (indeterminate)degree. 479. economic protection to private actors has also expanded considerably,

with over 2,500 bilateral investment treaties now in force : see Crawford,Brownlie’s Principles of Public International Law, op. cit., Chap. 28 for a sum-mary. 480. See, e.g., International Covenant on Civil and Political Rights, new

York, 16 December 1966 (in force, 23 March 1976), 999 UNTS 171, arts. 14, 6, 7 ; Convention for the Protection of Human Rights and FundamentalFreedoms, Rome, 4 november 1950 (in force, 3 September 1953), 213 UNTS221, arts. 6, 2, 3 ; american Convention on Human Rights, San José, 21 november 1969 (in force, 18 July 1978), 114 UNTS 213, arts. 8, 4, 5 ;african Charter of Human and Peoples’ Rights, nairobi, 27 June 1981 (in force,21 October 1986), 1520 UNTS 217, arts. 7, 4, 5.

does not yet apply horizontally between individuals in parallel of, orsubstitution for, national law. It consists of obligations owed by theState to all people, nationals or aliens, within its territory. 258. It must also be noted that while the rights of the individual

are well developed, the rights of “peoples” as a category are lessadvanced. Certain rights relating to equality, international peace andsecurity, permanent sovereignty over natural resources, develop-ment, environment and minorities have been said to apply to personswho constitute a group or to the group itself 481, but the full embraceof third generation rights, particularly of an economic, social or cultural kind, is still in its infancy — no longer precocious after 30 years. Some progress has occurred, however, with the interna-tional recognition of indigenous rights 482, further expanding the rageof rights recognized on the international plane as well as the range ofbeneficiaries.259. There is also a marked disjunction between the considerable

expansion of individual rights under international law and the limited scope of individual responsibility. The direct responsibilityof individuals is still limited to the field of international criminallaw, and only to certain crimes within that field. with respect tothese crimes however, individual responsibility is a matter of cus-tomary international law : the customary international law origins ofindividual criminal responsibility stem from the nuremberg trialsand have been advanced through the rapid creation of internationaland hybrid criminal courts and tribunals since 1992 483.260. In any event, to have rights or obligations is not the same

thing as being a subject of international law : otherwise the termwould lack all utility. For one thing, the obligations are disconnectedfrom the rights, forming a different sub-field, for another thing, there

158 J. Crawford

481. See J. Crawford, “The Rights of Peoples : ‘Peoples’ or ‘Governments’ ?”,in International Law as an Open System : Selected Essays (London, CameronMay, 2002), Chap. 7. For discussion of substantive rights consequent to recog-nition of peoples before the law, see Chap. XI.482. See, e.g., australian Law Reform Commission, The Recognition of

Aboriginal Customary Laws (Canberra, australian Government, Report no. 31,1986). See generally J. Crawford (ed.), The Rights of Peoples (Oxford,Clarendon Press, 1988). On recognition at the international level, see discussionin Chap. X.483. agreement for the Prosecution and Punishment of the Major war

Criminals of the european axis, London, 8 august 1945 (in force, 8 august1945), 82 UNTS 279, art. 7 ; International Military Tribunal, Trial of the MajorWar Criminals before the International Military Tribunal, Nuremberg, 14 November 1945-1 October 1946 (1947), vol. 22, p. 466.

is no inherent capacity for the individual to vindicate those rights,except for specific procedures established by States, and so farunevenly available 484. This is not to deprecate or devalue theprogress made in the field of human rights since 1948, but to put itin its proper legal context.

C. Conclusion

261. The present situation of participation in the internationallegal system is untidy, and is far from Hersch Lauterpacht’s vision ofthe individual as the “ultimate subject” of international law 485. Thereis an expanding range of actors in the international system, butStates very much remain the key-holders and gatekeepers of per-sonality 486. we may ask whether the central function of the State inthe international order is still sustainable. a fashionable scholarlyview is that it is not, given that there are now other participants who are vital to international relations, such as the european union, a protean entity which is sometimes treated like a State, sometimeslike an international organization, as may be thought convenient. Butit remains the case that States retain the prerogative of domestic andinternational governance, delegated ad hoc or by treaty but not aban-doned. The international law of personality is no doubt more opentoday, but for key purposes it is still a law of exclusion, not partici-pation.


484. See Crawford, Brownlie’s Principles of Public International Law, op.cit., Chap. 29 for a summary. 485. H. Lauterpacht, International Law and Human Rights (London, Stevens

& Sons Ltd., 1950), p. 70. See also H. Lauterpacht, “Règles générales du droitde la paix”, reproduced in e. Lauterpacht (ed.), International Law, Being theCollected Papers of Hersch Lauterpacht (Cambridge, CuP, 1970), vol. 1.486. For example vCLT, op. cit., art. 53, limits determination of the status

and content of peremptory norms to States.

CHaPTeR vII

InTeRnaTIOnaL anD naTIOnaL Law : SeRvInG TwO MaSTeRS ?

“no man can serve two masters : for either hewill hate the one, and love the other ; or else he willhold to the one, and despise the other.”

Matthew 6 : 24 (King James edition).

“Laws, in their most general signification, arethe necessary relations arising from the nature ofthings. In this sense all beings have their laws : theDeity His laws, the material world its laws, theintelligences superior to man their laws, the beasttheir laws, man his laws.”

Baron de Montesquieu, The Spirit of the Laws (1750) 487.

A. The Myth of the “Dédoublement Fonctionnel”

262. Mayors (maires) occupy an important role in French society,possessing a distinctive dual function as both ex officio agents of theState and agents of the commune that they represent as Chiefexecutive — all 36,681 of them 488. They are appointed from withinthe ranks of the municipal council members (conseil municipal), byabsolute majority in a secret ballot 489. The mayors’ hold on localpower is ensured by the electoral Code, which prevents dismissalexcept by a decree of the national Council of Ministers 490.263. In exercising their powers, mayors have a double allegiance.

The mayor-as-executive wields local powers that are essentially

160

487. The Spirit of the Laws (trans. T. nugent, new York, Macmillan, 1949),p. 1. 488. Institut national de la statistique et des études économiques, Le Code

officiel géographique as at 1 January 2013. For a historic perspective on the duality see w. B. Munro, “The Office of the Mayor in France” (1907) 22Political Science Quarterly 645-662.489. See v. Hoffman-Martinot, “The French Republic, One Yet Divisible ?”,

in n. Kersting and a. vetters (eds.), Reforming Local Government in Europe :Closing the Gap between Democracy and Efficiency (Opladen, Leske-Budrich,2003), pp. 157-183. 490. electoral Code, Bk I, Title Iv, Chap. I, arts. L1, L2, L227 ; General

Code of Territory, Legislative Party, Part II, Bk I, Chaps. I-II, arts. L2122-1,L2122-4, L2122-7, L2122-16. See also art. 72 of the French Constitution du 4 octobre 1958 for principles governing prefects and local authorities.

immune from challenge by the council 491. The mayor is the solemunicipal official in charge of local education, town planning, tariffsand certain financial borrowings 492. The mayor-as-State-agent alsohas considerable delegated authority for such matters as police, lawenforcement and other matters as requested by the State 493. Here,however, mayoral discretion is not complete — refusal to exercisepowers or their negligent discharge triggers the right of the centralGovernment to resume those powers 494. In addition, the centralGovernment may initiate administrative review of mayoral decisionsconsidered contrary to law 495. These procedures provide a limitedsystem of checks and balances. 264. at the international level there are no equivalent arrange-

ments for national agencies, including courts and tribunals, whosedecisions impact upon international law or relations. If internationallaw is part of a general fabric of law, one might expect there to besystematic links with national legal systems. On the other hand ifinternational law is more like a set of primary rules generated byStates exclusively on the international plane, links between the twowould more likely be haphazard, pragmatic and unsytematized. Sowhat is the position ?

B. The Need for Interaction between International and National Laws

265. Many of the things international law tries to do have to bedone at the national level. If diplomatic or State immunities are to be


491. L. Schapp, H Daemen and a. Ringeling, “Mayors in Seven europeanCountries : Part II — Performance and analysis” (2009) 35 (2) LocalGovernment Studies 240-241. Strikingly this commentary does not reference therole of duality. 492. General Code of Territory, Legislative Party, Part Two, Bk I, Part II,

Chap. II, art. L2122-22. Per art. L2122-21 the mayor also has power over cer-tain local matters (“under the control of the council and under the adminis-trative control of the state representative in the department”). See also e. Mouritzen and J. H. Svara, Leadership at the Apex : Politicians andAdministrators in Western Local Government (Pittsburgh, univ. PittsburghPress, 2002), p. 57 ; Schapp, Daemen and Ringeling, op. cit., p. 240. For thepowers of the municipal council see General Code of Territory, LegislativeParty, Part Two, Book I, Part II, Chapter I, ss. 5.493. arts. L2122-24, L2122-27 ; General Code of the Territory, Legislative

Party, Part Two, Bk I, Part II, Chapter II, ss. 3. 494. General Code of Territory, Legislative Party, Part Two, Bk I, Part III,

Chap. I, art. L2131-5 ; General Code of Territory, Legislative Party, Part Two,Bk I, Part II, Chap. II, art. L2122-34.495. General Code of Territory, Legislative Party, Part Two, Bk I, Part III,

Chap. I.

given effect by a receiving State, that State has to recognize andimplement them as part of its national law — it is not enough to doso as a matter of administrative concession. If it is to avoid commit-ting an internationally wrongful act, the law of the receiving Statemust grant immunity from prosecution to diplomats, regardless ofthe merits of the case. If persons are to be extradited, some Stateofficial must have authority to lawfully detain them and another offi-cial must be authorized to decide when the legal conditions forextradition are met. Human rights are in the first place a matter ofnational law ; so is foreign investment. International law acts in suchcases not as a first order definer of rights and duties but as a secondorder criterion of appropriateness or lawfulness — a critical stan-dard. It operates in many respects in relation to national law ; andthat might be thought to cast into question its status as a system.These days there are international civil servants — including interna-tional judges (full or part time) whose primary allegiance may besupposed to be to the international system as such (or at least to theirbit of it). But until the twentieth century there were none or almostnone 496 : everyone’s primary allegiance was elsewhere. 266. This presents a challenge. It might seem that international

law is a thing of shreds and patches ; shreds of international stan-dards, patches upon national law — nothing whole or integrated. Butgiven that the coexistence of the various systems requires interactionon some basis, it is worth exploring what the position might be. 267. a solution was offered by Georges Scelle, who analogized

the work of the national courts and other agencies in relation tointernational law to that of the French mayor. On the one hand, such

162 J. Crawford

496. The term “international organization” is thought to have been first usedby Lorimer ; the first recognizable international organizations were the rivercommissions of the Rhine (1815), elbe (1821), Duoro (1935), Po (1849) andDanube (1856), the legal powers of which varied and included standard setting,judicial determination and immunities. See R. Kolb, “History of InternationalOrganizations or Institutions”, in R. wolfrum, Max Planck Encyclopaedia ofPublic International Law (Oxford, 2008, online), paras. 3, 14-15, 17. The firstfull-time employee of a general international organization seems to have beenMr. le Baron Melvil de Lynden, who served as the first Secretary-General of thePermanent Court of arbitration from 1 October 1900 to 1 august 1901. SirJames eric Drummond was the first international official of the League ofnations. as Secretary-General of the League Drummond enjoyed privileges andimmunities “when engaged on the business of the League”. See F. P. walters, AHistory of the League of Nations (Oxford, OuP, 1960), p. 3 ; Covenant of the League of Nations, Paris, 29 april 1919 (in force, 10 January 1920), [1920] aTS 1, 3, art. 7.

agencies perform their designated functions in the internal consti-tutional order. On the other hand, he said, they also act as agents forthe international legal system, which has always been and to a largeextent still is afflicted by an institutional deficit. Scelle describedthis as a dédoublement fonctionnel or “role splitting” :

“dans l’ordre interétatique, où il n’existe pas de gouvernants etagents spécifiquement internationaux, les agents et gouvernantsétatiques qui les remplacent sont investis d’un double rôle. Ilssont agents et gouvernants nationaux lorsqu’ils fonctionnentdans l’ordre juridique étatique ; ils sont agents et gouvernantsinternationaux lorsqu’ils agissent dans l’ordre juridique interna-tional. C’est ce que nous appellerons la loi fondamentale dudédoublement fonctionnel.” 497

It was through the action of State agents — the individuals operatingwithin the respective arms of national or local government — “thatinternational law comes to be given flesh and blood” 498.

C. Models of Interaction

268. Two standard solutions to the problem of municipallaw/international law interaction are normally put forward 499. One is


497. G. Scelle, “Règles générales du droit de la paix” (1933) 46 Recueil descours 358. Translation :

“in the interstate order, where there is no specifically international authori-ties and agents, the agents and authorities of the State that substitute forthem are invested with a dual role. They are national agents and authoritieswhen operating in the legal order of the State ; they are international agentsand authorities when acting in the international legal order. This is what wecall the fundamental law of role splitting.” (emphasis in original.)

See also Scelle’s latter definition in “Le phénomène juridique du dédoublementfonctionnel”, in w. Schätzel and H. Schlochauer (eds.), Rechtsfragen derInternationalen Organization : Festschriftfür H. Wehberg (Frankfurt am Main,vittorio Klostermann, 1956), p. 331. See further a. Cassese, “Remarks onScelle’s Theory of ‘Role Splitting’ (dédoublement fonctionnel) in InternationalLaw”” (1990) 1 EJIL 213 ; w. Schroeder and a. T. Müller, “elements ofSupranationality in the Laws of International Organizations”, in u. Fastenrath etal. (ed.), From Bilateralism to Community Interest : Essays in Honour of JudgeBruno Simma (Oxford, OuP, 2011), p. 364 ; O. Diggelmann, “Georges Scelle(1878-1961)”, in B. Fassbender and a. Peters (eds.), The Oxford Handbook ofthe History of International Law (Oxford, OuP, 2012), p. 1165 ; H. Thierry,“The Thought of Georges Scelle” (1990) 1 EJIL 203.498. Cassese, op. cit., 221.499. For discussion see J. G. Starke, “Monism and Dualism in the Theory of

International Law” (1936) 17 BYIL 74-75.

to posit that national and international law are at some level thesame ; part of a general fabric of law, which calls for integration as asort of self-recognition of coherence — this goes under the rubric ofmonism 500. The other is to stipulate that the substance of domesticlaw or national law is different organically from that of internationallaw — this is known as dualism but it might better be referred to as pluralism 501. How to reconcile the divisions of power and theinteraction of systems with the idea of effective authority has beenexamined for centuries. For example Samuel Pufendorf “termed theGerman empire a ‘mis-shapen monster’ because sovereignty wasdivided between the emperor and the estates” 502. 269. But international law is not well suited to such binary classi-

fications, despite their academic popularity. The misleading terms“monism” and “dualism” are more appropriately framed as the endsof a continuum of domestic legal structuring, with much variation in between. Classifying a State’s constitutional design as eithermonist or dualist is not so much an exercise in absolutes as a matterof degree. Countries with constitutional arrangements reflectingdegrees of monism include the civil law States of France, Germany,the netherlands, Russia and Switzerland. Those that reflect dualisttendencies include States within the common law tradition such asthe united Kingdom, united States, South africa and australia. Butno two States treat foreign relations law or international law inexactly the same way. even if a State is constitutionally disposed toone approach, State practice demonstrates the concessions made topractical demands of interaction : “intermingling between national

164 J. Crawford

500. The multivariant ideas of monism and dualism originate in the sub-discipline the philosophy of mind and have various etymologies, including ofancient Greece. Renee Descartes’ work on “mind-body dualism” (i.e. the dis-tinction between physical and non-physical substances) reinvigoratedenlightenment scholarship on the topic, which later engaged Baruch Spinoza,Christian wolff and Georg Hegel. See R. audi (ed.), The Cambridge Dictionaryof Philosophy (Cambridge, CuP, 1995), pp. 597-606 ; J. Schaffer, “Monism”, ine. n. Zalta (ed.), The Stanford Encyclopedia of Philosophy (winter 2012 ed.,online).501. See, e.g., J. Cohen “Sovereignty in the Context of Globalization : a

Constitutional Pluralist Perspective”, in S. Besson and J. Tasioulas (eds.), ThePhilosophy of International Law (Oxford, OuP, 2010), Chap. 12. 502. S. Zurbuchen, “Samuel Pufendorf and the Foundation of Modern

natural Law : an account of the State of Research and editions” (1998) 31Central European History 418. See also S. von Pufendorf, The Present State ofGermany (1696) (trans. e. Bohun and J. Seidler (eds.), Indianapolis, LibertyFund, 2007), Chap. vI, para. 9.

and international legal orders” continues to increase 503. neither theory provides a satisfactory explanation for the practice of interna-tional and national courts in articulating the content and design oflegal systems 504.

D. Relations between the Law of Nations and Domestic Legal Systems

1. Underpinnings to constitutional classification

270. Scholarship on these two standard approaches is historicallyentwined with an early twentieth-century, primarily european, poli-tical debate about the position of the State and the individual andindeed about the separate identity of international law 505. Monismwas consistent with a new-found focus on the individual as seen inthe works of Scelle, Kelsen and Lauterpacht. Dualism on the otherhand favoured “restoration of the old classical tradition of volun-tarism”, emphasizing the central role of sovereignty, political willand a categorical disjunction of national from international laws 506. 271. Hans Kelsen is perhaps the most notable proponent of

monism. at the centre of his theory is the Basic norm or Grundnormof international law. Kelsen thought of this norm as the ultimatesource of validity for all municipal laws — their authority being delegated from international law — rendering international lawapplicable domestically and superior to municipal law to the extentof any inconsistency 507. Kelsen was deeply influenced by moral andpolitical concerns such as democracy, limitations of State power andthe flourishing of individual freedoms — as was his student Hersch


503. S. Besson, “Theorizing the Sources of International Law”, S. Besson andJ. Tasioulas (eds.), The Philosophy of International Law (Oxford, OuP, 2010), p. 184. 504. J. Crawford, Brownlie’s Principles of Public International Law (Oxford,

OuP, 8th ed., 2012), p. 50. 505. J. nijman and a. nollkaemper (eds.), New Perspectives on the Divide

between National and International Law (Oxford, OuP, 2007), pp. 6-7, 10. Thedebate has a historic pedigree when examined in the context of sovereignty, asdiscussed in Chap. III.506. nijman and nollkaemper, op. cit., pp. 6-7.507. H. Kelsen, Das Problem der Souveränität und die Theorie dies Völker-

rechts-Beitrag zu einer reinen Rechtslehre (Tübingen, JC Mohr, 1920) ; H. Kelsen, Reine Rechtslehre ; Einleitung in die rechtswissenschaftliche Proble-matik (Leipzig and vienna, Deuticke, 1934) ; J. von Bernstorff, The Public Inter-national Law Theory of Hans Kelsen : Believing in Universal Law (Cambridge,CuP, 2010) ; nijman and nollkaemper, op. cit., p. 8.

Lauterpacht, an early supporter of human rights 508. Kelsen was alsoa fellow sceptic of State sovereignty, though he accepted that it wasthe “State”, not the civil servants as such, that did the work of inter-national law 509. as for Georges Scelle, his sociological theory of law“denies essential differences between international and municipallaw” 510, not only because of the links to a voluntarist vision of inter-national law, but because he viewed dualism as destroying the unityof the “law of people”, which he considered supreme law — the civitas maxima (“world community”) 511. 272. It is, unfortunately, not so easy, nor does the “world commu-

nity” work other than as metaphor. Monists ignore the reality thatdomestic or national law dictates the terms on which internationallaw “comes in” to domestic law. and this preliminary competence isactually allowed or contemplated by international law itself. Forinternational law allows States to have diverse constitutionalarrangements, including as to the relations with international law. Itmay in one sense “delegate” authority to national systems, but it is arather formalistic notion of delegation, encapsulated in the notion ofdomestic jurisdiction.273. If States are in a sense “created” by international law, most

States most of the time do not think of themselves as so created ;rather they think of themselves as created by their own efforts, as ifthey were the only State in the world — in the words of the song “ifyou were the only State in the world and I were the only govern-ment” 512. On this view, international law is subsequent to nationallaw. If international law is the law between States (as historically conceived), and in some sense is derived beyond the control ofdomestic constitutional arrangements, by contrast the locus of validity of municipal law is a matter which is the first and usuallythe last place of local constitutional ordering. 274. On this view there is a fundamental discontinuity between

the relational order of international law and the “constitutional orga-

166 J. Crawford

508. nijman and nollkaemper, op. cit., p. 397. 509. Cassese, op. cit., 221.510. O. Diggelmann, “Georges Scelle (1878-1961)”, op. cit., p. 1164. See

generally G. Scelle, “Règles générales du droit de la paix” (1933) 46 Recueil descours 358-359, 421-690. 511. Cassese, op. cit., 211-212 ; H. Thierry, “The Thought of Georges Scelle”

(1990) 1 EJIL 200.512. n. D. ayer and C. Grey, “If You were the Only Girl (in the world)”

(1916).

nization”, or “structure”, of international society ; the latter is a func-tion of the thus construed pre-legal entities, the States. 275. a similar view was put by vattel, that “[e]very sovereign

state is free to determine for itself the obligations imposed upon it” 513. The individual will of the State was thus interlaced with the collective will of States as the foundation of international law. Thisview, notably held by Hans Triepel and Dionisio anzilotti, underpinsdualistic thinking 514. Dualists conceive of international law as exter-nal to that of the State ; in Hegel’s words, “the state in and by itselfis an ethical whole” 515. In this regard, the dualism of internationallaw and municipal law applies a priori — there is no “international”sovereign to ensure compliance 516. 276. If one is forced to choose, the better view is that dualism, or

rather pluralism, provides a closer approximation to reality — “eachsystem is supreme in its own field ; neither has hegemony over theother” 517. But this is acceptable only if at the same time it is nottreated as a recipe for autarchy. This is Jean Cohen’s vision of “constitutional pluralism”, which at its core entails a commitment toco-operation as between the States in managing competing interpre-tations of legal order 518.

E. International Law in National Legal Orders

1. Preliminary points

277. The defining feature of the international legal system is theabsence of a central organ with general legislative authority, a situa-tion we describe by reference to the principle of the independenceand equality of States 519. Multilateral treaties substitute for legis-lation, as we saw in Chapter Iv. They are the nearest analogy to


513. See Starke, op. cit., p. 68.514. See, e.g., D. anzilotti, Corso di Diritto Internazionale (Padua, CeDaM,

1955), vol. I. 515. T. M. Knox (ed.), Hegel’s Philosophy of Right (Oxford, OuP, 1967),

p. 279. 516. J. austin, The Province of Jurisprudence Determined (1832) (repr. In

w. e. Rumble (ed.), Cambridge, CuP, 1995), p. 123.517. Crawford, op. cit., p. 110. 518. Cohen, op. cit., p. 274. See also n. walker, “The Idea of Constitutional

Pluralism” (2002) 65 Modern L. Re. 337-338. 519. Cf. Austro-German Customs Union case, (1931) PCIJ, Ser. A/B, No. 41,

p. 57 (Judge anzilotti).

legislation that a horizontal, decentralized system can achieve.Conclusion of treaties is attained through the agency of States, and,in particular, the officials of States who negotiate texts then haveinternational agreements implemented domestically through the legislative or judicial process. But this creates separation of powerissues which international law does not face and to which manywriters looking at the issues from an international law perspectiveare curiously insensitive. The point is that different separation ofpowers constraints apply to treaties, to issues of customary interna-tional law and to non-justiciability at the international and municipallevels and necessarily so.278. It is true that courts sometimes use language consistent with

dédoublement fonctionnel. In Eichmann the Supreme Court of Israelstated that “in the capacity of a guardian of international law and anagent for its enforcement”, Israel was entitled to prosecute crimesthat “shake the international community to its very foundations” 520.Similar assertions of guardianship were made by the uS federalcourts in Filártiga 521 and Yunis. In the latter case the District Courtof the District of Columbia held :

“not only is the united States acting on behalf of the worldcommunity to punish alleged offenders of crimes that threat-ened the very foundations of world order, but the united Stateshas its own interest in protecting its nationals.” 522

But these assertions were not only equivocal ; they were effectivelyself-judging. There was no specific international mandate in eithercase, such as the clear authority to exercise jurisdiction in varioussuppression treaties such as the Montreal Convention for theSuppression of unlawful acts against the Safety of Civil aviation,article 5 523.279. Scelle himself considered the dédoublement fonctionnel a

temporary “makeshift construct” prior to the advancement of theinternational legal system to a supra-State society 524. But the inter-

168 J. Crawford

520. 23 September 1971 (in force, 26 January 1973), (1968) ILR 300, 304. 521. Filártiga v. Peña-Irala, 630 F. 2d 876 (2d Cir. 1980).522. United States v. Yunis, 681 F. Supp. 896 (DDC 1988), p. 903 ; Cassese,

op. cit., pp. 219, 229-230. Cf. discussion on Kiobel v. Royal Dutch PetroleumCo, 133 S. Ct. 1659 (2013), in Chap. vI.523. 974 UNTS 178.524. G. Scelle, Manuel élémentaire de droit international public (Paris,

Domat-Montchrestien, 1943), pp. 22, 190 ; G. Scelle, “Quelques réflexions

national system has not developed in this manner, with the exceptionof eu supranationality. The european union operates on the basisthat national courts and officials apply the whole body of europeanlaw ex officio to cases within their jurisdiction, by virtue of anexpress legislative or even constitutional mandate to do so, and sub-ject to reference of points of european law to the Court of Justice ofthe european Community as required. There is no such general man-date at the international level, and of course no reference procedure.In the absence of an express mandate, the assertion that a given courtis acting “on behalf of the world community to punish allegedoffenders of crimes” is self-imposed.

2. State obligations to effectuate international law domestically

280. International law, or its domestic reflection, is often treatedwith suspicion by municipal courts, though perhaps with less fre-quency than used to be the case. Rosalyn Higgins identifies nationallegal culture and the personal culture of the judges as conditioningresponses to international law 525. In the International Tin Councilcase, Lord Oliver interpreted the International Organisations act1968 as creating — rather than recognizing — entities with inter-national legal personality. Higgins criticizes this decision for its“striking” and “disturbing” terminology 526. The extreme manifes-tation of insularity was Lord Donaldson’s suggestion that an interna-tional organization “becomes a person” when “touched by the magicwand of the Order in Council” ; such an organization “is not a native,but nor is it a visitor from abroad. It comes from the invisible depthsof outer space.” 527

281. as it happens, international organizations inhabit our oneworld : for example, regional development banks to which theunited Kingdom is not a party hold huge deposits in uK accounts.The uK legislation only deals with the legal personality of organiza-tions of which the united Kingdom is a member, but it neitherexpressly provides nor implies that other organizations do not exist


hétérodoxes sur la technique de l’ordre juridique interétatique”, in Hommaged’une génération de juristes au Président Basdevant (Pedone, Paris, 1960), p. 477. 525. R. Higgins, Problems and Process : International Law and How We Use

It (Oxford, OuP, 1994), pp. 206-207.526. Ibid., p. 48.527. Arab Monetary Fund v. Hashim (no. 3) [1990] 2 all eR 769 at 775.

or may not operate within the united Kingdom — they obviously doboth. Yet in Arab Monetary Fund v. Hashim (No. 3) the House ofLords was only able to find that the arab Monetary Fund existed as a legal entity by a renvoi to the law of its seat, Bahrain 528.apparently international law (like ultra-violet light) has to passthrough the filter of a national legal system to be visible.282. Similar circumspection can be seen in the united States. The

Supreme Court allowed the execution of convicted persons despiteInternational Court provisional measures under article 41 of itsStatute requiring postponement of execution to preserve the legalrights of detained subjects under the vienna Convention on ConsularRelations 529. The Supreme Court held that judgments of the Inter-national Court are not binding on domestic courts 530, but rather matters for executive consideration 531. This serves to highlight thedifficulty of the dédoublement fonctionnel in practice. The “leewaysfor judicial choice” do not permit a judge to circumvent the domesticconstitutional process to integrate international law into the muni-cipal order at will 532. It would be absurd to say that judges cannotapply international law under any circumstances, but to do so, theappropriate processes must be adopted ; and these vary with historyand constitutional arrangements.

3. Implementing treaty and custom in four systems

283. Generally speaking, the constitutional arrangements of eachindividual State will be determinative as to the process of implemen-tation and application of international law in its own domestic sphere.There is no unanimity or standardization of approach in this respect— no two constitutions are the same — and it is unhelpful to speak

170 J. Crawford

528. Op. cit. footnote 527. In Westland Helicopters v. Arab Organization forIndustrialisation (1995) 80 ILR 596, the uK courts moved towards LordOliver’s view that the proper law creating and governing international organi-zations is public international law.529. See, e.g., Avena and Other Mexican Nationals (Mexico v. United States

of America), Judgment, ICJ Reports 2004 ; LaGrand (Germany v. United Statesof America), Judgment, ICJ Reports 2001.530. Medellin v. Texas, 552 uS 491 (2008), p. 508 (which concerned the

domestic effect within the united States of Avena). See also Beard v. Greene,523 uS 371 (1998).531. The united Kingdom maintains a similar position : R v. Lyons [2003]

1 aC 976, p. 995 (Lord Hoffman). 532. J. Stone, Legal Systems and Lawyers Reasoning (London, Stevens,

1964), p. 13.

in broad generalizations. It is more instructive to reflect on the dif-ferent approaches to implementation of treaties and custom adoptedin four domestic systems — the common law systems of the unitedKingdom and the united States, which reflect dualist tendencies, and the predominantly monist civil law traditions of France andGermany.

(a) The United Kingdom

284. In the united Kingdom, signature and ratification of treatiesis the constitutional prerogative of the Crown. But as a consequence(since the Crown has no prerogative to change the law) treaties arenot self-executing or of direct effect upon ratification : they are bind-ing only on the plane of international law until the Parliament enactsimplementing legislation which incorporates or gives effect to therelevant treaty rights and obligations. Only once so incorporated may the provisions of the relevant

treaty, as reflected in the domestic legislation, then be applied bymunicipal courts 533. although the law may mirror the terms of thetreaty implemented, it is not the treaty itself but the statute thatforms part of english law 534.285. This is not to say that unincorporated treaties are of no

domestic relevance. There are a number of means by which muni-cipal courts can have regard to standards reflected in unincorporatedtreaties, including, notably, the presumption of compatibility 535. Thiscanon of construction creates a “strong presumption in favour ofinterpreting english law . . . in a way that does not place the unitedKingdom in breach of an international law obligation” 536. This pre-sumption applies to both statutes and to the common law. It is appli-cable to the former on the basis that it must be presumed thatParliament intended to legislate in conformity with international law,and not in conflict with it ; but only insofar as the words of the


533. J. H. Rayner (Mincing Lane) Ltd. v. Department of Trade and Industry[1990] 2 aC 418, pp. 476, 499-500 (Lord Oliver) ; British Airways v. LakerAirways [1985] aC 58 ; Maclain, Watson & Co. Ltd. v. Department of Trade andIndustry [1990] 2 aC 418, p. 500 (Lord Oliver).534. R v. Lyons [2003] 1 aC 976, para. 27 (Lord Hoffman).535. See further Crawford, op. cit., pp. 62-64.536. R v. Lyons [2003] 1 aC 976, p. 992 (Lord Hoffmann). See also Attorney

General v. Guardian Newspapers Ltd. (No. 2) [1990] 1 aC 109, 283 (LordGoff).

statute are capable of bearing the desired meaning, i.e. are ambigu-ous or not clear on their face 537.286. as to the relationship between common law and customary

international law, the approach traditionally taken by the unitedKingdom, that of “incorporation” 538, is best described in the wordsof Lord Denning MR in Trendtex Trading Corp. v. Central Bank ofNigeria :

“Seeing that the rules of international law have changed —and do change — and that the courts have given effect to thechanges without any act of Parliament, it follows . . . inexor-ably that the rules of international law, as existing from time to time, do form part of english law.” 539

But despite the continued recognition of the “general truth” 540 of thisproposition, a more nuanced statement of the relationship betweencustomary international law and the common law of england may bethat custom is not a part of english common law but, rather, that itis one of its sources, of which the courts may take judicial notice asappropriate541. as put by Roger O’Keefe :

“what Lord Denning MR in Trendtex referred to as the doc-trine of incorporation is in reality a statement to the effect thatthe english courts, where appropriate, may and must createrules of common law by reference to rules of customary inter-national law. It is a licence and a direction to the judge to look,

172 J. Crawford

537. R v. Secretary of State for the Home Department ; ex parte Brind [1991]1 aC 696, p. 760 (Lord ackner) ; Attorney-General v. Associated Newspapers[1994] 2 aC 238, pp. 261-262 (Lord Lowry) ; J. A. Pye (Oxford) Ltd. v. Graham[2003] 1 aC 419, p. 444 (Lord Browne-wilkinson) ; R v. Lyons [2003] 1 aC976, p. 987 (Lord Bingham).538. english courts have subscribed to an incorporationist approach since the

eighteenth century : R. O’Keefe, “The Doctrine of Incorporation Revisited”(2008) 79 BYIL 9-10. See also R. v. Keyn (1876) 2 ex. D. 63 ; West RandCentral Gold Mining Co v. R. [1905] 2 KB 391 ; Mortensen v. Peters (1906) 8 F.(J.) 93 ; Commercial and Estates Co. of Egypt v. Board of Trade [1925] 1 KB 271,p. 295.539. [1977] QB 529, p. 554 ; reiterated in R. (Campaign for Nuclear

Disarmament) v. Prime Minister of the United Kingdom, (2002) 126 ILR 738 ;and as a general principle underlying R. (European Roma Rights Centre) v.Immigration Officer at Prague Airport [2005] 2 aC 1. Cf. R. v. Secretary of StateEx parte Thakrar [1974] QB 684, p. 701.540. R. v. Jones (Margaret) [2007] 1 aC 136, p. 155 (Lord Bingham).541. J. L. Brierly, “International Law in england” (1935) 51 LQR 31 ; R. v.

Jones (Margaret) [2007] 1 aC 136, p. 155 (Lord Bingham).

where relevant, to a rule of customary international law opera-tive on the international plane and, in appropriate circum-stances, to coin in near enough its image a rule of common law applicable in an english court.” 542

(b) The United States

287. as regards the domestic practice and process of the unitedStates, article vI, paragraph 2, of the Constitution (the “SupremacyClause”) describes treaties as “the supreme Law of the Land ; and the Judges in every state shall be bound thereby ; anything in the Constitution or Laws of any State to the Contrary notwithstand-ing” 543. Treaties will accordingly prevail over state constitutions and laws. as expressed by Justice Sutherland in United States v.Belmont, “[i]n respect of all international negotiations and compacts,and in respect of our foreign relations generally, state lines dis-appear” 544.288. It is because of this express constitutional mandate that the

united States’ approach to the incorporation of treaties divergesfrom that of the united Kingdom. But a further distinction is drawn,going back to the 1829 decision of Chief Justice Marshall in Fosterv. Neilson 545, between self-executing treaties, which have the sameeffect as and may even amend acts of Congress, and non-self-executing treaties, which require implementing legislation to beeffective 546.289. The characterization as a treaty as “self-executing” or not,

and the test to be applied in making this determination is a matter ofuS, not international, law and is subject to significant uncertainty.Since the 1970s, some lower courts have adopted a nuanced “multi-factor” test for determining whether a treaty provision is self-execut-


542. R. O’Keefe, op. cit., p. 59 (emphasis added).543. Restatement (Third) of the Foreign Relations Law of the United States

(1987) (Restatement (Third)), § 111, comment (d).544. 301 uS 324 (1937), p. 331.545. 27 uS 253 (1829).546. a treaty is

“to be regarded in courts of justice as equivalent to an act of the legislature,whenever it operates of itself without the aid of any legislative provision.But when the terms of the stipulation import a contract, when either of theparties engages to perform a particular act, the treaty addresses itself to thepolitical, not the judicial department ; and the legislature must execute thecontract before it can become a rule for the Court.” Ibid., p. 314.

ing 547. However, in its 2008 decision in Medellin v. Texas 548 theSupreme Court rejected such a multi-factored approach in determin-ing whether a treaty provision was self-executing, explaining that itwas too indeterminate and would improperly “assign to the courts —not the political branches — the primary role in deciding when andhow international agreements will be enforced” 549.290. The Supreme Court also implicitly rejected the proposition

that, in cases of ambiguity, there was a strong presumption that atreaty was self-executing 550. The Court made it clear that each indi-vidual treaty should be considered on its own facts 551, but alsoseemed to confirm that the intent of the united States treaty-makersis dispositive on the question of non-self-execution 552.291. Medellin was also ambiguous as to the domestic effect of

unimplemented non-self-executing treaties 553. However, united

174 J. Crawford

547. See, e.g., Frolova v. Union of Soviet Socialist Republics, 761 F. 2d 370,p. 373 (7th Cir. 1985) ; United States v. Postal, 589 F. 2d 862, p. 877 ; People ofSaipan v. U.S. Department of Interior, 502 F. 2d 90, p. 97 (9th Cir. 1974).Relevant factors cited include :

“the purposes of the treaty and the objectives of its creators, the existenceof domestic procedures and institutions appropriate for direct implementa-tion, the availability and feasibility of alternative enforcement methods, andthe immediate and long-range consequences of self- or non-self-execution”(United States v. Postal, p. 877, quoting People of Saipan v. U.S.Department of Interior, p. 97).

548. Medellin v. Texas, 552 uS 491 (2008).549. 552 uS 491, p. 516.550. Such a presumption had previously been advocated by commentators

and the Restatement (Third) : Restatement (Third), § 111, reporters’ note 5. Seealso L. Henkin, Foreign Affairs and the United States Constitution (Oxford,Clarendon Press, 2nd ed., 1996), p. 201 ; C. M. vásquez, “Laughing at Treaties”(1999) 99 Col. LR 2154. Other commentators argue for a presumption againstself-execution : e.g. J. C. Yoo, “Treaties and Public Lawmaking : a Textual andStructural Defense of non-Self execution” (1999) 99 Col. LR 2218.551. 552 uS 491, pp. 518, 520 (2008) ; C. a. Bradley, “Intent, Presumptions,

and non-Self-executing Treaties” (2008) 102 AJIL 545-547 ; R. Crootof,“Judicious Influence : non-Self-executing Treaties and the Charming BetsyCanon” (2011) 120 Yale LJ 1787. See also Al-Bihani v. Obama, 619 F. 3d 1, pp. 15-16 (DC Cir., 2010) (Judge Kavanaugh).552. 552 uS 491, pp. 519, 521, 523 (2008) ; Restatement (Third), § 111, com-

ment (h). Cf. S. a. Reisenfeld, “The Doctrine of Self-executing Treaties and u.S. v. Postal : win at any Price ?” (1980) 74 AJIL ; C. M. vásquez, “The FourDoctrines of Self-executing Treaties” (1995) 89 AJIL ; C. M. vásquez, “TheSeparation of Powers as a Safeguard of nationalism” (2008) 83 Notre Dame LR ; D. H. Moore, “an emerging uniformity for International Law” (2006) 75George Wash. L. Rev. 1.553. The Court held that a non-self-executing treaty does not give rise to dom-

estically enforceable federal law (552 uS 491, p. 505, n 2), but failed to clarifywhether a non-self-executing treaty is simply judicially unenforceable, or whether itmore broadly lacks the status of domestic law. See also Bradley, op. cit., pp. 547-550.

States law provides for a similar presumption of compatibility to thatof the united Kingdom. as expressed by Marshall CJ in 1804 inMurray v. Schooner Charming Betsy, “an act of Congress oughtnever to be construed to violate the law of nations if any other con-struction remains” 554. as with the united Kingdom, this canon onlyapplies where the statute to be interpreted is ambiguous 555.292. The classic statement of the relationship between uS

domestic law and custom was formulated in The Paquete Habana in1900 :

“International law is part of our law, and must be ascertainedand administrated by the courts of justice of appropriate juris-diction as often as questions of right depending on it are dulypresented for their determination. For this purpose, where thereis no treaty and no controlling executive or legislative act orjudicial decision, resort must be had to the customs and usagesof civilized nations.” 556

The Restatement (Third) further clarifies this relationship : “[c]usto-mary international law is considered to be like common law in theuS, but is federal law” 557. But

“[c]ustomary law does not ordinarily confer legal rights onindividuals or companies, even rights that might be enforced bya defensive suit such as one to enjoin or to terminate a violationby the united States (or a State) of customary internationallaw” 558.

293. Claims under the alien Tort Statute (aTS) 559 have hithertoincreased the focus on custom as an element of applicable law in theunited States. The aTS, which goes back to 1789, provides for federal court jurisdiction for actions brought by an alien for a tortwhich violates international law (“the law of nations or a treaty ofthe united States”). Dozens of actions have been brought under thislegislation since its “revival” in the 1980s, following the decision in


554. 6 uS 64, p. 118 (1804). See also Restatement (Third), § 114.555. e.g. United States v. Yousef, 327 F. 3d 56, 92 (2nd Cir. 2003). Cf.

eskridge, Frickey and Garrett, Statutes and the Creation of Public Policy([united States]. Thomson west, 4th ed., 2007), p. 884.556. 175 uS 677, 700 (1900).557. Restatement (Third), § 111, comment (d).558. Restatement (Third), § 111, reporters’ note 4.559. 28 uSC, § 1350.

Filártiga v. Peña-Irala that the aTS applied to a wide array ofclaimed human rights violations 560. But the Supreme Court hastwice moved to narrow the permissible scope of claims under theaTS, perhaps to address its concern that “there is no indication thatthe aTS was passed to make the united States a uniquely hospitableforum for the enforcement of international norms” 561. In Sosa v.Alvarez-Machain, the Supreme Court limited federal courts to recog-nizing causes of action only for alleged violations of internationallaw norms that are “specific, universal, and obligatory” 562. It furthernarrowed the potential scope of alien tort claims in its 2013 decisionin Kiobel 563, unanimously holding that the presumption againstextraterritoriality applies to claims under the aTS, and that nothingin the aTS rebuts that presumption :

“nothing in the text of the aTS evinces the requisite clear indi-cation of extraterritoriality. nor does the historical backgroundagainst which the aTS was enacted overcome the presumptionagainst application to conduct in the territory of anothersovereign.” 564

(c) France

294. It is misleading to generalize too broadly about a “civil law”approach to international law. However, it is true to say that civil lawjurisdictions commonly adopt a monist approach to the incorporationof customary international law into the domestic sphere, frequentlythrough constitutional recognition.295. In France, the mode of reception of customary international

law is somewhat ambiguous 565. The Preamble to the Constitution ofthe Fourth Republic of 27 October 1946 relevantly provided that :“the French republic, true to its traditions, conforms to the rules ofinternational public law” 566. according to this declaratory provision,international law was applied per se :

176 J. Crawford

560. 630 F. 2d 876 (2d Cir. 1980).561. Kiobel, Individually and on Behalf of Her Late Husband et al. v. Royal

Dutch Petroleum Co. et al., 569 uS (2013), p. 12.562. 542 uS 692 (2004), p. 732.563. 569 uS (2013).564. Ibid., 8.565. e. Decaux, “France”, in D. Shelton, International Law and Domestic

Legal Systems (Oxford, OuP, 2011), p. 235.566. Constitution of the Fourth Republic, Preamble, § 14.

“Like other Continental courts, the French tribunals haveregarded the rules of customary international law as directlyapplicable whenever they are relevant to the adjudication of anissue of which they have jurisdiction, and concerning whichthere is no controlling legislative or executive act. They appearnever to have doubted that they, as well as other organs of theFrench state, were obligated to apply the rules of internationallaw in any appropriate case, although they have developed nocoherent doctrine of ‘adoption’ or ‘incorporation’ as the basisof this obligation.” 567

296. The applicability of custom in French domestic law is stillindirectly governed by the Preamble to the 1946 Constitution — the1958 Constitution of the Fifth Republic contains a renvoi to the 1946Constitution of the Fourth Republic 568. Cassese argues the fact thatthe new French Constitution lacks an explicit reference to customaryinternational law is “eloquent evidence of the wide-spread skepti-cism and reserve about the contents, scope and impact of the tradi-tional rules of international law” 569. However, the predominant doctrine in France seems to regard the reference of the Constitutionof 1958 to the preamble of the Constitution of 1946 as a sufficientlegal basis for applying international law 570.

(d) Germany

297. By contrast, the German approach to the incorporation ofcustomary international law is set out in clear and unambiguous


567. L. Preuss, “The Relation of International Law to Internal Law in theFrench Constitutional System” (1950) 44 AJIL 641 ff.568. Constitution of the Fifth Republic, Preamble (“The French people

solemnly proclaim their attachment to the Rights of Man and the principles of national sovereignty as defined by the Declaration of 1789, confirmed and complemented by the Preamble to the Constitution of 1946”).569. a. Cassese, “Modern Constitutions and International Law” (1985) 192

Recueil des cours 393.570. D. nguyen Quoc, “Le Conseil constitutionnel français et les régles du

droit public international” (1976) RGDIP 1001 ff., 1010 ff., 1027 ff, 1034 ; S. Sur,“L’application du droit international”, in H. Thierry et al., Droit internationalpublic (Paris, editions Montchrestien, 2nd ed., 1979), pp. 155 ff., 180, 183, 190.See also Re Self-Determination of the Comoros Islands, 30 December 1975,Rec. 41, 74 ILR 91 ; Nationalization Law, 16 January 1982, Rec. 18, 75 ILR 700 ;Nationalization Law (No. 2), 11 February 1982, Rec. 31, 75 ILR 700 ; Law onthe Evolution of New Caledonia, 8 august 1985, Rec. 63 ; Law on the Evolutionof New Caledonia (No. 2), 23 august 1985, Rec. 70 ; Treaty on European Union,9 april 1992, Rec. 55, 93 ILR 337.

terms in the Basic Law. article 25 of the Basic Law not only recog-nizes custom but also elevates it to a superior status over municipallaw :

“The general rules of international law shall be an integralpart of federal law. They shall take precedence over the lawsand directly create rights and duties for the inhabitants of thefederal territory.” 571

On this basis, German domestic courts may take judicial notice ofcustomary rules, as appropriate 572.

(e) Treaty making processes in civil law jurisdictions

298. european civil law jurisdictions generally recognize theprecedence of incorporated treaties over national law. as a result, theconstitutions of such States commonly provide for strict signatureand ratification, including prior approval by the parliament.299. In France, the power to negotiate and ratify treaties is con-

stitutionally vested in the President of the Republic 573. However, theConstitution further requires that certain treaties “may be ratified orapproved only by an act of Parliament. They shall not take effectuntil such ratification or approval has been secured.” 574 The subject-matter of the treaties subject to this requirement is wide-ranging, andexpressly includes : peace treaties ; trade agreements ; treaties oragreements relating to international organizations, committing thefinances of the State, modifying provisions which are the preserve ofstatute law, relating to the status of persons, or those which involvethe ceding, exchange or acquisition of territory. On this basis, theFrench process in respect of incorporation of most treaties mightseem to reflect an essentially dualist rather than monist approach 575.However, although the required authorization takes the form of anact of Parliament, it has no normative effect. That is, if the Presidentratifies a treaty without the required consent of the Parliament, the only sanction is political 576. The Conseil Constitutionnel has

178 J. Crawford

571. 23 May 1949, amended by the unification Treaty, Berlin, 31 august1990 (in force, 29 September 1991), 30 ILM 457.572. H. P. Folz, “Germany”, in Shelton, op. cit., p. 245.573. article 52, Constitution of the Fifth Republic.574. article 53, Constitution of the Fifth Republic.575. Decaux, “France”, in Shelton, op. cit., p. 212.576. a. aust, Modern Treaty Law and Practice (Cambridge, CuP, 2000),

p. 147.

authority to determine that a treaty contains provisions contrary tothe Constitution 577. Ratification of a treaty or agreement that is inconflict with the Constitution may only occur after the Constitutionis amended 578.300. article 55 of the Constitution provides for the supremacy of

treaties over French domestic law, in the following terms :

“Treaties or agreements duly ratified or approved shall, uponpublication, prevail over acts of Parliament, subject, withrespect to each agreement or treaty, to its application by theother party.”

This article thus expressly contemplates three separate requirements :(1) ratification or approval ; (2) publication ; and (3) reciprocity inapplication. a treaty that has not yet achieved publication in theJournal Officiel has no domestic legal effect in France, even if it hasentered into force on the international plane 579. But once published,municipal courts will apply it as of the date it entered into force forFrance 580. The Conseil Constitutionnel has somewhat narrowed thescope of application of the requirement of “reciprocity” 581. Forexample, the Conseil held that the obligations provided for under theICC Statute 582 “apply to each of the State parties independently fromconditions for their execution by other parties ; that thus the reserva-tion of reciprocity mentioned in article 55 of the Constitution is notto be applied” 583.301. In Germany, the treaty-making power of the executive is

regulated by article 59 (2) of the Basic Law, which states :

“Treaties that regulate the political relations of the Federa-tion or relate to subjects of federal legislation require the consent or participation, in the form of a federal statute, of the


577. But only if the question is referred by the President, the Prime Minister,the President of the Senate or the national assembly or 60 Senators : article 54,Constitution of the Fifth Republic.578. Ibid.579. National Federation of Guardianship Associations [2000] Rec. Lebon

781 ; Prefect of La Gironde v. Mhamedi [1992] Rec. Lebon 446, 106 ILR 204(suspension of application of treaty must also be subject to publication). See further Decaux, in Shelton, op. cit., p. 226.580. aust, op. cit., p. 147.581. Finance Act 1981, 30 December 1980, Rec. 53 ; Higher Education

Framework Act, 30 October 1981, Rec. 31. Decaux, in Shelton, op. cit., p. 227.582. 17 July 1998, 2187 UNTS 3.583. Re ICC Statute, 22 January 1999, Rec. 29, 125 ILR 475.

bodies competent in any specific case for such federal legis-lation.”

Thus, many treaties concluded by Germany require prior legislativeratification 584. The law approving ratification has the effect of mak-ing the treaty part of German law with effect from the date the treatyenters into force for Germany. The treaty may then be applied byGerman domestic courts as part of national law 585. However, unlikecustom, treaty law is not superior to municipal law.302. The task of ensuring uniformity in the application of general

international law is entrusted to the Federal Constitutional Court ofGermany 586, which has a particularly wide-ranging authority toensure domestic compliance with public international law principles.In 2004, it held :

“the Federal Constitutional Court is also competent to preventand remove, if possible, violations of public international lawthat consist in the incorrect application or non-observance by German courts of international law obligations and maygive rise to international law responsibility on the part ofGermany . . . In this, the Federal Constitutional Court is indi-rectly in the service of enforcing international law and in thisway reduces the risk of failing to comply with internationallaw. For this reason, it may be necessary, deviating from thecustomary standard, to review the application and interpretationof international law treaties by the ordinary courts.” 587

F. Conclusions : A Return to Dédoublement Fonctionnel ?

303. as we have seen, Scelle’s theory of the dédoublement fonc-tionnel fails at a general level for want of any mandate from inter-national law to domestic organs, notably courts, requiring them toact as international agencies. when such a mandate is intended it isgiven explicitly, as in the criminal law field with the suppressionconventions. Generally speaking international law demands fromnational agencies only the required result, leaving it up to those

180 J. Crawford

584. a. Paulus, “Germany”, in D. Sloss (ed.), The Role of Domestic Courts inTreaty Enforcement (Cambridge, CuP, 2009), pp. 214-18.585. Ibid., pp. 217-218. 586. Basic Law, art. 100 (2).587. 111 BverfGe 307, 328 (2004).

agencies to achieve that result by such means as the national systemwills, in accordance with its own constitutional requirements. 304. Indeed at a deeper level this is true even when such a

general mandate is or would be given, whether at the international or regional level, as with the european union. national judges arehardly more european than they are international ; if they are either it is only contingently, though the contingency differs. witheuropean as with international law, the Grundnorm (from the perspective of a national institution) is pacta sunt servanda, whereasfor a national institution the Grundnorm comes from the constitution,or perhaps the national society whose constitution it is. For theeuropean union may falter or fail or be repudiated ; just as treatiesmay be terminated. The primary rules are (it is hoped) the same andto be interpreted in the same way ; but the secondary rules are dif-ferent.305. among these secondary rules is the principle of equivalent

protection, which manifests itself in the case-law concerning theimplementation of Security Council resolutions within the europeanunion 588, which will be discussed further in Chapter XII. Perhapsthe governing principle is the Bosphorus principle, after the leadingcase. In Bosphorus an apparent conflict arose between Ireland’s obli-gations under the european Convention on Human Rights (eCHR)and under eC law 589. The applicant claimed that its right to peacefulenjoyment of property — guaranteed by article 1, Protocol One, tothe eCHR — was breached when Irish authorities impounded itsleased aircraft. Ireland argued that it could not provide the pro-cedural guarantees required under its eCHR obligations when it actedpursuant to the eC community regulation, which itself implementeda Security Council resolution. The Court determined that despiteappearances, no genuine conflict existed. Because the eC regulationin question was directly effective, Ireland was bound to implementits obligations. However, this did not abrogate european humanrights obligations ; there was a rebuttable presumption that Commu-nity law provides “equivalent” or “comparable” human rights pro-


588. C. Brölmann, “Deterritorialization of International Law”, in J. nijmanand a. nollkaemper (eds.), New Perspectives on the Divide between Nationaland International Law (Oxford, OuP, 2007), pp. 93-100. See generally P. DeSena and M. C. vitucci, “The european Courts and the Security Council :Between Dedoublement Fonctionneel and Balancing of values” (2009) 20 EJIL193.589. Bosphorus Hava Yollari v. Ireland (app. no. 45036/98), 30 June 2005.

tection to that of the eCHR ; on this basis if Irish measures compliedwith eC standards, they would also comply with eCHR standards.Such a principle was first articulated in Solange in relation to pos-sible breach by eC law of guarantees under the German FederalConstitution. The Federal Constitutional Court held that “so long as”the european Community did not have its own catalogue of fun-damental rights, German courts reserved the right to examine thecompatibility of eu law with the fundamental rights guaranteed bythe German Constitution : national constitutions prevail unless equi-valent standards can be met by a regional system, displacing anypossible conflict. The two Kadi cases, discussed in Chapter XII 590,are to similar effect. Thus even for the purposes of dédoublementfonctionnel the european union is not systematically integrated,although it is integrated pro tem. and the same principle must applyat the international level.

182 J. Crawford

590. Kadi and Al Barakaat v. Council and Commission, Joined eCJ Casenos. C-402/05 P and C-415/05 P, [2008] eCR I-6351, paras. 248-376. Reaf-firmed by Kadi v. Commission, GCeu (ex-CFI) Case T-85/09, [2010] eCR II-5177, paras. 112-121, 126-129, which held that reviewing eC law implement-ing a Security Council resolution does not amount to reviewing that resolution.appeals by the european Commission, Council and united Kingdom dismissedin Commission, Council, United Kingdom v. Yassin Abdullah Kadi, Joined eCJCases C-584/10 P, C-593/10 P and C-595/10 P, Judgment, 18 July 2013.

CHaPTeR vIII

THe IMPOSSIBILITY OF MuLTILaTeRaLISM ?

“am I my brother’s keeper ?”Genesis 4 : 9 591.

306. For a long time, it seemed that international lawyers could only count up to two : two sides to a treaty, two parties to adispute. The great peace treaties of early modern europe, despiteinvolving a multiplicity of States, tended to allocate them to one orother “side”. even today, what we think of as multilateral treatiesoften create networks of bilateral rights and obligations — subsistingbetween pairs of States — rather than obligations genuinely owed byor to multiple States. In this chapter we will consider to what extentinternational law has become truly multilateral. we will trace thedevelopment of multilateral norms in international law, particularlyof multilateral treaties, from their prehistory, to the apparent break-through of 1815, to the aggregation of quasi-universal “law-making”treaties after 1945. a recent development meriting attention is theemergence of hierarchically superior categories of multilateral norms :peremptory norms and obligations erga omnes. This will lead to adiscussion of an important practical consequence of multilateralism :which States may have standing to bring claims about matters ofcommon or global concern.

A. Bilateral, Multilateral, Plurilateral 592

307. we begin with some definitions. The vienna Convention onthe Law of Treaties (vCLT) refers to “bilateral” and “multilateral”treaties without defining either word 593. But there is no mysteryabout the meaning of the terms. “Bilateral”, which was always a

183

591. (6th-5th century BCe) trans. King James version (1604-1611).592. Parts of this chapter are adapted from a previous series of lectures given

at the Hague academy of International Law in 1997 and revised for publicationin 2006 : J. Crawford, “Multilateral Rights and Obligations in International Law”(2006) 319 Recueil des cours, particularly Chaps. I, III, Iv.593. vienna, 23 May 1969 (in force, 27 January 1980), 1155 UNTS 331

(vCLT).

legal term, entered english in the late eighteenth century and Frenchin the early nineteenth 594. It simply means two-sided : a bilateraltreaty is hence one with two sides or parties. Since the vCLT contains separate termination provisions for bilateral and multilateraltreaties that, together, are clearly meant to deal with all treaties, anytreaty with more than two sides must logically be “multilateral”,meaning many-sided.308. we might question these simple definitions. what if one

“side” of a treaty is defined to consist of multiple legal persons ? Forexample, under the Trusteeship agreement for nauru, three States —australia, new Zealand and the united Kingdom — were designatedas “the administering authority” 595. australia relied on this pro-vision in Certain Phosphate Lands in Nauru to argue that it did nothave any independent obligations under the treaty 596. But the Courtheld, in effect, that since there was no suggestion that the admi-nistering authority was a separate legal person, the use of a collec-tive designation to refer to the three States did not convert a multi-lateral treaty into a bilateral one 597. we can apply the same reasoningelsewhere. The Treaty of Guarantee in relation to Cyprus was concluded by Cyprus “of the one part” and Greece, Turkey and theunited Kingdom “of the other part” and expressly reserved the

184 J. Crawford

594. english usage : New Shorter Oxford English Dictionary (Oxford, OuP,4th ed., 1993), vol. 1, pp. 226, 1855. French usage : Le Petit Robert lists the firstuses as respectively in 1812 (“bilatéral”) and 1948 (“multilatéral”). “Pluri-latéral” (the english term is not listed in the SOED) dates from 1933, but is notdistinguished in meaning from “multilatéral”. a. Rey and J. Rey-Debouve,Dictionnaire de la langue française (Paris, 1984), pp. 184, 1243, 1463. note thatin common law parlance a “bilateral contract” is contrasted with a “unilateralcontract” ; in the latter, one party has actually performed, in exchange for apromise of performance on the other part : S. williston, A Treaties on the Law ofContracts (Mount Kisco, Baker, voorhis, 3rd ed., 1957), vol. 1, p. 26. The common law does not use the term “multilateral”, but rather “multi-party”. SeeB. a. Garner, A Dictionary of Modern Legal Usage (new York, OuP, 2nd ed.,1995), pp. 108, 576.595. Trusteeship agreement for the Territory of nauru, new York, 1

november 1947 (in force, 1 november 1947), 10 UNTS 3. as art. 4 of theagreement recorded, the three Governments had arranged that australia would,unless otherwise agreed “continue to exercise full powers of legislation, admin-istration and jurisdiction in and over the Territory”.596. Preliminary Objections of the Government of Australia, December 1990,

p. 131 (para. 321). The other element of australia’s argument was based onMonetary Gold Removed from Rome in 1943 (Italy v. France, United Kingdomand United States of America), Preliminary Question, Judgment, ICJ Reports1954, p. 32.597. Certain Phosphate Lands in Nauru (Nauru v. Australia), Preliminary

Objections, Judgment, ICJ Reports 1992, pp. 258-259 (paras. 48-50).

power of the three guarantor States to act unilaterally 598. But thetreaty still regulated the relations of those States as between eachother. Thus, whatever terms a treaty uses for its parties, if it isbetween more than two legal persons then it is a multilateral treaty.309. The term “plurilateral”, meaning several-sided, sometimes

appears. This usually means a multilateral treaty to which only alimited number of States can be parties 599. The vCLT provides for aspecial rule of unanimous acceptance of reservations “[w]hen itappears from the limited number of the negotiating States and theobject and purpose of a treaty” that that is an essential condition ofconsent 600. Seemingly, the object and purpose of a treaty can havethis effect only if there is a limited number of negotiating States 601.


598. nicosia, 16 august 1960 (in force, 16 august 1960), 382 UNTS 3. Thefirst paragraph of art. Iv provides that the three guarantors should consult in theevent of breach of the Treaty with a view to taking “measures necessary toensure observance”. The second paragraph provides that if concerted action isnot possible, “each of the three guaranteeing Powers reserves the right to takeaction with the sole aim of re-establishing the state of affairs created by the present Treaty”.599. In the wTO “plurilateral” has a slightly different meaning. The

Plurilateral Trade agreements are those, included in annex 4 of the wTOagreement, which wTO member States may but need not become parties to. See Marrakesh agreement establishing the world Trade Organization,Marrakesh, 15 april 1994 (in force, 1 January 1995), 1867 UNTS 3, art. II (3)reads :

“The agreements and associated instruments included in annex 4 (here-inafter referred to as ‘Plurilateral Trade agreements’) are also part of thisagreement for those Members that have accepted them, and are binding onthose Members. The Plurilateral Trade agreements do not create either obligations or rights for Members that have not accepted them.”

annex 4 appears at 1869 UNTS 508.600. vCLT, op. cit., art. 20 (2).601. The contrast between reservations to plurilateral treaties under art. 20 (2)

and reservations to multilateral treaties under art. 20 (4) is a subtle one. vCLT,op. cit., art. 19 (c), adopts the “object and purpose” test as a criterion for thepermissibility of a particular reservation to any multilateral treaty. If one takesthe view that art. 19 lays down an objective test for the permissibility of a reser-vation (as distinct merely from articulating certain criteria for determining theiropposability or non-opposability to other parties), then a reservation impermis-sible under art. 19 (c) will preclude the reserving State from becoming, in truth,a party to the treaty. But if an impermissible reservation is actually accepted byall States parties to a multilateral treaty, it will be effective, since the States par-ties to a treaty may collectively dispense with its requirements. The difference isthat whereas in art. 19 (c) the “object and purpose” test is concerned to test thecompatibility of a particular reservation with the treaty, in art. 20 (2) the test isapplied to the treaty as a whole, and is concerned with the question of whetherany reservations at all may be accepted, whatever their content. This appears tobe the real difference between the two categories of restricted (plurilateral) andmultilateral treaty under vCLT, op. cit., art. 20.

But beyond that special rule, plurilateralism is a matter of form or degree. Indeed, some plurilateral treaties, such as the europeanConvention on Human Rights 602, may even have more parties thanother multilateral treaties, such as the vienna Convention on Succes-sion of States with respect to Treaties 603.310. There is also an important distinction between norms and

obligations 604. Bilateral treaties (at least prima facie) create bilateralnorms ; multilateral treaties, multilateral norms. But a multilateraltreaty may generate bilateral obligations — as many bilateral obliga-tions as there are pairs of parties 605. In this way, it has been arguedthat even the wTO is a conglomerate of bilateral obligations 606. Ortake the case of the right to innocent passage under the unConvention on the Law of the Sea 607. The right of State a’s ships toinnocent passage through State B’s territorial sea may be a bilateralobligation, owed by State B to State a, even though it arises under amultilateral convention. This does not mean that other States parties, like State C, have no interest in compliance with obligationsunder a multilateral treaty ; they do have an interest in general com-pliance with unCLOS. But equally it may not make much sense tosay that State C has a “right” that State a’s ships should enjoy inno-cent passage through State’s B’s territorial sea. why should State anot be free to qualify or forego its right as against State B in a certain case or generally ? Just because State B’s obligation to Statea under a multilateral treaty is the same as its obligations to otherStates parties does not necessarily make that obligation multilateral. 311. Some of the tensions between bilateralism and multilateral-

186 J. Crawford

602. Convention for the Protection of Human Rights and FundamentalFreedoms, Rome, 4 november 1950 (in force, 3 September 1953), 213 UNTS221 (eCHR). It currently has 47 parties.603. vienna, 23 august 1978 (in force, 6 november 1996), 1946 UNTS 3. It

currently has 22 parties.604. as Roberto ago said in defining authoritatively the scope of the ILC’s

State Responsibility project : “it is one thing to define a rule and the content ofthe obligation it imposes, and another to determine whether that obligation hasbeen violated and what should be the consequence of the violation” : ILC Ybk1970/II, p. 306, cited in the introductory commentary to ILC articles on theResponsibility of States for Internationally wrongful acts, ILC Ybk 2001/II (2).605. a treaty between 4 States could generate 6 bilateral links ; a treaty

between all 193 un member States could generate 18,528 bilateral links ; atreaty between n States could generate n × (n – 1) / 2 bilateral links.606. J. Pauwelyn, “a Typology of Multilateral Treaty Obligations : are wTO

Obligations Bilateral or Collective in nature ?” (2003) 14 EJIL 925-941. 607. Montego Bay, 10 October 1982 (in force, 16 november 1994), 1833

UNTS 396 (unCLOS), art. 17.

ism in the context of multilateral treaties also exist in customaryinternational law. In both cases, specific international legal normstend to prevail over more general norms. Thus in the Right ofPassage case the International Court held that there may also bebilateral customary norms and that they will prevail over any generalrule that could be derived from custom or general principles of law 608. The notions of opposability and the persistent objector canalso be seen as restraining the emergence of general customarynorms 609. and the express specific consent of States to treaties usu-ally prevails over their general and implied consent to customarynorms (and even more so over the other category of internationallegal norms identified in article 38 (1) (c) of the Statute of theInternational Court — general principles of law). This primacy ofmore specific norms seems to undermine the possibility of genuinemultilateral norms that are not dependent on the continued accep-tance of individual States.312. Indeed for the most part what we see are contingent multi-

lateral norms : treaties that happen to be universal, customs that happen to be general, principles of such a high level of generalitythat no conscientious objection appears or is necessary. Specific and general norms relate to each other by way of interpretation, assent,practice and the operation of the lex specialis rule. One issue to consider in this chapter is the extent to which multilateral norms canexist in international law without being contingent in these respects.

B. Development of Multilateralism

1. Thinking multilaterally

313. we might see multilateral treaties — especially those agreedto at quasi-universal “law-making” conferences — as simply a moreefficient way of generating bilateral relations. To some degree, thisconclusion may be unavoidable in the absence of a vertical, centra-lized system for enacting international law. It is also historicallysound, for it acknowledges the late historical emergence of multi-lateral treaties, which were unknown in early modern internationallaw.


608. Right of Passage over Indian Territory (Portugal v. India), Merits,Judgment, ICJ Reports 1960. See Chap. X.609. See Chap. II.

314. This is well illustrated by the Peace of westphalia of 1648,which settled a major european conflict : the Thirty Years’ warbetween a mostly Protestant alliance on one side and the HolyRoman empire plus various mostly Catholic allies on the other. atthat time there were more than 150 semi-independent entities inGermany alone, under the aegis of the empire, and not all on thesame side. we should be careful in applying modern assumptionsabout statehood, legal personality and treaty-making to this very different period. nonetheless, the wide-ranging settlement at west-phalia comprised three bilateral treaties : between the netherlandsand Spain ; between Sweden and the empire ; and between Franceand the empire 610. These treaties were expressed to have two “sides”— though the participants seem to have taken a relaxed view of rep-resentation. France and Sweden insisted during negotiations that allthe German princes should sign separately, but in the event only 15were added and emperor Ferdinand III’s signature was treated asvalid for all the princes 611. equally, the Queen of Sweden repre-sented “her allies and adherents” 612. almost everyone was roped into one “side” or the other in this way rather than signing as separateentities 613, with the result that the treaties took a bilateral formdespite the multiplicity of warring States.315. This pattern of pan-european treaty-making continued to be

influential in the seventeenth and eighteenth centuries. The Peace ofutrecht in 1713-1715 was made up of no less than 11 treaties thatadhered even more closely to a bilateral form than the Peace of

188 J. Crawford

610. Peace of Münster (netherlands-Spain), 30 January 1648, 1 CTS 1 ;Treaty of Osnabruck (Sweden-Holy Roman empire), 14 (24) October 1648, 1CTS 119 ; Treaty of Münster (France-Holy Roman empire), 12 (24) October1648, 1 CTS 271. There was no treaty between Spain and France. On the Peaceof westphalia see, e.g., a. Osiander, The States System of Europe, 1640-1990.Peace-making and the Conditions of International Stability (Oxford, ClarendonPress, 1994), pp. 16-89.611. a. M. de Zayas, “westphalia, Peace of (1648)” (1984) 7 Enc. PIL 537.612. Treaty of Osnabruck, op. cit., art. I.613. On the bilateral structure of the westphalia treaties and their relationship

to earlier treaty practice, see R. Lesaffer, “The westphalia Peace Treaties and theDevelopment of the Tradition of Great european Peace Settlements prior to1648” (1997) 18 Grotiana (New Series) 71. Lesaffer notes that third powers thatwere not signatories were not necessarily bound by the treaties in the same wayas the parties, although the condition of peace established by the treaty was considered to include them : ibid., p. 83. See further R. Lesaffer, “Peace Treatiesfrom Lodi to westphalia”, in R. Lesaffer (ed.), Peace Treaties and InternationalLaw in European History : From the Late Middle Ages to World War One(Cambridge, CuP, 2004), p. 9, and other chapters in the same volume.

westphalia 614. It must be noted, as Krystyna Marek has, that thereexisted some treaties plainly phrased in multilateral terms 615. TheTreaty of aix-la-Chapelle of 1748 included what can only be seen asa multilateral guarantee : “Toutes les Puissances contractantes etinteressées au présent Traité, en garantissent réciproquement etrespectivement l’exécution.” 616 But our predominant impressionfrom the period is that the notion of multilateral treaties was slow todevelop. vattel, for instance, does not evince much interest in thequestion ; in his view, the contents of treaties were mere matters offact or history, not law 617. and this view was evidently influential.anthony Carty writes :

“the growing influence of vattel in the early nineteenth centurywas to have the very curious effect of directing lawyers’ atten-tion away from treaty issues as they were understood by nine-teenth century statesmen, without offering a radical newapproach to the issue of State necessity. That issue tended to besubmerged in vattel’s wider theoretical approach to law,[which] consisted of a theory of natural rights — or legal powers/capacities, to be more exact — which included the rightto acquire property and to conclude contracts, but did not provide any resolution of the difficulty that, at the internationallevel, States were free and equal, without any sovereign


614. France-Great Britain, 11 april 1713, 27 CTS 475 ; France-netherlands,11 april 1713, 28 CTS 37 ; France-Savoy, 11 april 1713, 28 CTS 123 ; France-Prussia, 11 april 1713, 28 CTS 141 ; France-Portugal, 28 CTS 169 ; Spain-Savoy, 13 July 1713, 28 CTS 269 ; Spain-Great Britain, 13 July 1713, 28 CTS295 ; Spain-netherlands, 26 June 1714, 29 CTS 97 ; empire and Spain-France,Rastadt, 6 March 1714, 29 CTS 1 ; empire and Spain-France, Baden, 7 September 1714, 29 CTS 141 ; Spain-Portugal, 6 February 1715, 29 CTS 201.On the Treaty of utrecht see Osiander, op. cit., pp. 90-165.615. K. Marek, “Contribution à l’étude de l’histoire du traité multilatéral”, in

Festschrift für Rudolf Bindscheidler (Bern, Stampfli, 1980), p. 23.616. 38 CTS 297, art. XXIII. Translation : “all the Powers contracting and

interested in the present Treaty guarantee reciprocally and respectively its exe-cution.” Marek’s thesis that real multilateral treaties existed well before theCongress of vienna is supported by the series of treaties constituting the SwissConfederation and dating back to 1215. This culminated in act XI of theCongress of vienna, the Declaration respecting the affairs of the SwissConfederation, vienna, 20 March 1815, 64 CTS 5. The Declaration, to whichSwitzerland acceded on 27 May 1815, recognized the neutrality and terri-torial integrity of the Swiss Confederation and regulated various inter-cantonalissues.617. e. de vattel, Le droit des gens (1758) (repr. Indianapolis, Liberty Fund,

2005), § 21.

authority to judge their actions, including the extent of their treaty obligations. There was no international civil society.” 618

316. all this is not to suggest that international law, in its “origi-nal” incarnation, was simply a network of bilateral relations. Thatwas true of treaties between some or even many States. But custom-ary international law, under the influence of natural law theoriessuch as vattel’s, was different. we might almost say (at some risk ofoversimplification) that in the classical period customary interna-tional law was assumed to be general, if not universal, not so mucha network as a field of action, a common ground — whereas treatieswere merely in the nature of particular contracts concluded againstthat background.317. There is, however, no a priori reason why treaty rights and

obligations should always be bilateral. There may be policies or out-comes in which States have a common interest without having anyspecial or distinctive interests of their own. whether we conceive of responsibility in such cases as being bilateral or multilateral, as separate or solidary, can have a range of practical consequences.Must, for example, two States that hold a right jointly act together invindicating it ? Distinguishing truly multilateral obligations frommere bundles of bilateral relations also affects how we think aboutthe international legal system. It is thus no surprise that the emer-gence and later development of multilateral treaties has coincidedhistorically with attempts to establish overarching regimes for order-ing international society, such as the Concert of europe and theunited nations Charter.318. The major shift is usually dated to 1815. although (contrary

to the conventional view) the Congress of vienna between austria,France, Prussia, Russia and Great Britain was not the first multi-lateral treaty, it includes three novel annexes framed in general terms 619. act XvII was a Règlement on the Precedence of Diplomatic agents that applied to all diplomatic agents and adopted a principle of equality among agents in each of several

190 J. Crawford

618. a. Carty, The Decay of International Law ? A Reappraisal of the Limitsof Legal Imagination in International Affairs (Manchester, Manchester uP,1986), p. 68. 619. Definitive Treaty of Peace between austria, Great Britain, Prussia and

Russia, and France, 20 november 1815, 65 CTS 251. a series of instrumentswas annexed.

classes 620. act XvI related to navigation on international rivers and included, as well as articles on specific rivers, some general articles binding “les Puissances dont les etats sont séparés ou tra-versés par une même rivière navigable” (i.e. only the signatory powers) 621.and act Xv was a Declaration relative to the universal abolition

of the Slave Trade 622, “une déclaration solennelle des principes”made “à la face de l’europe”, which was not subject to ratificationor accession. Its sentiments were splendidly progressive, though itwas little more than a record of a failure to agree : the slave tradewas not abolished until much later 623, and the main multilateralinstrument dates from 1926.319. Despite these equivocations and the fact that the former

two acts purported to bind only the contracting powers, the actsembodied certain general regulations that were multilateral in sub-stance as well as form and that went on to become part of the corpusjuris. The evident success of the Congress tended to support the view

that it had established a new european order through multilateralprocess and through the establishment of general principles, ratherthan just a bundle of bilateral bargains. Friedrich von Gentz ofPrussia wrote in 1818 :

“The whole of the european powers have since 1813 beenunited, not by an alliance properly so called, but by a system ofcohesion founded on generally recognised principles, and on


620. 64 CTS 1. The Règlement was concluded at vienna on 19 March 1815between austria, France, the united Kingdom, Portugal, Prussia, Russia, Spainand Sweden. It was modified by a Protocol of 21 november 1818, 69 CTS 385.It is now reflected in vienna Convention on Diplomatic Relations, vienna, 18 april 1961 (in force, 19 March 1967), 500 UNTS 95, arts. 14-18. See ILCYbk 1956/II, pp. 132-134.621. Règlement for the Free navigation of Rivers, vienna, 24 March 1815,

64 CTS 13, art. I. The term “états” is used in its original sense, which wascloser to a landed “estate” than to the notion of sovereignty.622. The declarants were austria, France, the united Kingdom, Portugal,

Prussia, Russia, Spain and Sweden. See 63 CTS 473.623. It was not until 1841 that a multilateral treaty was concluded on the sub-

ject : Treaty for the Suppression of the african Slave Trade (austria, unitedKingdom, Prussia and Russia), 20 December 1841, 92 CTS 437. France signedbut did not ratify the Treaty. a treaty in an “all States” form had to wait until1926 : Slavery Convention, Geneva, 25 September 1926 (in force, 9 March1927), 60 LNTS 253. Despite its importance, the General act of the BrusselsConference relating to the african Slave Trade, Brussels, 2 July 1890, 173 CTS293, is not in an “all States” form and dealt only with one region.

treaties in which every State, great or small, has found itsproper place . . .” 624

320. It was this european system that found common currencyaround the world in the twentieth century. By 1868, when a multi-lateral treaty established the first public international organization,the International Telegraphic Bureau (later the International Tele-graphic union), essentially all the techniques of multilateral law-making had been developed 625. a further advance was the Court’sReparation advisory Opinion, which established not only that inter-national organizations had legal personality, including the capacityto bring claims against States on their own behalf or on behalf oftheir agents, but also that it was opposable to non-member States 626.we will look at what some see as the culmination of this process inChapter XIv on the “constitutionalization” of international law.

2. A residual bilateralism ?

321. what we find after 1945 is almost a reversal of the approachof classical international law. Due partly to the declining influenceof natural law theories, customary international law risked dissolvinginto a network of particular relations. In contrast, the most visible andfar-reaching multilateral relations are now to be found in the greatmultilateral “law-making” treaties, whether they deal with interna-tional organizations, international peace and security, the law of thesea, world trade, human rights or the environment. This historicalprocess must in itself be accounted a major step towards multi-lateralism. But multilateral treaties may serve simply as a con-venient vehicle for the creation of new bilateral obligations. So westill need to consider to what extent the emergence of these multilateral treaties has resulted in multilateral obligations.322. This depends partly on the integrity of the text — whether,

under the law of treaties, multilateral obligations can be laid down

192 J. Crawford

624. Cited by Osiander, op. cit., pp. 248-249. See also pp. 186-187 for a catalogue of similar descriptions by, among others, Castlereagh, nesselrode and Talleyrand.625. International Telegraphic Convention, vienna, 21 July 1868, 136 CTS

292, revising a convention of 17 May 1865, 130 CTS 198.626. Reparation for Injuries Suffered in the Service of the United Nations,

Advisory Opinion, ICJ Reports 1949, p. 185. The Intrnational Court’s findings inthis case on the international legal personality of international organizations arediscussed in detail in Chap. vI (para. 246).

for all parties without fear of derogation, either from the beginningor subsequently. That would seem to be a necessary prerequisite forthe existence of multilateral obligations. The answer seems simpleenough. If States parties wish the text of a multilateral treaty to beintegral, they can say so expressly. They can exclude reservations atthe point of entry to the treaty and insist that the exclusion cannot beevaded by such devices as interpretative declarations. Or they canalso exclude the possibility of subsequent variation inter se by onlysome States parties. In short they can create more-or-less self-contained mandatory regimes. an example of a treaty regime that is widely considered to be self-contained is that of the europeanunion 627. In some treaties, such as the “single undertaking” of theworld Trade Organization 628, unCLOS and the Rome Statute forthe International Criminal Court 629, clauses have been inserted pro-hibiting reservations and insisting on the integrity of the negotiatedtext. But despite this possibility, it is still of practical importancewhat the default rule is that determines how far States are limited inmaking reservations to multilateral treaties.323. Between the world wars, the League of nations depository

practice was to decline to accept reservations not specificallyallowed by the treaty text on the basis that that was what the nego-tiating parties must have intended 630. But in the Reservations case, anarrow majority of the International Court adopted the more liberalpractice of Latin american States : to subject reservations only to theprotean limitation that they not be “inconsistent with the object andpurpose of the treaty” 631. This is as clear a policy decision as theCourt has ever made. It appears that the majority preferred treatiesof variable content but more States parties to integral treaties with(presumably) fewer States parties. It held that the object and purpose


627. See, e.g., T. C. Hartley, “International Law and the Law of the europeanunion — a Reassessment” (2001) 72 BYIL 1-35. For further discussion on self-contained regimes, see Chap. IX.628. wTO agreement, op. cit., art. XvI (5). See, however, Paulwelyn, “a

Typology of Multilateral Treaty Obligations : are wTO Obligations Bilateral orCollective in nature ?”, op. cit. 629. Rome Statute of the International Criminal Court, Rome, 17 July 1998

(in force, 1 July 2002), 2187 UNTS 155, art. 120.630. On reservations to multilateral treaties see the successive reports of the

ILC’s Special Rapporteur (alain Pellet), summarized at <http ://untreaty.un.org/ilc/summaries/1_8.htm> accessed 24 June 2013.631. Reservations to the Convention on the Prevention and Punishment of the

Crime of Genocide, Advisory Opinion, ICJ Reports 1951. This case is detailedfurther in Chap. X (paras. 436-437).

of the Genocide Convention implied “that it was the intention of theGeneral assembly and of the States which adopted it that as manyStates as possible should participate” and that the complete exclu-sion of some States “would detract from the authority of the moraland humanitarian principles which are its basis” 632. Despite this, it isclear that, in the Court’s view, at a certain point it was a question forobjective determination whether a particular reservation could bemade — whether it was consistent with the object and purpose of thetreaty — and what consequences that would have for the reservingState. The rule illustrates how multilateral concepts and regimeshave been overlaid on the bilateral foundations of traditional interna-tional law. 324. against the majority view permitting reservations, it could

be argued — as the travaux préparatoires indicated — that theGenocide Convention dealt “with the preservation of an element ofinternational order” and therefore that reservations of a generalscope, at least, had no place in it 633. Judge alvarez, in dissent,described such treaties, “reflecting the new orientation of the legalconscience of the nations”, as “almost real international laws” towhich reservations could not be made 634. Considerations such as these influenced countervailing attempts to introduce special categories of multilateral norms into international law.

C. Special Categories of Multilateral Norms

325. These attempts have challenged the traditional view thatthere can be only contingent multilateral obligations. One of them,the notion of international crimes of State, was considered but ulti-mately not adopted in the articles on the Responsibility of States forInternationally wrongful acts 635. But two other categories are now firmly established : first, obligations owed to the internationalcommunity as a whole (also called obligations erga omnes or com-

194 J. Crawford

632. ICJ Reports 1951, p. 24.633. Draft Convention on the Crime of Genocide, 26 June 1947, cited in

Reservations to the Convention on the Prevention and Punishment of the Crimeof Genocide, op. cit., p. 40 (Judges Guerrero, Sir arnold Mcnair, Read and HsuMo, joint diss. op.).634. Reservations to the Convention on the Prevention and Punishment of the

Crime of Genocide, op. cit., p. 51. 635. See J. Crawford, State Responsibility : The General Part (Cambridge,

CuP, 2013), excursus to Chap. 11.

munitarian norms) ; and secondly, peremptory (or jus cogens) norms.Some have criticized these categories for producing an ill-digestedmass of “relative normativity” 636. Certainly, they remain poorlyarticulated, and the differences and relationships between communi-tarian and peremptory norms are unclear. But if these categoriesmean anything at all, they involve the introduction of decidedly multilateral elements that are not restrained by such contingencies as requiring States to become parties to integral multilateral treaties.They appear to operate hierarchically, or “vertically”, in contrast tothe apparent flatness of traditional sources of international law,which seem to create only “horizontal” and bilateral relationshipsbetween States.326. The differentiation of these new categories occurred later

than the emergence of multilateral treaties and the recognition of theobjective personality of international organizations. unlike thoseother developments, it occurred largely outside the realm of Statepractice — through actions and reactions by the International Courtand the ILC and, to a lesser extent, scholars.327. at the time the ILC was finalizing the text of the vCLT, in

the mid-1960s, there was a great struggle in the Court over the pro-ceedings brought by ethiopia and Liberia claiming breaches ofSouth africa’s League of nations mandate for South west africa(now namibia). In 1962, the Court upheld jurisdiction 8 to 7, but in1966 it effectively reversed itself on the casting vote of the thenPresident, Sir Percy Spender 637. It held that the principle of “sacredtrust” had no residual juridical content that could give rise to legalrights and obligations “outside the system as a whole” concerning aparticular mandate :

“Once the expression to be given to an idea has beenaccepted in the form of a particular régime or system, its legalincidents are those of the régime or system. It is not permis-sible to import new ones by a process of appeal to the originat-


636. P. weil, “Towards Relative normativity in International Law ?” (1983)77 AJIL 413-442.637. South West Africa (Liberia v. South Africa ; Ethiopia v. South Africa),

Second Phase, Judgment, ICJ Reports 1966. For discussion of the controversythat followed see, e.g., S. Bastid, “L’affaire du Sud-Ouest africain devant la Cour internationale de Justice” (1967) 3 Journal du droit international ; J. Dugard (ed.), The South West Africa/Namibia Dispute (Berkeley, university ofCalifornia Press, 1973).

ing idea — a process that would, ex hypothesi, have no naturallimit. Hence, although . . . the members of the League had aninterest in seeing that the obligations entailed by the mandatessystem were respected, this was an interest which, according to the very nature of the system itself, they could exercise only through the appropriate League organs, and not indivi-dually.” 638

328. ethiopia and Liberia could thus not invoke the compromis-sory clause in the mandate. Politically, this caused an outcry ; legally,it seemed to presage a narrow view of the range and effect of inter-national law in matters of public interest, if not an actual rejection ofthe possibility of multilateral rights and obligations. The majoritywas not categorical on the point (its leading light, Judge Fitzmau-rice, was too good a lawyer for that). But the tone of the decisiontended to contradict the 1962 decision, since if the claimants couldnot invoke the compromissory clause then there was no jurisdiction.not only that, it also contradicted the approach to the “sacred trust”that the Court had taken in earlier advisory opinions on South westafrica, and it was hostile to multilateralism generally.329. The Court’s decision in Barcelona Traction a few years

later, in 1970, is often seen as a sort of apology for South WestAfrica. what had happened in between the two decisions was the adoption of the vCLT, by a majority of 79 to 1 (France) with 19 abstentions. The most controversial question at vienna had beenthat of peremptory norms 639 (the idea of which, incidentally, owednot a little to Fitzmaurice, who, as Special Rapporteur, noted the dis-tinction between “jus dispositivum” and “jus cogens” 640). under thefinal text of the vCLT, a treaty is void if it conflicts with a peremp-tory norm of general international law ; and, exceptionally, there isprovision for the settlement of disputes about peremptory norms 641.

196 J. Crawford

638. South West Africa, op. cit., p. 35 (para. 54).639. See united nations Conference on the Law of Treaties, 2nd session,

vienna, 9 april-22 May 1969, un doc. a/COnF.39/11/add.1, p. 207, and forthe explanation of the French vote, ibid., p. 203 : “jus cogens was no doubt alofty concept but it was liable to jeopardize the stability of treaty law, which wasa necessary safeguard in international relations”. Certain other delegationsagreed with the French view, but not to the point of voting against theConvention : e.g., australia, ibid., p. 205. Cf. P. Reuter, Introduction to the Lawof Treaties (London, Kegan Paul, 2nd ed., 1995), pp. 142-146, 184-185.640. G. Fitzmaurice, Third Report on the Law of Treaties, article 17, ILC

Ybk 1958/II, p. 27. 641. vCLT, op. cit., arts. 53, 64, 66 (a).

330. In Barcelona Traction, the Court avoided the term “peremp-tory norms” — to have used it a mere year after the adoption of thevCLT, when its entry into force could not be assumed, was evidentlya leap too far. Instead of that leap, we got a new concept : “obliga-tions of a State towards the international community as a whole . . .obligations erga omnes” 642. It was a confusing addition, since theexamples the Court gave of obligations erga omnes were more orless the same as the typical examples of peremptory norms : obliga-tions derived “from the outlawing of acts of aggression, and of geno-cide, as also from the principles and rules concerning the basic rightsof the human person, including protection from slavery and racialdiscrimination” 643. Most of these examples had also been listed bythe ILC in its commentary to article 53 of the vCLT 644.331. Indeed it is not too much to say that the terminology in this

area is an unresolved shambles 645. There is even confusion betweenthe idea of rights erga omnes and obligations erga omnes, as if inter-national law has not always had a concept of status relative to theinternational community as a whole (for example, in the institutionof territorial sovereignty) or as if the general notion of a legal rightis explicated by the addition of Latin words. The Court has referredto a public statement by the French President as having been made“erga omnes” 646 : for all the clarification this affords it might as wellhave said “urbi et orbi” 647. For a long time the Court itself con-tinued to avoid the term “peremptory norm” (except when describingarguments made by the parties) and preferred synonyms. In NuclearWeapons, for instance, it described the basic rules of humanitarianlaw as “intransgressible” 648. Since such rules are, regrettably, often


642. Barcelona Traction, Light and Power Company, Limited (Belgium v.Spain), Second Phase, Judgment, ICJ Reports 1970, p. 32 (paras. 33-35).643. Ibid., p. 32 (para. 34).644. ILC Ybk 1966/II, p. 248.645. See the comments by C. Tams, Enforcing Obligations erga Omnes in

International Law (Cambridge, CuP, 2005), pp. 101-102.646. Request for an Examination of the Situation in Accordance with

Paragraph 63 of the Court’s Judgment of 20 December 1974 in the nuclearTests (new Zealand v. France) Case, Request for the Indication of ProvisionalMeasures, Order, ICJ Reports 1995, p. 305 (para. 61). The Court held that thestatement, whether made erga omnes or not, could not be invoked by newZealand.647. “To the city [of Rome] and to the world” — denoting certain papal

addresses.648. Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion,

ICJ Reports 1996, p. 257 (para. 79).

transgressed, the Court presumably meant that they were peremptorywithout wanting to say so.332. The concept of peremptory norms finally made its Court

debut — 36 years after that of “erga omnes” — in DemocraticRepublic of the Congo v. Rwanda in 2008. The DRC argued thatRwanda’s express reservation to the Court’s jurisdiction under theGenocide Convention was invalid because it prevented the Courtfrom adjudicating claims of genocide. while agreeing with the resultreached by the majority, which referred to jus cogens norms, JudgeDugard commented in a separate opinion :

“This is the first occasion on which the International Courtof Justice has given its support to the notion of jus cogens. It isstrange that the Court has taken so long to reach this pointbecause it has shown no hesitation in recognizing the notion ofobligation erga omnes, which together with jus cogens affirmsthe normative hierarchy of international law. Indeed, the Courtitself initiated the notion of obligation erga omnes in 1970 . . .The approval given to jus cogens by the Court in the present

Judgment is to be welcomed. However, the Judgment stressesthat the scope of jus cogens is not unlimited and that the concept is not to be used as an instrument to overthrowaccepted doctrines of international law.” 649

333. The majority of the Court affirmed that the mere fact thatobligations erga omnes or peremptory norms are at issue “cannot initself constitute an exception to the principle that its jurisdictionalways depends on the consent of the parties” 650.334. a desirable, but unlikely, consequence of the Court’s recog-

nition of the category of “peremptory norms” would be for us toabandon the term “erga omnes” altogether. Since that outcome isnot likely, we might consider the following to be the distinction,albeit provisional, between the two concepts. The historically olderconcept is that of peremptory norms, prohibitions so fundamentalthat two or several States cannot derogate from them, not even bytreaty between themselves. The peremptory character of such normsattaches to the substance of the protected interest : the norm itself is

198 J. Crawford

649 Armed Activities on the Territory of the Congo (New Application : 2002)(Democratic Republic of the Congo v. Rwanda), Jurisdiction and Admissibility,ICJ Reports 2006, pp. 87-88 (paras. 4, 6) (Judge ad hoc Dugard, sep. op.).650. Ibid., pp. 51-52 (para. 125).

superior to all treaties (unless the treaty reflects the will of the inter-national community as a whole to modify the norm) 651. By contrastwith the substantive impact of peremptory norms, obligations ergaomnes are concerned with who may complain of a violation, i.e. withstanding.

D. Standing

1. An actio popularis?

335. we have already seen that the Court rejected a notion of“public interest standing” in South West Africa. ethiopia and Liberia,the applicants, had not suffered direct injury due to South africa’spractice of apartheid in South west africa, and the Court consideredthat they lacked standing to claim on behalf of the people of Southwest africa. In Barcelona Traction, Belgium complained of a vio-lation of the rights of a company that was Belgian-owned but incor-porated in Canada. under traditional conceptions of diplomatic protection, protecting a Canadian company is a matter held in rightof Canada, not Belgium or any other State. But the Court held thatobligations erga omnes would take the matter outside the purelyinterstate realm of diplomatic protection 652. The dominant element isthe universality of the interest in complying with such obligationsand thus, by extension, the standing of all States to claim non-com-pliance, whether or not they have suffered any actual harm from thebreach. But no question of public right or interest was actuallyinvolved in Barcelona Traction on the facts. Obligations erga omnescan be provisionally defined as multilateral rights and obligations,established in the interest of and owed to the international commu-nity as a whole, entailing a recognized legal interest of each of itsmembers to invoke compliance. But in introducing this new concept,the Court also introduced new uncertainties. after distinguishingobligations erga omnes from obligations derived from diplomaticprotection, it went on to say :

“with regard more particularly to human rights . . . it shouldbe noted that these also include protection against denial of jus-


651. The possibility of such modifications is envisaged by vCLT, op. cit.,art. 64.652. Cf. Mavrommatis Palestine Concessions, (1924) PCIJ, Ser. A, No. 2.

tice. However, on the universal level, the instruments whichembody human rights do not confer on States the capacity toprotect the victims of infringements of such rights irrespectiveof their nationality. It is therefore still on the regional level thata solution to this problem has had to be sought ; thus, within theCouncil of europe, of which Spain is not a member, the prob-lem of admissibility encountered by the claim in the presentcase has been resolved by the european Convention on HumanRights, which entitles each State which is a party to theConvention to lodge a complaint against any other contractingState for violation of the Convention, irrespective of the natio-nality of the victim.” 653

336. To start with, the contrast drawn here between instruments“on the universal level” and the european Convention is difficult toaccept. The relevant provision of the International Covenant on Civiland Political Rights (article 41) is not limited by considerations of the nationality of claims, any more than article 48 (b) of theeuropean Convention then was 654.337. More significantly, it is not clear how this is to be reconciled

with the Court’s view that obligations erga omnes are the concern ofall States “[b]y their very nature” and that hence “all States can beheld to have a legal interest in their protection” 655. The Court evengave the example of “the principles and rules concerning the basicrights of the human person”. One possibility is that in giving thatexample, it was limiting itself to those “principles and rules” thatwere clearly recognized by general international law as obligationserga omnes in 1970. This was arguably true for the prohibitions ofslavery and racial discrimination but not for the guarantee againstdenial of justice. On this view, only some aspects of human rightslaw give rise to obligations erga omnes (or at least, only some then

200 J. Crawford

653. Barcelona Traction, Light and Power Company, Limited, op. cit., p. 47(para. 91). 654. It is true that eCHR, op. cit., art. 48 (b), specifically envisaged cases

being brought to the european Court of Human Rights by a party “whosenational is alleged to be the victim”. But this was not an exclusive provision,and if the drafters of the International Covenant had wished to insert a similarproviso they could have easily done so. The issue addressed by art. 48 (b) does not now arise under the eCHR as amended by Protocol 11 (in force, 1 november 1998), ETS no. 155.655. Barcelona Traction, Light and Power Company, Limited, op. cit., p. 32

(paras. 33-35).

did so). a second possibility is that the Court envisaged the exis-tence of obligations erga omnes but without States necessarily having any of the “corresponding rights of protection” that it notedhad “entered into the body of general international law”. This mightsuggest that existing conditions of admissibility of claims (includingthe nationality of claims) would continue to apply to breaches ofcommunitarian norms. This view deprived the Court’s recognition ofthe new category of much of its significance.338. Perhaps a more reasonable interpretation is that the Court

was seeking to distinguish between the remedies available for diplo-matic protection claims and for claims brought within the frameworkof human rights. On that basis, the possibility that general interna-tional law recognized an obligation not to deny justice to personsbefore national courts (even an obligation erga omnes) would notproduce the result of assimilating the two fields. nor would it equatethe remedies available to concerned States for breaches of humanrights with those available to States whose own rights had beeninfringed by a denial of justice to their nationals. But whicheverview is preferred, uncertainties of interpretation remained, and the result is that Barcelona Traction was only the beginning of thestory 656.

2. Articles on State Responsibility 657

339. It is now generally accepted that in certain circumstancesany State has standing to protest against breaches of fundamentalnorms, and if necessary to institute proceedings to vindicate its inter-est as a member of the international community. The most signifi-cant attempt to articulate what those circumstances are is article 48of the articles on State Responsibility. It provides that “[a]ny Stateother than an injured State is entitled to invoke the responsibility ofanother State” in two cases. First, if the obligation breached is eitherowed “to a group of States including that State, and is established forthe protection of a collective interest of the group”. This categorycomprises what are sometimes called obligations erga omnes partes.Secondly, if the obligation is owed “to the international community


656. See, e.g., Tams, op. cit., pp. 48-96.657. See further Crawford, State Responsibility, op. cit., Chap. 11, from

which this section is partly drawn.

as a whole” : i.e. general obligations erga omnes. note the differencefrom article 53 of the vCLT, which describes peremptory norms as“a norm accepted and recognized by the international community ofStates as a whole”. In practice both terms are used, and the broaderone is this time more accurate : the european union, for example, isnot a State but it is a full contributor to the making of international lawand must certainly be counted part of any “international community”. 340. article 48 represents a genuine break from the older view,

expressed by the Court in the Reparation Opinion, that “only theparty to whom an international obligation is due can bring a claim inrespect of its breach” 658. That approach would work well enough ifinternational law consisted of bilateral norms — which, as we haveseen, it did almost entirely until fairly recent times — but it is non-sensical if the “party” to which the obligation is owed is a collectiveone without the capacity to act. and it would be extravagant to adoptthe method of the early modern treaties we discussed earlier andtreat obligations in areas such as international environmental andhuman rights law as being owed individually to each State. Instead,every State has standing (or in the case of obligations erga omnespartes, every State belonging to the group of States parties).341. article 48 was not adopted without criticism. Some States

expressed concern about its breadth and the danger of “opening theflood gates” of litigation 659 ; some scholars 660 complained that it wastoo weak compared with the rejected notion of international crimesof State. So far there is nothing to substantiate these contrastingfears. States do not seem inclined to bring international legal pro-ceedings without good reason ; but if they do bring them, acting (orpurporting to act) in the common interest of the international com-munity, they should not be hindered by procedural technicalities orleft with only with non-judicial means of dispute settlement.342. The question came before the Court in 2012 in Belgium v.

Senegal. Hissène Habré, the former president of Chad, had been

202 J. Crawford

658. Reparation for Injuries Suffered in the Service of the United Nations, op.cit., pp. 181-182.659. J. Crawford, Fourth Report on State Responsibility, ILC Ybk 2001/II (1),

p. 11 for the comments of Japan and France. See also J. Crawford, Third Reporton State Responsibility, ILC Ybk 2000/II (1), pp. 27-28, nn. 142-145, referringto the comments of Italy (ILC Ybk 1998/II (1), p. 104), venezuela (un doc.a/C.6/54/SR.23, § 54), austria (ILC Ybk 1998/II (1), p. 138) and the unitedStates (ILC Ybk 1998/II (1), p. 142).660. a. Cassese, International Law (Oxford, OuP, 2nd ed., 2005), pp. 269-271.

accused of torture and crimes against humanity but had been grantedasylum in Senegal. In 2005, Belgium sought his extradition inreliance on the Convention against Torture 661, and five years later itcommenced proceedings against Senegal. Senegal argued that theclaim was inadmissible because none of the torture victims hadBelgian nationality at the time of the alleged offences. Belgiumrelied, among other things, on another provision of the articles onState Responsibility : article 42 (b) (i), which entitles a State toinvoke responsibility “as an injured State” where the obligation isowed to “a group of States including that State, or the internationalcommunity as a whole” but also requires that the breach “speciallyaffects that State” 662. Belgium claimed to be specially affected byreason of the Belgian proceedings and the extradition request. TheCourt declined to decide the case on that narrower ground. Instead itheld that the relevant provisions of the Convention against Torturewere “‘obligations erga omnes partes’ in the sense that each State party has an interest in compliance with them in any givencase” 663. It concluded :

“The common interest in compliance with the relevant obliga-tions under the Convention against Torture implies the entitle-ment of each State party to the Convention to make a claimconcerning the cessation of an alleged breach by another Stateparty. If a special interest were required for that purpose, inmany cases no State would be in the position to make such aclaim. It follows that any State party to the Convention mayinvoke the responsibility of another State party with a view toascertaining the alleged failure to comply with its obligationserga omnes partes, such as those under article 6, paragraph 2,and article 7, paragraph 1, of the Convention, and to bring thatfailure to an end.” 664


661. Convention against Torture and Other Cruel, Inhuman or DegradingTreatment or Punishment, new York, 10 December 1984 (in force, 26 June1987), 1465 UNTS 85.662. emphasis added.663. Questions relating to the Obligation to Prosecute or Extradite (Belgium

v. Senegal), Merits, Judgment, International Court of Justice, 20 July 2012, p. 26(para. 68).664. Ibid., p. 26 (para. 69). See also p. 6 (para. 23) (Judge Owada, sep. op.) ;

pp. 3-4 (paras. 21-23) (Judge Skotnikov, sep. op.) ; and cf. pp. 3-4 (paras. 11-18)(Judge Xue, diss. op.).

343. although the Court did not actually refer to article 48, thedecision is firmly in line with it. Its use of the term erga omnespartes is significant, suggesting that it may be more parsimoniouswith erga omnes obligations in future 665. australia’s claim againstJapan in Whaling in the Antarctic, in which it invokes an obligationerga omnes partes under the International Convention for theRegulation of whaling, may present it with another opportunity toconsider the evolving law in this area 666.

E. Conclusion

344. The concepts of peremptory and communitarian norms rep-resent a major step towards multilateralism. But they are also justthe latest step on a path that international law has been followingsince at least 1815 and arguably even before then : away from purebilateralism and towards a system combining both bilateral and multilateral norms, including norms articulated through mass “law-making” treaties. In the primacy of the specific over the general(except in the case of jus cogens norms) and the system of reser-vations to multilateral treaties, there is still a residual bilateralism in international law. But even this is changing. In the next two chap-ters — on the proliferation of regimes and on universality — we will explore some of the tensions between bilateralism and multi-lateralism that continue to shape international law.

204 J. Crawford

665. The Court held that the dispute was exclusively one under the TortureConvention and not under general international law : Questions relating to theObligation to Prosecute or Extradite, op. cit., p. 23 (para. 55). 666. Whaling in the Antarctic (Australia v. Japan), Application Instituting

Proceedings, International Court of Justice, 31 May 2010.

CHaPTeR IX

FRaGMenTaTIOn, PROLIFeRaTIOn anD “SeLF-COnTaIneD ReGIMeS”

“Turning and turning in the widening gyreThe falcon cannot hear the falconer ;Things fall apart ; the centre cannot hold ;Mere anarchy is loosed upon the world . . .”

Yeats, The Second Coming (1919) 667.

“at one point midway on our path in life I came around and found myself now searchingthrough a dark wood, the right way blurred andlost.”

Dante, Inferno (c. 1321) 668.

A. The Problem

1. A case study : Hirsi Jamaa v. Italy

345. In Hirsi Jamaa v. Italy 669, the Grand Chamber of the euro-pean Court of Human Rights was faced with a claim by Somali anderitrean asylum seekers, part of a group of about 200 who attemptedto cross by boat from Libya to Italy. while still on the high seas,they were intercepted by the Italian authorities, transferred to Italianmilitary vessels, and returned to Tripoli. The Italian Minister of theInterior relied on a series of bilateral agreements with Libya, enteredinto pursuant to an Italian policy to curb clandestine immigration.346. The asylum seekers alleged that their treatment was incon-

sistent with various provisions of the european Convention onHuman Rights 670. In particular, the Grand Chamber found that Italy

205

667. w. B. Yeats, The Collected Poems of W. B. Yeats (ware, wordsworth,1994), p. 158. 668. D. alighieri, Inferno (trans., R. Kirkpatrick, London, Penguin, 2006), I.

1-3.

“nel mezzo del cammin di nostra vitami ritrovai per una selva oscura,ché la diritta via era smarrita.”

669. Hirsi Jamaa & Others v. Italy, [2012] eCtHR no. 27765/09.670. Originally the Convention for the Protection of Human Rights and

Fundamental Freedoms, Rome, 4 november 1950 (in force, 3 September 1953),213 UNTS 221.

had breached eCHR article 3, on the basis that by returning the asylum seekers to Libya without any form of individual assessment,Italy knowingly exposed the asylum seekers to inhuman and degrad-ing treatment 671. Most notably, the Grand Chamber found that“[i]rregular immigrants, such as the applicants, were destined tooccupy a marginal and isolated position in Libyan society, renderingthem extremely vulnerable to xenophobic and racist acts” 672. It wasfurther held that Italy had breached eCHR article 3 by exposing theasylum seekers to the risk of arbitrary repatriation 673.347. So far, so arguable. But the applicants in Hirsi Jamaa made

a further submission on the basis of article 4 of Protocol 4 674, whichprovides that “[t]he collective expulsion of aliens is prohibited”. Theplain meaning of that provision — and particularly the word “expul-sion” — would suggest that in order to be “expelled” from some-where, one must first be “in” the place that one is being expelledfrom 675. One has to be expelled from somewhere, whether theGarden of eden or uganda. It is difficult to see how someone couldbe expelled from the high seas.348. This reading is consistent with the drafting history 676.

article 4 was a latecomer to Protocol 4 ; it only emerged after theassembly’s first draft was submitted to the Committee of experts.

206 J. Crawford

671. Hirsi Jamaa & Others v. Italy, op. cit., para. 138.672. Ibid., para. 125.673. Ibid., para. 158.674. Protocol 4 to the Convention for the Protection of Human Rights and

Fundamental Freedoms, Securing Certain Rights and Freedoms other than thosealready included in the Convention and in the First Protocol thereto, 16 Sep-tember 1963 (in force, 2 May 1968), ETS no. 46.675. Cf. the definition provided in The New Shorter Oxford English

Dictionary (Oxford, OuP, 4th ed., 1993) :

“expulsion, n. The action of expelling ; the fact or condition of beingexpelled.expel, v.t.1. eject ; cause to depart or emerge ; esp. by the use of force ; banish

from a place ; discharge (a bullet).2. Compel the departure (of a person) from a society, community, etc. ;

esp. enforce the departure (of a student) from an educational establishmentas a punitive measure.3. Reject from attention or consideration.4. Keep off, exclude, keep out (rare).”

676. See Council of europe, “Collected edition of the Travaux Préparatoiresof Protocol no. 4”, 1976 <http ://www.echr.coe.int/Library/DIGDOC/Travaux/eCHRTravaux-P4-BIL2907919.pdf> ; explanatory Report to Protocol no. 4,ETS, no. 46, 1963 <http ://conventions.coe.int/Treaty/en/Reports/Html/046.htm>.

This much longer draft 677 contained a territorial limitation : expul-sion could only occur where the expellee was “in the territory of aContracting Party” 678. This version of article 4, moreover, wasbased on article 13 of the International Covenant on Civil andPolitical Rights 679, a procedural guarantee concerning expulsionwhich also contains a territorial limitation 680. Removal of the limi-tation does not mean that the drafters intended that the provision beterritorially unlimited 681. nowhere in the Committee of experts’deliberations does there appear any intention to create a territoriallyunlimited prohibition. 349. The Grand Chamber relied on article 1 of the eCHR, which

provides that “the High Contracting Parties shall secure to everyonewithin their jurisdiction the rights and freedoms defined in Section Iof this Convention”. eCHR article 1 jurisdiction is presumed to be“essentially territorial” 682 ; but it can apply in circumstances where aState exercises effective control of an area outside of its territory,e.g. in the case of a military operation 683. In Hirsi Jamaa, the GrandChamber found that Italy’s exclusive control over the interceptingmilitary vessels justified the extension of eCHR article 1 juris-


677. explanatory Report to Protocol no. 4, op. cit., para. 30 provided :

“1. an alien lawfully residing in the territory of a High Contracting Partymay be expelled only if he endangers national security or offends against‘ordre public’ or morality. 2. except where imperative considerations of national security otherwise

require, an alien who has been lawfully residing for more than two years inthe territory of a Contracting Party shall not be expelled without first beingallowed to avail himself of an effective remedy before a national authority,within the meaning of article 13 of the Convention. 3. an alien who has been lawfully residing for more than ten years in the

territory of a Contracting Party may be expelled only for reasons of nationalsecurity or if the other reasons mentioned in paragraph 1 of this article areof particularly serious nature.”

678. Ibid.679. new York, 16 December 1966 (in force, 23 March 1976), 999 UNTS

171.680. Collected edition of the Travaux Préparatoires of Protocol no. 4, op.

cit., p. 479.681. Ibid., 505-506.682. See, e.g., Banković & Others v. Belgium & 16 Other Contracting States,

[2001] eCtHR no. 52207/99, paras. 61, 67 ; Ilaşcu & Others v. Moldova &Russia, [2004] eCtHR no. 71503/01, para. 312. 683. See, e.g., Banković & Others, op. cit., para. 67 ; Ilaşcu & Others, op.

cit., para. 314. See further Loizidou v. Turkey, Preliminary Objections, (1995)103 ILR 642 ; Al-Skeini & Others v. United Kingdom, [2011] eCtHR no. 55721/07, paras. 132, 136 ; Medvedyev & Others v. France, [2010] eCtHRno. 3394/03, para. 67.

diction 684. This is all well and good — indeed, a similar conclusionwas reached with respect to maritime interception in the case ofMedvedyev v. France 685. But the point was not the meaning ofarticle 1, it was the meaning of article 4. The Chamber went on toargue that the concept of extraterritorial jurisdiction in eCHR,article 1, should be used to expand the scope of article 4 ofProtocol 4 on the basis that :

“[t]o conclude otherwise . . . would result in a discrepancybetween the scope of application of the Convention as such andthat of article 4 of Protocol no 4, which would go against the principle that the Convention must be interpreted as awhole” 686.

This reasoning is specious : one struggles to see how the meaning ofthe term “jurisdiction” in eCHR article 1 can control the meaning ofthe term “expulsion” in article 4 of Protocol 4. The Grand Chamberwas no doubt correct to say that article 4 applies in situations fallingwithin the jurisdiction in the sense of article 1. But that does notmean that the substantive elements of article 4 do not need to bemade out.350. The conclusion is at odds with other important norms of

international law, notwithstanding the Grand Chamber’s unconvin-cing assertion that its ruling did not compromise a State’s ability toset its own immigration policy 687. Of course it did. In principle, aState has the right to determine who shall enter its territory, subjectto a few legal restrictions 688. among these, collective expulsion of aliens is a serious breach of international law, and article 4 isexpressed as an absolute and non-derogable prohibition. as such, itmust be interpreted narrowly and precisely. If any measure prevent-

208 J. Crawford

684. Hirsi Jamaa & Others, op. cit., paras. 76-82.685. Medvedyev & Others, op. cit., paras. 66-67.686. Hirsi Jamaa & Others, op. cit., paras. 178-180.687. Ibid., para.179 :

“The above considerations do not call into question the right of States toestablish their own immigration policies. It must be pointed out, however,that problems with managing migratory flows cannot justify havingrecourse to practices which are not compatible with the State’s obligations.”

688. This was perhaps amongst the earliest and widely recognized powers ofsovereignty, and was recognized in england as a prerogative power of theCrown as early as the seventeenth century : East India Co v. Sandys, (1684) 10 ST 371, 530-531 (Jeffreys CJ). See further R (European Roma Rights Centre& Others) v. Immigration Officer at Prague Airport, [2005] 2 aC 1, 27-28 (LordBingham).

ing groups of aliens from entering the territory of a Contracting Stateis prohibited, then the words of article 4 cease to have meaning. Forexample, in the “Eurotunnel” decision, at the request of cross-Channel train operators, an arbitral tribunal found that it was incum-bent on the united Kingdom and France “to maintain conditions ofof normal security and public order” to prevent asylum seekers from illegally boarding trains travelling from France to the unitedKingdom 689. under the Grand Chamber’s decision, it would appearthat this would be equivalent to collective expulsion. 351. The decision, moreover, ignores the purpose served by the

principle of non-refoulement that is expressed in the RefugeeConvention 690. article 33 (1) of the Convention prohibits the expul-sion or return (refoulement) of refugees “to the frontiers of territorieswhere his life or freedom would be threatened” on account of racial,ethnic or religious factors. The generally accepted position is that thenon-admittance of a refugee must occur from within or at the veryleast at the frontier of the State refusing entry 691. For the scope ofarticle 4 to expand so as to cover as preventive measures taken out-side the territory of the State would render article 33 (1) of theRefugee Convention redundant as far as groups of persons are con-cerned and would do so irrespective of the treatment to be expectedin the State of return : there would no longer be any requirement ofa well-founded fear of persecution. 352. Hirsi Jamaa is also at odds with the International Law

Commission’s (ILC) work on expulsion. Draft article 2 (a) of theCommission’s proposed Draft articles on the expulsion of aliensprovides that “expulsion” means “a formal act, or conduct consistingof an action or omission, attributable to a State, by which an alien is compelled to leave the territory of that State” 692. as the Special


689. The Channel Tunnel Group Limited and France-Manche S.A. v. UnitedKingdom and France, Partial Award, (2007) IIC 58, para. 395 (2). 690. Convention relating to the Status of Refugees, Geneva, 28 July 1951 (in

force, 22 april 1954), 189 UNTS 150. 691. See European Roma Rights Centre & Others, op. cit., 29-30 (Lord

Bingham) ; Sale, Acting Commissioner, Immigration and Naturalisation Servicev. Haitian Centers Council Inc. (1993) 509 uS 155 ; Minister for Immigrationand Multicultural Affairs v. Ibrahim, (2000) 204 CLR 1 ; Minister for Immi-gration and Multicultural Affairs v. Khawar (2002) 210 CLR 1. See generally T. Gammeltoft-Hansen, Access to Asylum : International Refugee Law and theGlobalisation of Migration Control (Cambridge, CuP, 2011), Chap. 3.692. Draft arts. 1-32 were provisionally adopted on first reading by the ILC’s

Drafting Committee at its 64th session in 2012. See un doc. a/Cn.4/L.797, 24 May 2012 (emphasis added).

Rapporteur noted, expulsion “refers to the displacement of an individual under constraint beyond the territorial frontier of theexpelling State to a State of destination” 693. 353. Finally, the decision is at odds with the decision of other

courts or tribunals at a domestic level. The House of Lords in theRoma Rights case had to consider whether the united Kingdom’spolicy of intensively questioning Roma passengers prior to boardingat Prague airport (pursuant to an agreement with the CzechGovernment) was contrary to the Refugee Convention. Following anextensive analysis of the Convention, the eCHR and customaryinternational law, the House concluded that there was no obligationon the part of the united Kingdom to extend legal protections of therelevant kind to asylum seekers in situations that did not arise at orwithin its frontiers 694.

2. Defining fragmentation

354. In short the Grand Chamber in Hirsi Jamaa laid down the“law” on collective explusion without due regard to the relationshipof article 4 of eCHR Protocol 4, with other rules of internationallaw. It employed textual and teleological reasoning to reach a pre-determined result. Other international courts and tribunals, notablythe International Court, would be most unlikely to reach the sameconclusion. But the issue may never come before the InternationalCourt. under the eCHR, there is no right of appeal to another courtor tribunal. There is no avenue for the correction of the decision,especially where it is a unanimous decision of a Grand Chamber. aswas observed by the appeals Chamber of the International CriminalTribunal for the former Yugoslavia in Tadić :

210 J. Crawford

693. un doc. a/Cn.4/573, 20 July 2006, p. 57 (para. 178).694. European Roma Rights Centre & Others, op. cit., pp. 45-47 (Lord

Steyn), pp. 53-55 (Lord Hope), p. 55 (Lady Hale), p. 66 (Lord Carswell). Seeespecially ibid., p. 38 (Lord Bingham) :

“There would appear to be general acceptance of the principle that a person who leaves the state of his nationality and applies to the authoritiesof another state for asylum, whether at the frontier of the second state orfrom within it, should not be rejected or returned to the first state withoutappropriate inquiry into the persecution of which he claims to have a well-founded fear. But that principle, even if one of customary international law,cannot avail the appellants, who have not left the Czech Republic nor presented themselves, save in a highly metaphorical sense, at the frontier ofthe united Kingdom.”

“International law, because it lacks a centralized structure,does not provide for an integrated judicial system operating anorderly division of labour among a number of tribunals, wherecertain aspects or components of jurisdiction as power could becentralized or vested in one of them but not the others. In inter-national law, every tribunal is a self-contained system (unlessotherwise provided).” 695

355. In this light, the decision in Hirsi Jamaa typifies the problemcaused by the proliferation of international courts and tribunals, andthe consequent “fragmentation” of international law. In short, frag-mentation is the product of a system of laws that, by and large, lacksa sense of vertical integration, of hierarchy. 356. The issue was addressed by the ILC Study Group under the

chairmanship of Martti Koskenniemi 696. This recognized that :

“[S]pecialized law-making and institution-building tends totake place with relative ignorance of legislative and institu-tional activities in the adjoining field and of the general prin-ciples and practices of international law. The result is conflictsbetween rules or rule-systems, deviating institutional practicesand, possibly the loss of an overall perspective on the law.” 697

Put more pithily by Sir Robert Jennings, fragmentation reflects aconcern that international law is growing “without any overall plan”“there is a danger that international law as a whole will become frag-mented and unmanageable” 698.

3. Species of fragmentation

357. So far I have proceeded without defining fragmentation. Infact two broad concepts may be identified : “fragmentation” and“proliferation”. Fragmentation will be used here to refer to substan-


695. Prosecutor v. Dusko Tadić (Jurisdiction), (1995) 105 ILR 458.696. The topic was originally included on the ILC’s programme under the

title “Risks ensuing from the Fragmentation of International Law” : ILC Ybk2000/II (2), p. 131. This was subsequently and finally changed in 2002 to “Frag-mentation of International Law : Difficulties arising from the Diversificationand expansion of International Law”. See ILC Ybk 2002/II (2), p. 98.697. Report of the ILC Study Group, un doc. a/Cn.4/L.682, 13 april 2006,

p. 11 (para. 8).698. R. Y. Jennings, “The Role of the International Court of Justice” (1997)

68 BYIL 58-60.

tive fragmentation, i.e. the product of conflicting but equally autho-ritative pronouncements on international law by courts and tribunals.Proliferation refers to institutional fragmentation, i.e. the greatexpansion in the number of international courts and tribunals ofcompeting jurisdiction since the 1980s 699.358. To the problems of fragmentation and proliferation may be

added a further concern : the so-called self-contained regime 700. aself-contained regime is said to occur where a particular substantiveor institutional field of international law develops to such an extentthat it effectively displaces the general law. In such a situation, theregime so created no longer interacts with international law, eventhough it may owe to international law its existence and externallegitimacy.

B. Substantive Aspects of Fragmentation

1. Fragmentation as a crowded bar

359. according to article 38 (1) (d) of the Statute of the Inter-national Court, judicial decisions are a subsidiary source for theidentification of international law : nowhere is it said that the judicialpronouncements of one court carry more weight than any other. Thislack of hierarchy can result in a situation in which many differentvoices shout from many different rooftops, with none able to beheard over the other.360. a key example of this process arose from the International

Criminal Tribunal for the former Yugoslavia, which has engaged inseveral differences of opinion with the International Court 701. whilstthese have usually taken the form of the two institutions talking pasteach other, on at least one occasion the situation has become a full-fledged dialogue, with contrary views forcefully presented.

212 J. Crawford

699. See, e.g., Symposium, “The Proliferation of International Courts andTribunals : Piecing Together the Puzzle” (1999) 31 NYU JILP 679 ; Y. Shany,The Competing Jurisdictions of International Courts and Tribunals (Oxford,OuP, 2003), pp. 1-11 ; C. Brown, A Common Law of International Adjudication(Oxford, OuP, 2007), Chap. 1.700. See B. Simma, “Self-contained Regimes” (1985) 16 NYIL 111-136 ;

B. Simma and D. Pulkowski, “Of Planets and the universe : Self-ContainedRegimes in International Law” (2006) 17 EJIL 483-529 ; B. Simma and D. Pulkowski, “Leges Speciales and Self-Contained Regimes”, in J. Crawford,a. Pellet and S. Olleson (eds.), The Law of International Responsibility (Oxford,OuP, 2010), pp. 139-163.701. Koskenniemi and Leino, “Fragmentation of International Law ? Post-

modern anxieties” (2002) 15 LJIL 562-567.

361. One point of divergence occurred in the context of belliger-ent reprisals. In the Martić case, Trial Chamber I held that armedreprisals against civilians were in all cases forbidden 702, a positionwhich was contradicted by the International Court two months laterin the Nuclear Weapons advisory Opinion, in which it held that suchreprisals could indeed be lawful if necessary and proportional 703. 362. a more profound divergence occurred in the context of

attribution to the State of the conduct of non-State actors 704. InNicaragua, the International Court held that attribution required it tobe shown that the State in question possessed effective control of therelevant non-State actor in relation to the alleged conduct 705.applying that test to the uS support for the nicaraguan contras, itwas held that the united States could not be held generally respon-sible for all acts of these paramilitary groups, notwithstanding that a considerable measure of logistical and other support had been pro-vided. This finding was called into question by the ICTY in Tadić.as an international criminal tribunal with jurisdiction limited to indi-viduals, the ICTY would not need to address questions of Stateresponsibility. In Tadić, however, the ICTY treated State responsi-bility as an antecedent question in order to classify the conflict inBosnia as an international rather than a non-international armed conflict : many of the charges against Tadić depended on that classi-fication. The Trial Chamber — Judge McDonald dissenting — reliedon the test proposed in Nicaragua, and held that the actions of theRepublika Srpska could not be sheeted home to the Federal Republicof Yugoslavia, notwithstanding that Republika Srpska was largelydependent upon the FRY for its existence 706.363. On appeal by the prosecution, the appeals Chamber

affirmed the Trial Chamber’s decision, but criticized its reliance onNicaragua, arguing that the International Court’s use of the effectivecontrol test was contrary to the logic of State responsibility. Rather,the appeals Chamber argued for a lesser standard of overall control


702. Prosecutor v. Marcić (Rule 61), (1996) 108 ILR 47.703. Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion,

ICJ Reports 1996, p. 246 (para. 46).704. See generally J. Crawford, State Responsibility : The General Part

(Cambridge, CuP, 2013), pp. 146-156.705. Military and Paramilitary Activities in and against Nicaragua (Nica-

ragua v. United States), Merits, Judgment, ICJ Reports 1986, p. 65 (para. 115).706. Prosecutor v. Tadić, ICTY, IT-94-1-T (Trial Chamber, 7 May 1997),

para. 216.

in certain circumstances 707. This test was unsurprisingly seen asauthoritative by the Trial Chambers of the ICTY and was appliedsystematically in later cases 708.364. The International Court returned to the question of responsi-

bility for non-State actors in Bosnian Genocide 709. This arose fromthe same factual matrix as Tadić : the question was whether the FRY(and, later, Serbia) was responsible for acts of genocide committedby Bosnian Serb militias during the Bosnian war. The Court heremade a concerted effort to engage with the jurisprudence of theICTY, but it affirmed the effective control test in Nicaragua, rejectedthe reduced threshold advocated by the appeals Chamber in Tadić,stressing that the question “was not indispensable in the exercise of[the ICTY’s] jurisdiction”, and that the ICTY’s expertise did notextend to

“issues of general international law which do not lie within thespecific purview of its jurisdiction and, moreover, the resolu-tion of which is not always necessary for deciding the criminalcases before it” 710.

214 J. Crawford

707. Prosecutor v. Tadić (Appeal against Conviction), (1999) 124 ILR 62,109.708. See, e.g., Prosecutor v. Aleksovski, ICTY, IT-95-14/I-T (Trial Chamber,

25 June 1999) ; Prosecutor v. Aleksovski, ICTY, IT-95-14/I-a (appealsChamber, 24 March 2000) ; Prosecutor v. Blaškić, ICTY, IT-95-14/T (TrialChamber, 3 March 2000) ; Prosecutor v. Naletilić, ICTY, IT-98-34-T (TrialChamber, 31 March 2003) ; Prosecutor v. Kordić and Čerkez, ICTY, IT-95-14/2-T (Trial Chamber, 26 February 2001) ; Prosecutor v. Kordić and Čerkez, ICTY,IT-95-14/2-a (appeals Chamber, 17 December 2004) ; Prosecutor v. Delalić etal., ICTY, IT-96-21-a (appeals Chamber, 20 February 2001). See further e. LaHaye, War Crimes in Internal Armed Conflicts (Cambridge, CuP, 2008), pp. 15-19. See also a. J. J. De Hoogh, “articles 4 and 8 of the ILC articles on State Responsibility, the Tadić Case and attribution of acts of Bosnian Serbauthorities to the Federal Republic of Yugoslavia” (2001) 72 BYIL 279-281 ; M. Milanović, “State Responsibility for Genocide” (2006) 17 EJIL 581. Cf. also the dissenting judgment of Judge Shahabuddeen : (1999) 124 ILR 62, 194 ff.709. The Court had flagged its intention to hold fast to the effective control

test in Nicaragua in the Armed Activities case. The Court there was called uponto decide whether the conduct of certain auxiliaries operating in the DemocraticRepublic of the Congo could be attributed to uganda. It held that the requiredlevel of instruction, direction or control within the meaning of the provision hadnot been made out on the facts, citing nicaragua in support : Armed Activities onthe Territory of the Congo (Democratic Republic of the Congo v. Uganda),Judgment, ICJ Reports 2005, p. 226 (para. 160).710. Application of the Convention on the Prevention and Punishment of the

Crime of Genocide (Bosnia and Herzegovina v. Serbia and Montenegro),Judgment, ICJ Reports 2007, p. 209 (para. 403).

2. Substantive fragmentation and its resolution

365. although the level of disagreement between the Interna-tional Court and ICTY evidenced in these exchanges are prima facie perturbing, it may fairly be asked whether fragmentation isreally a problem 711.366. The notion that different elements of a judicial system can

have a different idea of what the law is will come as no surprise toany lawyer who has been trained in a federal jurisdiction in whichseparate regions possess separate judicial hierarchies. To take oneexample concerning the uS alien Tort Statute 712, a significant splitdeveloped between the different circuit courts of appeal 713 as towhether foreign corporations could be held liable for violations “ofthe law of nations”. It is true that the uS federal system contains aresolving element, namely a court of final appeal in the form of the uS Supreme Court. But as we have seen in Kiobel 714, the mere fact of a hierarchy of courts within a system does not mean that arguments will inevitably be resolved : a court may elide an issue, ora split may develop within the Bench that masks the long-term significance of the decision 715. In such a situation, litigants may beleft with contradictory pronouncements of formally equal weight.But this does not signify the collapse of the system. 367. The second point to be made is that the existence of multiple

formally equal statements does not prevent international law fromexercising a preference, although it may take time for the preferenceto emerge. article 38 (1) (d) of the Statute of the International Court


711. On which see M. Koskenniemi and P. Leino, op. cit., pp. 553-579 ; e. Benvenisti and G. Downs, “The empire’s new Clothes : Political economyand the Fragmentation of International Law” (2007) 60 Stanford L. Rev. 602-608. 712. 28 uSC, § 1345. Following the “rediscovery” of the act in the case of

Filartiga v. Peña-Irala, 630 F. 2d 876 (2nd Cir. 1980), the Torture victimsProtection act of 1991 was passed, providing a cause of action for any victim oftorture or extrajudicial killing wherever committed : 106 Stat. 73 (1992).713. See, e.g., Presbyterian Church of Sudan v. Talisman Energy Inc, 582 F.

3d (2nd Cir. 2009) ; Kiobel, Individually and on Behalf of Her Late Husband etal. v. Royal Dutch Petroleum et al., 621 F. 3d 111 (9th Cir. 2010). See further O. Murray, D. Kinley and C. Pitts, “exaggerated Rumours of the Death of analien Tort ? Corporations, Human Rights and the Remarkable Case of Kiobel”(2011) 12 Melbourne JIL 1.714. Kiobel, Individually and on Behalf of Her Late Husband et al. v. Royal

Dutch Petroleum Co. et al., 569 uS (2013).715. For a further example, see R v. Keyn (1876) 2 ex. D. 63. See further

G. Marston, The Marginal Seabed : United Kingdom Legal Practice (Oxford,Clarendon, 1981), Chap. 5.

refers to highly qualified publicists as a subsidiary means for deter-mining the law. Scholarly criticism may thus play a role, sometimesan important one. The pronouncements of the ILC, charged underthe terms of its Statute with the “the promotion of the progressivedevelopment of international law and its codification” 716, carryunusual normative force and may help to resolve difficult legalissues. In the case of State responsibility for non-State actors, thearticles on the Responsibility of States for Internationally wrongfulacts 717 expressly approved the International Court’s formulation of effective control in its commentary to article 8 718, confining Tadić to its facts. It thereby avoided contradicting the appealsChamber while expressing a preference for the test of attribution inNicaragua 719. article 8, in turn, was applied by the Court in BosnianGenocide, confirming the approach in Nicaragua 720. The Court’sformulation must be seen as the default rule 721. 368. This episode highlights two further points. First, it suggests

that despite the lack of a formal hierarchy between internationalcourts and tribunals, the pronouncements of the International Court,the only permanent tribunal of general jurisdiction, carry particularweight 722. This may provide a centre of gravity. Second, it demon-strates that international law is not inherently hostile to the notion ofa rule confined to a particular factual or institutional context, as inTadić 723. 369. This suggests — Hirsi Jamaa notwithstanding — that frag-

mentation poses no real threat to international law as a system. Theinstances of disagreement between courts and tribunals rarely matureinto outright conflict, and even if they do, the system allows forlegal preferences to be expressed over time, such that a choicebetween two differing norms may eventually be made. Perhaps even

216 J. Crawford

716. Statute of the International Law Commission, annexed to Ga res. 174(II), 21 november 1947, as amended by : Ga res. 485 (v), 12 December 1950 ;Ga res. 984 (X), 3 December 1955 ; Ga res. 985 (X), 3 December 1955 ; Gares. 36/39 of 18 november 1981, art. 1 (1).717. ILC Ybk 2001/II (2), p. 20.718. Ibid., p. 47 (para. 4).719. Crawford, State Responsibility : The General Part, op. cit., p. 154.720. Application of the Convention on the Prevention and Punishment of the

Crime of Genocide, op. cit., p. 208 (para. 399).721. Crawford, State Responsibility : The General Part, op. cit., p. 156.722. See generally M. andenas and e. Bjorge (eds.), A Farewell to Frag-

mentation : Reassertion and Convergence in International Law (Cambridge,CuP, 2014).723. Prosecutor v. Tadić (Jurisdiction), op. cit., 434 ff.

Hirsi Jamaa may be only the beginning of the debate over collectiveexpulsion, not the last word on it.

C. Institutional Aspects of Fragmentation

1. The overlapping jurisdictions of international courts and tribunals

370. The post-Cold war era has seen the establishment of a number of new international courts or tribunals. By and large, theseare specialized in character and limited in terms of their jurisdictionratione materiae. But they still all function — or at least purport tofunction — within the same ocean of international law.371. Two broad reasons for this proliferation may be identi-

fied 724. In the first place, the traditional reluctance of States to submit bilateral disputes to third party adjudication has decreased. In the second, globalization and increasing interdependence hasresulted in complex problems of interaction, and the correspondingproduction of detailed norms of international law. One need onlycompare the provisions of the un Convention of 1982 with the 1958Geneva Conventions on various aspects of the law of the sea 725. Thesame may be said of a comparison of the provisions of the wTOcovered agreements 726 with the 1947 General agreement on Tariffsand Trade 727. The increasing complexity of such norms has resultedin a need for specialized systems of dispute resolution 728.


724. Brown, op. cit., pp. 22-23. Shany, op. cit., pp. 3-4, offers a more detailedtaxonomy, including : (1) the increased density, volume and complexity of inter-national norms, which require sufficiently complex forms of dispute resolution ;(2) greater commitment to the rule of law in international relations ; (3) the easing of international tensions post-Cold war ; (4) the positive experience ofStates with earlier specialist international courts, such as the european Court ofHuman Rights or the Court of Justice of the european Communities ; and (5) theperceived unsuitability of the International Court of Justice to adequatelyaddress specialized and technical disputes.725. See Convention on the Territorial Sea and the Contiguous Zone, Geneva,

29 april 1958 (in force, 10 September 1964), 516 UNTS 206 ; Convention on the High Seas, Geneva, 29 april 1958 (in force, 30 September 1962), 450 UNTS82 ; Convention on Fishing and Conservation of the Living Resources of theHigh Seas, Geneva, 29 april 1958 (in force, 20 Geneva March 1966), 559 UNTS286 ; Convention on the Continental Shelf, Geneva, 29 april 1958 (in force, 10 June 1964), 499 UNTS 312.726. Marrakesh agreement establishing the world Trade Organization,

Marrakesh, 15 april 1994 (in force, 1 april 1995), 1867 UNTS 14. See furtherworld Trade Organization, The Legal Texts : The Results of the Uruguay Roundof Multilateral Trade Negotiations (Cambridge, CuP, 1994).727. Geneva, 30 October 1947 (in force, 1 January 1948), 55 UNTS 194.728. Shany, op. cit., pp. 3-4.

372. But with such systems, new problems may arise. This maybe the case when the obligations incumbent on States in one arenaprevent them from relying on other international law rights justi-ciable before another tribunal. One example is the Mexico—SoftDrinks case before the wTO 729. In that case, Mexico had imposed aseries of taxes on soft drinks and other beverages that used a sweet-ener other than cane sugar. unsurprisingly, the united States —whose companies both produced in and imported to Mexico softdrinks sweetened with corn syrup — protested, claiming that themeasures were inconsistent with GaTT articles III (2) and III (4). 373. This could be considered a fairly banal case concerning the

national treatment provisions of the GaTT, but for the fact thatMexico imposed these measures as a consequence of a long-runningdispute with the united States under the north american Free Tradeagreement (naFTa) 730. Mexico had initially invoked naFTa’s dis-pute settlement procedures in relation to uS restrictions on importsof Mexican sugar, which Mexico claimed were contrary to a specificagreement under naFTa guaranteeing market access. naFTaChapter 20 envisaged that disputes would be resolved by a stand-ing roster of panellists, failing which naFTa panels were to beassembled on an ad hoc basis. unfortunately this permitted theunited States to obstruct dispute settlement by simply refusing toappoint its panellists 731. The Mexican Congress introduced theGaTT-inconsistent trade measures as a form of countermeasure,although they were not well articulated as such.374. Mexico argued that the wTO dispute was inextricably

linked to the pre-existing naFTa dispute, and asked that jurisdictionbe declined accordingly. at the centre of the Mexican argument wasthe notion that the wTO’s adjudicative bodies possessed certainimplied powers, including the capacity to decline jurisdiction “in circumstances where the underlying or predominant elements of adispute derive from rules of international law under which claims

218 J. Crawford

729. Panel Report, Mexico — Tax Measures on Soft Drinks and Other Bever-ages, wT/DS308/R, 7 October 2005 ; appellate Body Report, Mexico — TaxMeasures on Soft Drinks and Other Beverages, wT/DS308/aB/R, 6 March2006. See further C. Henckels, “Overcoming Jurisdictional Isolationism at thewTO-FTa nexus : a Potential approach for the wTO” (2008) 19 EJIL 576-579.730. washington, 8 and 17 December 1992, Ottawa, 11 and 17 December

1992, Mexico City, 14 and 17 December 1992 (in force, 1 January 1994), (1993)32 ILM 289 .731. Ibid., art. 2011.

cannot be judicially enforced in the wTO, such as naFTa provi-sions” 732. In this it was unsuccessful. The Panel considered that theterms of the wTO Dispute Settlement understanding 733 did not per-mit it to choose whether to decline an otherwise valid jurisdiction 734.Declining jurisdiction would amount to a failure to address the mat-ter before it and would, moreover, diminish the rights of the unitedStates 735. In light of DSu article 3.10, which States that “Membersshould not link complaints and counter-complaints in regard to distinct matters”, Mexico had no business linking its naFTa disputeto the wTO case, and of course the Panel had no authority to rule onbreaches of naFTa 736.375. This decision was upheld by the appellate Body, which said

that

“it is difficult to see how a panel would fulfill [its] obligation[to adjudicate a valid dispute] if it declined to exercise validlyestablished jurisdiction and abstained from making any findingon the matter before it” 737.

For the Panel to decline jurisdiction would be to “disregard or modify” the provisions of the DSu 738. The presumption of the lan-guage of the DSu and the reasoning of the Panel and appellateBody is that in declining jurisdiction, the united States would be deprived of any legal recourse against Mexico’s actions. as it


732. Panel Report, Mexico — Soft Drinks, op. cit., para. 4.103.733. understanding on Rules and Procedures Governing the Settlement of

Disputes, Marrakesh, 14 april 1994 (in force, 1 January 1995), 1869 UNTS 401(DSu). The DSu is annex 2 to the Marrakesh agreement establishing the worldTrade Organization.734. The Panel placed particular reliance on DSu, art. 11, which provides

relevantly that “a panel should make an objective assessment of the matterbefore it, including an objective assessment of the facts of the case and theapplicability of and conformity with the relevant covered agreements”.735. Panel Report, Mexico — Soft Drinks, op. cit., paras. 7.4-7.9, noting DSu,

arts. 3.2. and 19.2.736. Ibid., paras. 7.11, 7.15.737. appellate Body Report, Mexico — Soft Drinks, op. cit., para. 51.738. Ibid., para. 46. The appellate Body here noted its earlier decision India

— Patents, where it was said :

“[a]lthough panels enjoy some discretion in establishing their own workingprocedures, this discretion does not extend to modifying the substantiveprovisions of the DSu . . . nothing in the DSu gives the panel the authorityeither to disregard or to modify . . . explicit provisions of the DSu” :appellate Body Report, India — Patent Protection for Pharmaceutical andAgricultural Chemical Products, wT/DS50/aB/R, 19 December 1997,para. 92.

happened, the naFTa tribunal the united States had deliberatelystymied was capable of determining both sets of complaints together 739 — the result of the action, therefore, was to depriveMexico of access to justice under naFTa, Chapter 19.376. The particular wording of jurisdictional provisions may lead

to adverse outcomes, even within the same system. The signal example of such intra-system conflict occurred in the investmentarbitration cases of Lauder v. Czech Republic 740 and CME v. CzechRepublic 741, which resulted in what has been described as “the ulti-mate fiasco in investment arbitration” 742. Both cases concerned theinvestment by a uS national, Ronald Lauder, in the Czech privatebroadcaster Tv nova through an investment vehicle CMe, a Dutchcompany owned by Lauder. Following, inter alia, the alleged inter-ference of the Czech Media Council in the operations of Tv nova,Lauder brought a personal action under the Czech/united Statesbilateral investment treaty. The Lauder tribunal held unanimouslythat although the Czech Republic had committed certain breaches ofthe BIT, these nevertheless did not give rise to liability because oflack of causation of damage suffered. Shortly after the Lauder pro-ceedings were initiated, however, CMe launched its own investmentproceedings under the Czech/netherlands bilateral investment treatymaking substantially the same allegations on the basis of the samefacts. Shortly after the Lauder award was rendered, the CME tribunal issued a partial award that reached diametrically the oppo-site conclusion, eventually awarding the claimant uS$270 million indamages (a sum proximate to the Czech Republic’s annual healthbudget at the time). an attempt to have the award set aside beforethe Swedish courts 743 failed on the basis that the formal non-identityof Lauder as an individual and CMe as a company prevented theapplication of principles of res judicata or lis alibi pendens so as to

220 J. Crawford

739. Henckels, op. cit., 579.740. Lauder v. Czech Republic, (2001) 9 ICSID Reports 66.741. CME Czech Republic BV (The Netherlands) v. Czech Republic, Partial

Award, (2001) 9 ICSID Reports 121 ; Final Award, (2003) 9 ICSID Reports 264.742. a. Reinisch, “The Proliferation of International Dispute Settlement

Mechanisms : The Threat of Fragmentation vs. the Promise of a More effectiveSystem ? Some Reflections from the Perspective of Investment arbitration”, in I. Buffard, J. Crawford, a. Pellet and S. wittich (eds.), International Lawbetween Universalism and Fragmentation : Festschrift in Honour of GerhardHafner (Leiden, Martinus nijhoff, 2008), p. 116.743. Czech Republic v. CME Czech Republic BV (Judicial Review : Sweden),

(2003) 9 ICSID Reports 439.

prevent concurrent litigation of the same dispute before different tribunals 744.

2. Proliferation and comity

377. Such cases — one inter-systemic, one intra-systemic — givecause for concern. apart from diminishing confidence in theintegrity of international dispute settlement, and the systematicity ofinternational law, one cannot say justice was done in either of thesesituations. But, as with substantive issues associated with fragmen-tation, the problem is not necessarily pathological. Similar problemshave been faced and largely overcome as between domestic courts ortribunals by techniques of private international law. 378. within private international law, the willingness to surrender

jurisdiction or stay proceedings in the event of a more appropriateforum is generally justified by reference to “comity” 745. as theSupreme Court of Canada said in one case :

“Comity is the recognition which one nation allows withinits territory to the legislative, executive or judicial acts ofanother nation, having due regard both to international dutyand convenience, and to the rights of its own citizens or ofother persons who are under the protection of its law.” 746

379. Comity represents an exercise of discretion by the court ortribunal that weighs its own jurisdiction against the interests of theparties and the conflicting jurisdiction, actual or anticipated, of othercourts or tribunals747. It is

“a flexible doctrine enabling the cooperation of tribunals in theinternational legal order. . . . [I]t can rationalize the tensionbetween an international dispute settlement forum’s jurisdictionand the non-hierarchical nature of such fora. In the sense that it


744. See further C. n. Brower and K. Sharpe, “Multiple and Conflictingarbitral awards” (2003) 4 JWIT 211 ; a. Reinisch, “The use and Limits of ResJudicata and Lis Pendens and Procedural Tools to avoid Conflicting DisputeSettlement Outcomes” (2004) 3 LPICT 37. See further C. McLachlan, “LisPendens in International Litigation” (2008) 336 Recueil des cours 203.745. R. Fentiman, International Commercial Litigation (Oxford, OuP, 2010),

pp. 579-586.746. Morguard v. De Savoye, [1990] 3 SCR 1077, 1096, citing Hilton v.

Cuyot, 159 uS 113 (1895).747. See, e.g., Henckels, op. cit., 584-585 ; Shany, op. cit., pp. 260-266.

functions as a principle for resolving issues of overlappingjurisdiction, it operates to permit a tribunal to limit its ownjurisdiction where exercise of that jurisdiction would be unreasonable or inappropriate . . .” 748

380. This may result in the court or tribunal declining jurisdictionor staying proceedings over matters more appropriately heard else-where. It is a discretionary power, not a legal requirement ; no textual reading of a court or tribunal’s constitutive instrument or pro-cedural rules such as that undertaken by the Panel and appellateBody in Mexico—Soft Drinks will mandate its use. It derives fromthe inherent power of international courts and tribunals to protect theintegrity of the judicial process. as Judge Higgins said in the Use ofForce cases 749

“[t]he Court’s inherent jurisdiction derives from its judicialcharacter and the need for powers to regulate matters connectedwith the administration of justice, not every aspect of whichmay have been foreseen in the Rules” 750.

She referred to

“the very occasional need to exercise inherent powers mayarise as a matter in limine litis, or as a decision by the Court notto exercise a jurisdiction it has” 751.

381. an example of how such an approach might affect the contours of international dispute settlement occurred in serial liti-gation brought by Ireland against the united Kingdom concerningthe operation of the MOX nuclear reprocessing plant at Sellafield 752.Four decisions were produced : (1) a judgment of the Interna-tional Tribunal for the Law of the Sea on a request for provisionalmeasures 753 ; (2) an arbitration award under the OSPaR Conven-

222 J. Crawford

748. Henckels, op. cit., 584.749. Legality of Use of Force (Serbia and Montenegro v. Belgium) (Serbia

and Montenegro v. Canada) (Serbia and Montenegro v. Germany) (Serbia andMontenegro v. Italy) (Serbia and Montenegro v. Portugal) (Serbia and Monte-negro v. United Kingdom), Preliminary Objections, Judgment, ICJ Reports2004.750. Ibid., pp. 1361-1362 (para. 10) (sep. op.).751. Ibid., p. 1362 (para. 11) (sep. op.).752. McLachlan, op. cit., pp. 447-451.753. MOX Plant (Ireland v. United Kingdom), Provisional Measures, (2002)

126 ILR 334.

tion 754 in proceedings for access to information concerning the operation of the plant 755 ; (3) an order by an unCLOS annex vIItribunal 756 ; and (4) a judgment of the european Court of Justice inproceedings brought against Ireland by the european Commissionholding that commencing the unCLOS proceedings breached eulaw 757. The ITLOS proceedings determined that the distinct character of each of the relevant proceedings meant that unCLOSarticle 282 did not deprive it of jurisdiction 758. Similar conclusionswere reached by the PCa tribunal in the OSPaR proceedings 759. Theannex vII tribunal, however, decided to stay proceedings on thebasis that

“there is a real possibility that the european Court of Justicewill be seized of the question of whether the provisions of the Convention on which Ireland relies are matters in relationto which competence has been transferred to the europeanCommunity, and indeed . . . the exclusive jurisdiction of theeuropean Court of Justice . . . In these circumstances, and bear-ing in mind considerations of mutual respect and comity whichshould prevail between judicial institutions both of which maybe called upon to determine rights and obligations as betweentwo States, the Tribunal considers that it would be inappro-priate for it to proceed further with hearing the Parties on themerits of the dispute in the absence of the resolution of the


754. Convention for the Protection of the Marine environment of the north-east atlantic, Paris, 22 September 1992 (in force, 25 March 1998), 32 ILM 1069(OSPaR Convention).755. Dispute concerning Access to Information under Article 9 of the OSPAR

Convention (Ireland v. United Kingdom), (2003) 126 ILR 334.756. MOX Plant (Ireland v. United Kingdom), Annex VII, (2003) 126 ILR

310.757. European Commission v. Ireland, Case C-0459/03, [2006] eCR I-4635.758. MOX Plant, Provisional Measures, op. cit., 273-274. For the contrary

approach, see Southern Bluefin Tuna (Australia v. Japan ; New Zealand v.Japan), Annex VII, (2000) 119 ILR 508 ; a. Boyle, “The Southern Bluefin Tunaarbitration” (2001) 50 ICLQ ; C. McLachlan, op. cit., pp. 443-446. a similar situation threatened to arise as between ITLOS and the wTO in Conservationand Sustainable Development of Swordfish Stocks in the South-Eastern PacificOcean (Chile/European Union), ITLOS Case no. 7 (Order, 20 December 2000),but the matter was discontinued before it could materialize : M. a. Orellana,“The Swordfish Dispute between the eu and Chile at the ITLOS and the wTO”(2002) 71 Nordic JIL 55. On the operation of united nations Convention on theLaw of the Sea, Montego Bay, 10 October 1982 (in force, 16 november 1994),1833 UNTS 396, art. 282, see D. Rothwell and T. Stephens, The InternationalLaw of the Sea (Oxford, Hart, 2010), pp. 445-448.759. OSPAR Convention Dispute, op. cit., 378.

problems referred to. Moreover, a procedure that might resultin two conflicting decisions on the same issue would not behelpful to the resolution of the dispute between the Parties.” 760

382. as perhaps anticipated by the Tribunal, the european Courtof Justice subsequently held that Ireland had no right to bring arbi-tration proceedings before any other forum with respect to the MOXPlant 761.383. To apply this principle to the examples given above, comity

would operate so as to give the Panel in Mexico—Soft Drinks capacity to decline jurisdiction in favour of the naFTa tribunal 762.Similarly, the tribunal in CME, knowing that Lauder was on foot,should have stayed its hand notwithstanding the lack of formal identity between the claimants in the two arbitrations.384. Comity is not the only solution. The concepts of lis alibi

pendens and res judicata, flexibly applied, could provide a means toprevent duplication of proceedings. But despite attempts by somescholars to argue that a general principle of lis pendens exists ininternational law 763, no court or tribunal has yet pronounced on itsscope. 385. To conclude, the problems that result from proliferation tend

to emerge not due to a failure of the system as a whole, but due tothe rigidity of the jurisdictional provisions of individual courts or tribunals. It is true, as the appeals Chamber said in Tadić, that inter-national law does not provide for an integrated judicial system, butthis does not mean that judges and arbitrators are justified in throw-ing up their hands. The tools necessary to address problems of pro-liferation are already available.

D. Self-contained Regimes

386. Finally, I turn to self-contained regimes. The term has arespectable provenance, deriving from the decision of the PermanentCourt in The SS “Wimbledon”. The Court was faced with the ques-tion whether the provisions of the Treaty of versailles 764 relating toGerman waterways applied to entitle non-neutral passage through

224 J. Crawford

760. MOX Plant, Annex VII, op. cit., 318-320.761. European Commission v. Ireland, op. cit., paras. 123-127, 135.762. Henckels, op. cit., 592-598.763. McLachlan, op. cit., p. 500.764. versailles, 28 June 1919 (in force, 10 January 1920), 225 CTS 188.

the Kiel Canal at a time when Germany was not a belligerent. TheCourt noted that the Treaty contained a special section on the Canal,which differed from its general provisions relating to watercourses :

“The provisions relating to the Kiel Canal in the Treaty ofversailles are therefore self-contained ; if they had to be supple-mented and interpreted by the aid of those referring to theinland navigable waterways of Germany in the previous sections of Part XII, they would lose their ‘raison d’être’ . . .The idea which underlies [the specific provisions regarding theKiel Canal] is not to be sought by drawing an analogy fromthese provision but rather by arguing a contrario, a method ofargument which excludes them.” 765

387. The “SS Wimbledon” concerned the interaction betweengeneral and specific rules within the same treaty : the Court could aswell have applied the lex specialis rule as have referred to self-contained provisions. But the International Court went further inTehran Hostages, holding that the legal consequences described inthe vienna Convention on Diplomatic Relations766 (for example,declaring a diplomat persona non grata) were not merely self-contained but amounted to a regime :

“The rules of diplomatic law . . . constitute a self-containedregime which, on the one hand, lays down the receiving State’sobligations regarding the facilities, privileges and immunitiesto be accorded to the diplomatic missions and, on the other,foresees their possible abuse by members of the mission andspecifies the means at the disposal of the receiving State tocounter any such abuse. These means are, by their very nature,entirely efficacious.” 767

388. Both cases were taken into account in the ILC’s work onState responsibility. aRSIwa article 55 codifies the lex specialisprinciple 768, providing that


765. The SS “Wimbledon” (United Kingdom, France, Italy and Japan v.Germany ; Poland intervening), (1923) PCIJ, Ser. A, No. 1, p. 24.766. vienna, 18 april 1961 (in force, 24 april 1964), 500 UNTS 95.767. United States Diplomatic and Consular Staff in Tehran (United States of

America v. Iran), Judgment, ICJ Reports 1980, p. 38 (para. 86).768. Translated : “the specific derogates from the general”. See further

G. Fitzmaurice, “The Law and Procedure of the International Court of Justice1951-4 : Treaty Interpretation and Other Treaty Points” (1957) 33 BYIL 237.

“[t]hese articles do not apply where and to the extent that theconditions for the existence of an internationally wrongful act orthe content or implementation of the international responsibilityof a State are governed by special rules of international law”.

The Commentary makes reference to both “strong” and “weak”forms of lex specialis, with only the former constituting a self-contained regime, properly so-called 769. 389. Thus, in a broad sense, a self-contained regime is little more

than a strong form of lex specialis, by which a “geographically orfunctionally limited treaty series” attempts to contract out of the secondary rules of international law that underpin the system as awhole 770. More narrowly, however, such regimes may representcomprehensive sub-systems that cover a particular international lawproblem in a different manner from how it might be otherwise dealtwith 771. Such subsystems, some have suggested, threaten the coher-ence of international law : “when such deviations become generaland frequent, the unity of the law suffers” 772. One is reminded of thewords of the prophet ezekiel : “[t]heir appearance and their workwas as it were a wheel in the middle of a wheel” 773. 390. Hints of a dystopian future might be seen in those self-

contained regimes which are stronger than the system of diplomaticrelations. a commonly cited example is that of the wTO 774. It pos-sesses a detailed set of primary norms and a compulsory system ofdispute settlement. Its specialized regime of remedies — notably thesystem of suspension of concessions contained in DSu, articles 19and 22 — is comprehensive, leading some to the conclusions that intransiting from the GaTT, the wTO moved decisively towards aself-contained regime 775. 391. another example is that of the european union 776. The

226 J. Crawford

769. ILC Ybk 2001/II (2), p. 140 (para. 5).770. Conclusions of the ILC Study Group, un doc. a/Cn.4/L.702, 18 July

2006, para. 10.771. Report of the ILC Study Group, op. cit., para. 128.772. Conclusions of the ILC Study Group, op. cit., para. 10.773. ezekiel 1 :16 (King James version).774. See, e.g., Report of the ILC Study Group, op. cit., paras. 165-171 ;

Simma and Pulkowski, “Leges Speciales and Self-Contained Regimes”, op. cit.,pp. 155-158.775. P. J. Kuyper, “The Law of the GaTT as a Special Field of International

Law” (1994) 25 NYIL 252.776. See, e.g., Simma and Pulkowski, “Leges Speciales and Self-Contained

Regimes”, op. cit., 152-155.

Treaty establishing the european Community 777 has established itsown legal system 778, one which requires that members “shall not failto carry out their obligations and shall not take the law into theirown hands” 779. To the extent that the customary rules of Stateresponsibility are applicable, they are only residual — and as the system of european law matures, any remaining lacunae are gradually being plugged 780.

1. Self-contained regimes, lex specialis and the law of treaties

392. But again the problems posed by self-contained regimesshould not be exaggerated. If States wish to enter into comprehen-sive relationships that, in effect, contract out of the remainder of thelaw (peremptory norms aside 781) they are free to do so. The task ofinternational law in such a circumstance is to ensure that a frame-work “through which such systems may be assessed and managed ina legal-professional way” 782. Conflicts between general and specificnorms of international law may thereby be prevented, and a work-able compromise reached 783.393. However, it is rather difficult for an inter-State system

to contract completely out of general international law — indeed, there does not presently exist an entirely self-contained interstateregime 784. For example, the wTO appellate Body has acknowledgedthat the GaTT “is not to be read in clinical isolation from publicinternational law” 785. In Korea — Procurement, the Panel said thatcustomary law remains a source of law for the wTO “to the extent


777. 24 December 2012, 2002 OJ, C-325/35.778. Costa v. ENEL, Case 6/64, [1964] eCR 585 ; Van Gend en Loos, Case

26/62, [1963] ECR 12.779. European Commission v. Luxemburg & Belgium, Cases 90/63 and 91/63,

[1964] ECR 625. 780. Simma and Pulkowski, “Leges Speciales and Self-Contained Regimes”,

op. cit., 153.781. OSPAR Convention Dispute, op. cit., 364.782. Conclusions of the ILC Study Group, op. cit., para. 12.783. See, e.g., vienna Convention on the Law of Treaties, vienna, 23 May

1969 (in force, 27 January 1980), 1155 UNTS 331.784. Report of the ILC Study Group, op. cit., para. 120 ; Simma and

Pulkowski, “Leges Speciales and Self-Contained Regimes”, op. cit., 143.785. appellate Body Report, United States of America — Standards for

Reformulated and Conventional Gasoline, wT/DS2/aB/R, 26 april 1996, p. 19.See further L. Bartels, “applicable Law in wTO Dispute SettlementProceedings” (2001) 35 JWT 499-519 ; P. van den Bossche, The Law and Policyof the World Trade Organization (Cambridge, CuP, 2nd ed., 2008), pp. 55-63.

that the wTO treaty agreements do not ‘contract out’ from it” 786 :this is the default position in international law for any treaty. Similarconclusions may be reached with respect to the system’s treatment ofgeneral principles of law 787. Indeed the wTO is one of the mostinfluential users of the vCLT in matters of treaty interpretation 788.Similarly, the european Court of Justice has asserted that there is aresidue of general principles of international law within the contextof eu law, and in the interpretation of an international agreement,has indicated its willingness to defer to the decisions of tribunalsestablished pursuant to that agreement 789. as Simma and Pulkowskinote, “the term ‘self-contained regime’ should not be used to circumscribe the unrealistic hypothesis of a fully autonomous legalsystem” 790.

E. Conclusion : The Centre Holds

394. all this suggests that the problems to which the label of“fragmentation” has been attached may be little more than by-products of a maturing system of law, albeit one lacking in verticalintegration. Given that international law grew from bilateral relation-ships, it is difficult to see how anything has become more frag-mented than it was at the beginning : it has just become morediverse. Multilateralism never meant complete coherence of treatypractice or State interest. If States are free to join multilateraltreaties, they are free to create a partly fragmented system. 395. and yet, and yet. The instinct of any system is its own

preservation and perpetuation : international law is no different. Inthe midst of a period of immense, even exponential, growth, itretains the tools necessary to maintain its own coherence. This has been shown for substantive fragmentation (the allocation of

228 J. Crawford

786. Panel Report, Korea — Measures Affecting Government Procurement,wT/DS163/R, 1 May 2000, para. 7.96. See further J. Pauwelyn, Conflict ofNorms in Public International Law : How WTO Law Relates to Other Norms ofInternational Law (Cambridge, CuP, 2003), pp. 210-211, 470.787. See, e.g., appellate Body Report United States of America — Import

Prohibition of Certain Shrimp and Shrimp Products, wT/DS58/aB/R, 12 October 1998, para. 158. See further van den Bossche, op. cit., pp. 56-57 ; a. D. Mitchell, Legal Principles in WTO Disputes (Cambridge, CuP, 2008).788. See generally I. van Damme, Treaty Interpretation by the WTO

Appellate Body (Oxford, OuP, 2009).789. EEA I, Opinion 1/91, [1991] eCR I-6079, paras. 39-40.790. Simma and Pulkowski, “Leges Speciales and Self-Contained Regimes”,

op. cit., 143.

preferences within the system), institutional proliferation (the intro-duction of comity to international dispute settlement) and self-con-tained regimes (application of the lex specialis rule and the vCLT to manage and resolve conflict). The result is a system whichitself acts as guide, in the mode of Dante’s virgil :

“ ‘You, as you speak, have so disposed my heart in keen desire to journey on the way that I return to find my first good purpose. Set off ! a single will inspires us both.’. . . . . . . . . . . . . . . . . . . . . . .all this I said to him as he moved on. I entered on that deep and wooded road.” 791


791. Dante, op. cit., II.136-142.

“ ‘Tu m’hai con disiderio il cor dispostosi al venir con le parole tue,ch’i’ son tornato nel primo proposto.Or va, ch’un sol volere è d’ambedue : tu duca, tu segnore e tu maestro . . .’ Così li dissi ; e poi che mosso fue, intrai per lo cammino alto e silvestro.”

CHaPTeR X

unIveRSaLITY OF InTeRnaTIOnaL Law

“nothing could be more fallacious than to judgeof China by any european standard.”

George, 1st earl Macartney (1794) 792.

“The General AssemblyProclaims this universal Declaration of Human

Rights as a common standard of achievement forall peoples and all nations . . .”

Ga res. 217a (III), 10 December 1948.

A. Introduction

396. Modern international law claims to be universal as far as ourone world is concerned. It even has rules about how we are tobehave in outer space, and some of these are in “all States” form ;whether they are binding on extra-terrestrials is a so-far-unansweredquestion 793. But humanly speaking, one may think there is no greaterform of universality than that. 397. Yet our international law has a specific origin in european

political thought and practice. These european origins of what wenow call general international law create a considerable tension.Despite its claim to universality, it is often thought that internationallaw cannot attract the real allegiance of other cultures and peoples.at most they will observe it for pragmatic reasons, and then onlyapproximately. Meanwhile the peoples of the book (whether thebook is the De Jure Belli ac Pacis of Grotius or Le droit des gens of

230

792. J. L. Cranmer-Byng, An Embassy to China : Being the Journal kept byLord Macartney during his Embassy to the Emperor Ch’ien-lung 1793-1794(London, Longmans, 1962), p. 219, quoted in a. Peyrefitte, The Collision of Two Civilisations : The British Expedition to China in 1792–4 (trans. J. Rothschild, London, Harvill, 1993), p. vii.793. See, e.g., agreement Governing the activities of States on the Moon and

Other Celestial Bodies, new York, 5 December 1979 (in force, 11 July 1984),1363 UNTS 3 (applies to States (art. 3) who along with intergovernmental orga-nizations may bear international responsibility for conducting space activities(arts. 14, 16)) ; Treaty on Principles Governing the activities of States in theexploration and use of Outer Space, including the Moon and Other CelestialBodies, washington, Moscow, London, 27 January 1967 (in force, 10 October1967), 610 UNTS 205 (applies to States (art. XIII) ; intergovernmental organiza-tions may also be internationally responsible (art. vI)).

vattel or the Max Planck Encyclopedia of our own time) proclaim itas intimately associated with what is right. The tension also mani-fests itself when developments in international law are under discus-sion — whether these concern, for example, a putative right todemocratic governance going beyond the meagre provisions ofarticle 25 of the ICCPR or the ambitiously expressed rights ofindigenous peoples articulated in the united nations Declaration onthe Rights of Indigenous Peoples. as to the former, how can interna-tional law require democracy as the form of government when manyStates, including some of the world’s largest and most powerful, arenot democratic or are only superficially so 794 ? as to the latter, whatroom is there for special indigenous rights in non-settler societieswhich are still struggling to achieve national unity and economicdevelopment with all that entails 795 ?398. One way to resolve this tension is to redefine international

law as a coalition of the virtuous 796 or the like-minded 797, for


794. See T. M. Franck, “The emerging Right to Democratic Governance”(1992) 86 AJIL 46-91 ; G. H. Fox and B. R. Roth (eds.), DemocraticGovernance and International Law (Cambridge, CuP, 2000). For a scepticalview see J. Crawford, “Democracy and the Body of International Law”, in Foxand Roth, op. cit., pp. 91-120. The issue of democracy is further discussed inChap. XII. 795. See united nations Declaration on the Rights of Indigenous Peoples,

un doc. a/61/L.67, 12 September 2007, annex. See also S. J. anaya, Indige-nous Peoples in International Law (new York, OuP, 2nd ed., 2004) ; C. Charters, “Indigenous Peoples and International Law and Policy”, in S. Imai, B. J. Richardson and K. Mcneil, Indigenous Peoples and the Law : Comparativeand Critical Perspectives (Oxford, Hart, 2009), pp. 161-194 ; a. Xanthaki,Indigenous Rights and United Nations Standards : Self-Determination, Cultureand Land (Cambridge, CuP, 2010) ; M. Castan, “DRIP Feed : The SlowReconstruction of Self-determination for Indigenous Peoples”, in S. Joseph anda. McBeth (eds.), Research Handbook on International Human Rights Law(Cheltenham, edward elgar Publishing, 2010), pp. 492-511 ; K. engle, “OnFragile architecture : The un Declaration on the Rights of Indigenous Peoplesin the Context of Human Rights” (2011) 22 EJIL 121 ; S. allen and a. Xanthaki (eds.), Reflections on the UN Declaration on the Rights ofIndigenous Peoples (Oxford, Portland, Hart Publishing, 2011) ; e. Pulitano (ed.),Indigenous Rights in the Age of the UN Declaration (Cambridge, CuP, 2012).796. Hilf al-Fudul, meaning “The alliance of the virtuous”, was a convention

agreed in Mecca in the year 590 between the warring chiefs of various kingdomsin the region. Parties to the alliance pledged to resist oppression and injustice, toabide by a set of rules regulating war and peace, to safeguard trade and to pro-tect visiting delegates and merchants. The formation of the alliance was said tohave been attended by the prophet Muhummad 20 years before the advent ofIslam : see, e.g., S. Khatab and G. D. Bouma, “Islamic International Law”, inDemocracy in Islam (Routledge, Oxford, 2007), pp. 167-168.797 F. O. Hampson, “Climate Change : Building International Coalitions of

the Like-minded” (1989/1990) 45 International Journal 58-61.

example to condition statehood upon democracy 798, the observanceof human rights or l’état de droit 799. under this strategy (which asfar as I know has never been espoused by someone who was not anational of a virtuous State), international law continues to be —what some thought it was in the late nineteenth century — the law ofand between self-proclaimed civilized States only 800.399. But whether that view of international law ever prevailed, it

is completely unacceptable now. It is of the essence of modern inter-national law that it is universal. we are struggling towards the pointwhere there is a cadre of treaties to which every State without excep-tion is party ; let us not engage in the error of exclusivity by refer-ence to some subset of so-called liberal values and so-called liberalStates. That would only risk reintroducing colonialism in anotherguise. The “clash of civilizations” 801 is contained within interna-tional law ; it is not set against it by any principle of exclusion orsubordination.400. In this chapter we will begin by considering how this situa-

tion came to be, how a system that originated in europe now claimsuniversality. we will then consider what the universality of interna-tional law amounts to in practice today.

B. The History of Universality

1. Origins : from the seventeenth to the nineteenth century

401. It is true that the historical origins of international law liemainly in a particular time and place : early modern europe. But

232 J. Crawford

798. R. Rich, “Bringing Democracy into International Law” (2001) 12 (3)Journal of Democracy 23-29 ; J. Crawford and S. Marks, “The Global DemocracyDeficit : an essay in International Law and Its Limits”, in D. archibugi, D. Heldand M. Kohler (eds.), Re-Imagining Political Community : Studies in Cosmopo-litan Democracy (Cambridge, Polity Press, 1998), pp. 72-90 ; H. J. Steiner, “Poli-tical Participation as a Human Right” (1988) 1 Harvard Hum. Rights Yb. 129-134.799. S. M. H. nouwen, “Justifying Justice”, in J. Crawford and M. Kosken-

niemi (eds.), The Cambridge Companion to International Law (Cambridge,CuP, 2012), pp. 328-330 ; R. Peerenboom, “Human Rights and the Rule of Law :what’s the Relationship ?” (2004-2005) 36 Georgetown JIL 902 ; R. G. Teitel,“Humanity’s Law : Rule of Law for the new Global Politics” (2001-2002)Cornell Int’l. LJ 365-368. 800. See J. Hornung, “Civilisés et barbares” (1885) 17 RDILC 5-18, 447-470,

539-560 ; (1886) 18 RDILC 188-206, 281-298 ; G. Simpson, Great Powers and Outlaw States : Unequal Sovereigns in the International Legal Order(Cambridge, CuP, 2004), pp. 235-247.801. S. P. Huntington, “The Clash of Civilisations” (1993) 72 (3) Foreign

Affairs 22-49.

even during that period international law accepted neighbouringnon-european States as participants in the system, such as theOttoman empire as early as 1649 802. In effect, as Gerry Simpsonobserves, early modern international law was riven by two seem-ingly contradictory ambitions : “[f]irst, europe was to establish itselfas a unique and superior legal and cultural order” and “[s]econd, it was to export this order through the adoption of universalistforms” 803. The prevalent view was unashamedly chauvinistic — it asserted,

or rather assumed, that europe was a society uniquely amenable tothe international rule of law. 402. This line of thinking was still evident in international law

texts as late as the early twentieth century. Thus Hall’s treatise, theleading treatise by a British author of the time :

“It is scarcely necessary to point out that as international lawis a product of the special civilisation of modern europe, andforms a highly artificial system of which the principles cannotbe supposed to be understood or recognised by countries differ-ently civilised, such states only can be presumed to be subjectto it as are inheritors of that civilisation.” 804

But by that time, what Hall even called the “international society forexport” 805 had been exported to the rest of the world. It had travelledto the further reaches of africa and asia, to the americas, to


802. See, e.g., Instrument for the Prolongation of the Peace between theemperor of the Holy Roman empire and the Sultan of Turkey, 1 July 1649, 1 CTS 457.803. Simpson, op. cit., p. 236.804. w. e. Hall, A Treatise on International Law (London, Oxford & Co.,

6th ed., 1909), p. 39, quoted in Simpson, op. cit., p. 237. Lorimer displays asimilar sentiment. He suggests that States are fundamentally unequal and thatparticipation in the international system of States (what he terms full political“recognition”) is a right belonging only to “sufficiently developed” and“civilised nations”, not to those that are less powerful or whose actions contra-dict “natural law” (which he defines as an expression of “spirit” and the “ethicalideal” as manifested through a correct use of rationality). He contends that eth-nicity, race and religion, other than the superior Christian faith, present obstaclesto advancing a nation’s stage of development : “The right of undeveloped races . . . is not to recognition . . . but to guardianship — that is to guidance —in becoming that of which they are capable, in realising their special ideals”,“we ought, perhaps, to distinguish between the progressive and the non-progres-sive races”. See J. Lorimer, The Institutes of the Law of Nations — A Treatise ofthe Jural Relations of Separate Political Communities (edinburgh, London,william Blackwood and Sons, 1883), pp. 1, 12, 97-99, 102, 156-157.805. Hall, op. cit., p. 236.

australia and to even the most remote Pacific islands in the holds ofgunboats and the baggage of colonizers.403. The case of China is particularly striking. In 1792, Lord

Macartney led a British embassy to the Qīng court, under instruc-tions from his Government to open permanent diplomatic relationsbetween the two countries on the european model. He was treatedwith disdain, kept waiting and eventually sent home bearing a letterof rejection from the Qiánlóng emperor, ruler of all under heaven, toGeorge III of Great Britain. “O King”, it began, “inclining your hearttowards civilization, you have specially sent an envoy to present astate message” from which “your sincere humility and obedience can clearly be seen”. It advised the British monarch to “simply act in conformity with

our wishes by strengthening your loyalty and swearing perpetualobedience so as to ensure that your country may share the blessingsof peace” 806. To this day it is unclear whether Macartney actually kowtowed

(by touching the ground with his head, according to Chinese custom,to display submission), though the emperor’s letter asserts that hedid. 404. The internal rule of law was not a foreign concept in late

eighteenth-century China 807. But the incident shows that China conceived of relations with foreign rulers exclusively in terms ofsubordination ; it had no room for sovereign equality. at the timeChina had only a “Department of Foreign Tribute” rather than aeuropean-style department of foreign affairs 808. Yet eventually, in1860, China did create such a department and did send envoysabroad 809. That was the year that the Second Opium war culminatedin the sacking of the Summer Palace and the “unequal treaties” with

234 J. Crawford

806. a translation of the original Chinese version of the letter is in Peyre-fitte, op. cit., pp. 289-292. The version read by Macartney was translated byJesuit missionaries who had “carefully altered the most insolent formulations,openly proclaiming their desire to remove ‘any offensive turn of phrase’ ” : p. 288.807. It has been suggested that the debates from the eighth to third centuries

BCe between Confucians (who sought to organize society around rules of pro-priety and held that to rely on law was to admit a failure of virtue) and Legalists(who sought to use norms or law to regulate society) “may be roughly com-pared” to Greek debates about “rule of man” and “rule by law” : S. Chesterman,“an International Rule of Law ?” (2008) 56 AJCL 338-389.808. Peyrefitte, op. cit., p. 291 n ‡.809. Ibid.

France, Russia and the united Kingdom, which forced China to cedeterritory, notably Hong Kong 810. Thus even China was cowed.405. Other States that resisted european colonialism, most

notably Japan after the Meiji Restoration of 1868, embraced interna-tional law during processes of internal modernization 811. even at thistime, wheaton’s Elements of International Law (1836) — translatedinto Chinese in 1864 and also adopted in Japan 812 — characterizedinternational law as Christian, “civilized” and european 813. Yet itwas also clear that States such as China and Japan were sovereign,what Simpson refers to as :

“a recognition that states could be part of the international lawsociety while at the same time excluded from the inner circle orfamily. westlake [writing in the late nineteenth century] seemsto envisage a sort of staggered admission policy : ‘Our interna-tional society exercises the right of admitting outside states toparts of its international law without necessarily admittingthem to the whole of it.’ ” 814

406. By the 1920s only seven States that were neither europeannor former european settler colonies had managed to retain theirindependence without formal qualification to their sovereignty : theywere afghanistan, China, ethiopia, Japan, Liberia, Siam (Thailand)and the Ottoman empire (Turkey). There was no serious resistanceto the international system which had been imposed in various waysin the meantime and whose structure was by then recognizably inplace 815. 407. Since the mid-nineteenth century there has been only one

serious attempt on the part of a major State to reject and abandoninternational law. This was pursued by the Russian Socialist Fed-erative Soviet Republic (RSFSR) in the aftermath of the BolshevikRevolution of 1917. The revolutionaries saw international law as an


810. Treaties of Beijing : China-united Kindom, 24 October 1860, 123 CTS71 ; China–France, 25 October 1860, 123 CTS 79 ; China-Russia, 14 november1860, 123 CTS 125.811. S. Yamauchi, “Civilization and International Law in Japan during the

Meiji era (1868-1912)” (1996) 24 Hitotsubashi J. L. & Pol. 24.812. Ibid., p. 11.813. Simpson, op. cit., p. 237. 814. Ibid., p. 238, quoting J. westlake, Collected Papers on International

Law (Cambridge, CuP, 1914), p. 82.815. See J. Crawford, Brownlie’s Principles of Public International Law

(Oxford, OuP, 8th ed., 2012), pp. 4-5.

imperial tool of and between capitalist States 816. The RSFSR ini-tially identified the “State” in the sense of a subject of internationallaw with the State in the Marxist sense of an organization of the ruling class. Since the Russian Revolution had resulted in a changeof ruling class, the RSFSR could deny that it was the continuator ofthe international legal personality of tsarist Russia. what this meantin practice was that it could deny liability for tsarist debts 817. Sovietinternational lawyers such as evgeny Korovin conceptualized this byarguing that a community of ideology shared by the ruling classes ofdifferent nations was a necessary basis for international law. Therecould be particular systems of international law that applied to therights and obligations of States with comparable social structuresand that bound them only insofar as they preserved those structures.a change in social structure as fundamental as the RussianRevolution resulted in a State losing the status of a subject of inter-national law within the international system to which it previouslybelonged 818.408. Despite the emphasis placed on class in the international

relations of the RSFSR — and, from its foundation in 1922, theearly Soviet union — there was little to support Korovin’s views.The Soviet union was able to re-enter the international communitywithout abandoning its social system — indeed the twentieth centurywas full of examples of legal interaction between nominally socialistand capitalist States. as early as 1929, faced with the difficulty thata “socialist” system of international law could not explain relationsbetween the Soviet union and its capitalist neighbours, Korovinacknowledged that his original theories had been wrong 819.409. Thus even the Russian Revolution and the alternative world

view it posited did not lead to anything but a short-lived and incon-

236 J. Crawford

816. J. n. Hazard, “Cleansing Soviet International Law of anti-MarxistTheories” (1938) 32 AJIL 247. This view was put forward by eugenePushukanis, the then premier Soviet international law theorist. Pushukanis’sviews modified over time, exposing his apparent disconnect from the newRSFSR ideology, to the glee of eugeny Korovin, who seized the opportunity toreorient the Soviet view of international law and rise in stature. 817. R. Schlesinger, Soviet Legal Theory : Its Social Background and

Development (London, Kegan Paul, Trench, Trubner & Co., 1945), pp. 276-277.See further J. n. Hazard, “Socialism and International Public Law” (1985) 23Col. J. Transnat’l. L. 251-264.818. Ibid., pp. 275-278, citing e. Korovin, “La république des Soviets et le

droit international” (1925) RGDIP 290.819. Ibid., p. 279.

sequential rejection of international law. nor did the People’s Repub-lic of China repeat the experiment after 1949, despite the lengthydelay on the part of the united States in recognizing it (not until1979) 820. as for the Third world — to use a term then common 821

— the former colonies which won their independence in the mid-twentieth century, virtually without exception and without protest,took their place in the existing international order. If interna-tional law remains the preserve of the “civilized” — and of coursearticle 38 (1) (c) of the Statute of the International Court still refersto “civilized nations” — then all States are now considered to fallinto that category.410. This is not to suggest that the universalization of interna-

tional law was a one-way process of european expansion. arnulfBecker Lorca has argued that a belief in its exclusively europeanslant “limits the scope of analysis and prevents an understanding ofthe global character of the historical processes through which inter-national law became universal” 822. Its universalization was alsoinfluenced by the reinterpretation of certain elements of classicalinternational law — positivism, sovereignty and civilization — byinternational lawyers in “semi-peripheral” societies such as Japan,the Ottoman empire and Latin america :

“[B]y the dawn of the nineteenth century, a significant number of international legal regimes had governed, undersome degree of formal equality, the interaction between someeuropean and non-european sovereigns. During the course ofthe nineteenth century, semi-peripheral appropriations of inter-national legal thought and the global circulation of rules,lawyers, and legal ideas transformed existing international legalregimes into a universal international law. ‘universality’, as aconsequence, describes not only international law’s geographicexpansion, but also doctrinal changes, the global professionali-


820. Consulate General of the united States, Hong Kong and Macau, JointCommuniqué on the Establishment of Diplomatic Relations between the UnitedStates of America and the People’s Republic of China, 15 December 1978<http ://hongkong.usconsulate.gov/uscn_docs_jc1979010101.html> accessed 1 april 2013.821. J. Krieger (ed.), The Oxford Companion to the Politics of the World

(Oxford, OuP, 2001, online 2012).822. a. B. Lorca, “universal International Law : nineteenth-Century

Histories of Imposition and appropriation” (2010) 51 Harvard ILJ 546.

zation of international lawyers, and the transformation of thenature and functions of the international legal discourse.” 823

411. This is an important qualification to the european origins of international law. The expansion of international law was not amatter simply of enforcing european norms on States such as Chinabut also of abandoning such parochial notions as the distinctionbetween civilized and uncivilized States through a process of inter-action between different societies. adjustments were made on bothsides.

2. Challenges : the twentieth and twenty-first centuries

412. Did this close off the possibility of reviving a more paro-chial conception of international law ? It did not, for there persisttendencies in western writing that aim to do just that. Moreoverthere continue to be major divergences of policy, interest andapproach among groups of States. as well as the distinction between“civilized” and “uncivilized” States in the nineteenth century and thecontested notion of “socialist” international law in the twentieth,these have also produced claims to a special status for the law ofdecolonization and development 824 and the phenomenon of regional-ism. 413. But these factors do not contradict the universality of inter-

national law. Rather what they show is that this universality, at leastas we understand it today, is a historically contingent phenomenon,driven particularly by colonization and decolonization, shaped by theideological conflicts of the twentieth century and subject to futuredevelopments. nor is it the result of a Rawlsian negotiation amongStates in which they are postulated to have planned and consented toits basic structure 825. It is the fruit — perhaps more accurately, theearly blossom — of a process of organic growth shaped by countlesschance events, disconnected ideas and diverse influences that mightwell have had quite different results.414. we might then say that international law is subjectively uni-

238 J. Crawford

823. Op. cit. footnote 822, 548.824. See, e.g., a. Cassese, International Law in a Divided World (Oxford,

Clarendon Press, 1986), pp. 115-123 ; M. Bedjaoui, “non-alignment et droitinternational” (1976) 151 Receuil des cours 337.825. See generally J. Rawls, The Law of Peoples : With, the Idea of Public

Reason Revisited (London, HuP, 1999). See further Chap. Xv.

versal : its universality depends on the historical fact that non-european States gradually accepted it as binding law as a matter ofpractice and circumstance. They did not derive it from some uni-versal moral principle. But this does not prevent international lawfrom making claims of objective universality : that it has some moralauthority that would bind States regardless of the fact of their accep-tance. One such claim to moral authority may be the idea of “uni-versal” human rights — “meta-rights” that are pursued by specificlegal regimes but that would exist independently of them. Theuniversal Declaration of Human Rights refers not to the creation orenactment of rights but to “recognition” of the “inherent dignity andof the equal and inalienable rights of all members of the human family” 826. Such claims to universal moral authority will not convince everyone. They rest on the controversial assumptions thatthere is such a thing as moral principle binding on all and that inter-national law (or part of it, such as human rights law) draws authorityfrom such a principle.415. But the language of objective universality, whatever its

philosophical underpinning, shapes how international law operates.Koskenniemi argues that we have become accustomed to thinking ofpolitics as a process of separate identities or groups “seeking recog-nition” but that they can do so only in universal terms : “perhaps aright of self-determination, fair distribution of resources, equality ofopportunity, and so on”. no group, especially not a vulnerable one,“can claim a right merely in terms of its separate ‘value-system’ ” 827.416. There is a danger that this language of objective universality

will act as a cloak for parochial values and interests. Sundhya Pahujatraces how the promised universality of international law “served toconstrain, and ultimately to undermine the radical potential” ofdemands by decolonized States. This was because their attempts touse international law “were subsumed within a pervasive rationalitythat successfully made a claim for the universality of a particular, or‘provincial’ set of values originating in and congenial to the north”,especially “the concepts of development and economic growth” 828.Decolonization was channelled into the formation of the develop-


826. Ga res. 217a (III), 10 December 1948, Preamble.827. M. Koskenniemi, The Gentle Civilizer of Nations : The Rise and Fall of

International Law 1870-1960 (Cambridge, CuP, 2001), pp. 504-505.828. S. Pahuja, Decolonising International Law (Cambridge, CuP, 2011),

pp. 2-3.

mental State ; claims to permanent sovereignty over naturalresources were transformed into the protection of foreign investors ;and “the asserted rule of international law” was transformed into “the internationalisation of the rule of law as a developmentstrategy” 829. Pahuja argues that in order to realize the universalpromise of international law, to “decolonize” it of these parochialvalues and interests, “we need precisely to resist attempts to producea framework that ‘recognises’ the universality of certain values”,including “development as a proxy for human well-being”. Insteadwe must acknowledge “the contingency of any value put forth as universal and the frame of reference supporting the universalclaim” 830.417. Pahuja’s thesis is a warning against pretensions of objective

universality. not only are they philosophically controversial, in prac-tice they may also serve parochial or even anti-universalist interests.we need not accept her specific critique of the concept of develop-ment and other putatively universal values in order to agree withthis. at the same time, she urges us to acknowledge the subjectivityand contingency of the universality of international law. On the pre-mise that international law is subjectively universal — and hencesubject to dialogue and change — we can engage in what I havecalled “international law as process” : the process of claim andcounterclaim, assertion and reaction, in this case applied not to aparticular dispute but at the level of “universal” values and prin-ciples. we can observe this process at work everywhere in internatio-nal law, including in the tension between general and specific legalcategories and between universalism and regionalism.

C. Universality Today

1. Universality versus regionalism

418. an assumption of universality is evident in almost everyarea of international law. This is true not only of customary inter-national law but also of many multilateral treaties and other interna-tional instruments, particularly those that seek to codify or progres-sively develop custom. It is expressed, for example, in the references

240 J. Crawford

829. Pahuja, op. cit., pp. 2-3.830. Ibid., p. 260.

to “all States”, the “international community as a whole”, the “inter-national community of States as a whole” and other such phrases inmultilateral international instruments. Many of these have originatedin the work of the International Law Commission (ILC), whose veryfirst instrument, the draft Declaration on Rights and Duties of States,proclaimed in a resolutely universalist tone : “the States of the worldform a community governed by international law” 831. Later instru-ments drafted by the ILC have tended to avoid references to suchsubsets as “developing” or “newly independent” States 832.419. a number of reasons may be suggested for this tendency

towards universality in multilateral treaties, whose States partiesusually comprise only a fraction — though often a large fraction —of the whole number of States. They include the role of the ILC as a subordinate organ of the united nations — the prototype of theuniversal international organization and still the “most” universal —and the representation within it, according to its Statute, “of the mainforms of civilization and of the principal legal systems of the world” 833. 420. There are, however, also countervailing tendencies. One, in

the special context of the law of the sea, can be identified in instru-ments such as the united nations Convention on the Law of the Seaof 1982 (unCLOS), which recognizes such subsets, for the firsttime, as archipelagic, landlocked and “geographically disadvan-taged” States 834.421. while such conventional law developments as “archipelagic

States” provide a functional salience to a subset of “island States” —a term that is itself not free from definitional debate — it also creates problems. First, the unCLOS definition of an “archipelagicState” is arbitrary. It requires that an archipelagic State be “consti-tuted wholly by one or more archipelagos and may include otherislands” 835. applying this definition, together with the rules on


831. ILC Ybk 1949, p. 287 (Preamble, para. 1).832. See, e.g., J. Crawford, “universalism and Regionalism from the

Perspective of the work of the International Law Commission”, in InternationalLaw on the Eve of the Twenty-first Century : Views from the International LawCommission (new York, united nations, 1997), p. 99.833. Statute of the International Law Commission, Ga res. 174 (II),

21 november 1947 (as amended), art. 8.834. Montego Bay, 10 December 1982 (in force, 16 november 1994), 1833

UNTS 3 (unCLOS), arts. 7 (2), 38 (1), Part Iv, Part X, art. 70.835. Ibid., art. 46 (a) (emphasis added).

archipelagic baselines 836, seems to exclude any offshore archipelagothat is part of a mainland State, for example Hawaii. This amounts todiscrimination and is a deviation from formal equality under interna-tional law 837. Second, by establishing archipelagic State sovereigntyover archipelagic waters and superjacent air space 838, unCLOS created one of the largest expansions (comparable to exclusive economic zones) of maritime jurisdiction in the history of the law ofthe sea, based entirely on the on grounds of certain States’ “special circumstances” 839. even if an archipelagic State’s geographical lay-out is correlated with its developmental status, as a matter of legalprinciple it still gains a maritime windfall.422. The category of geographically disadvantaged States

presents another curious innovation. The essence of this category isthat both developed and developing coastal States are entitled tosome of the surplus living resources (i.e. fish) from the exclusiveeconomic zone of neighbouring States under two conditions : (1) the“disadvantaged” State is dependent on the other State’s exclusive economic zone for the “nutritional purposes” of their population, or (2) it has no exclusive economic zone itself. These categoriesmake very little sense. One category relates to dependencyupon another State’s waters for subsistence, the other relates to alegal consequence of State maritime boundaries. This means that inprinciple, subject to the technical rules for exploitation 840, a self-sufficient developed coastal State that does not have an exclusiveeconomic zone but has a healthy population, is just as entitled tosome excess fish of a neighbouring State as is a developing coastalState that has an exclusive economic zone which cannot support itspopulation but depends on sustenance from that other exclusive eco-nomic zone 841. In such cases as these the literal interpretation of

242 J. Crawford

836. unCLOS, art. 47.837. J. Crawford, “Islands as Sovereign nations” (1989) 38 ICLQ 297. See

also D. P. O’Connell, The International Law of the Sea (I. Shearer (ed.), Oxford,Clarendon Press, 1982), vol. 1, p. 256.838. Subject only to transit and innocent passage : unCLOS, op. cit.,

arts. 52-53.839. Crawford, “Islands as Sovereign nations”, op. cit., p. 296.840. unCLOS, op. cit., arts. 61, 62, 70 (3)-(4), (6).841. G. Hafner, “Geographically Disadvantaged States”, in R. wolfrum (ed.),

The Max Planck Encyclopedia of Public International Law (OuP, 2008, onlineedition), para. 1. One restriction, however, is that developed geographically dis-advantaged States may only participate in the exclusive economic zone of otherdeveloped States : unCLOS, op. cit., art. 70 (5).

“geographic disadvantage” seems not to be the primary reason thataccess is granted. 423. Thus even though decisions to establish special categories of

States may be forged with good intentions, care must be taken tomaintain the integrity of the principle of equality when consideringthe development of special rights 842. 424. More prominent is the phenomenon of regionalism.

although the situation of each State can be attributed to its “place inthe world”, that “place” tends to be seen in the first instance in termsof the State’s immediate neighbours. The matters that mostGovernments spend most time on and which can affect them mosttend to relate to their neighbours or their region. Indeed even whenStates focus on questions of apparently universal concern, theirapproach is likely to be influenced by regional postures and impli-cations — and not only in such matters as minority rights or the use of international watercourses. Since 1945, regional approacheshave developed in areas of international law such as peace and security 843, the marine environment844, exploitation of marine naturalresources845, the settlement of disputes 846, disarmament 847 (especially


842. Crawford, “Islands as Sovereign nations”, op. cit., pp. 286-287.843 See Charter of the united nations, Chap. vIII, the various regional

arrangements and alliances and the attempts to develop regional security doc-trines during the Cold war. See also w. Rogers and w. M. Reisman, “TheBrezhnev Doctrine and the Reagan Doctrine : apples and Oranges ?” (1987) 81ASIL Proceedings 561. There has been a significant regional involvement insome “peacekeeping” issues (e.g., Liberia, Haiti), and the european unionasserted priority of concern at various stages of the Yugoslav crisis.844. as in the “regional seas” conventions : see, e.g., B. Boer, “environ-

mental Law and the South Pacific : Law of the Sea Issues”, in J. Crawford andD. Rothwell (eds.), The Law of the Sea in the Asian Pacific Region (Dordrecht-London, Martinus nijhoff, 1995), p. 67.845. For example, the role of regional fisheries agencies under the agreement

for the Implementation of the Provisions of the united nations Convention onthe Law of the Sea of 10 December 1982 relating to the Conservation andManagement of Straddling Fish Stocks and Highly Migratory Fish Stocks, newYork, 4 august 1995 (in force, 11 December 2001), 2167 UNTS 3, arts. 9-13,17.846. See, e.g., Convention on Conciliation and arbitration of the Organiza-

tion for Security and Cooperation in europe, Stockholm, 15 December 1992 (in force, 5 December 1994), 32 ILM 551.847. See, e.g., the regional nuclear weapon free zones treaties : Treaty for the

Prohibition of nuclear weapons in Latin america, Tlatelolco, 14 February 1967(in force, 22 april 1968), 634 UNTS 281 ; South Pacific nuclear Free ZoneTreaty, Rarotonga, 6 august 1985 (in force, 11 December 1986), 24 ILM 1440 ;Treaty on the Southeast asia nuclear weapon-Free Zone, 15 December 1995 (in force, 27 March 1997), 35 ILM 635 ; african nuclear Free Zone Treaty,Cairo, 11 april 1996 (in force, 15 July 2009), 35 ILM 698. note the qualified

in the nuclear field) and of course economic development and freetrade 848.425. The proliferation of regional approaches to human rights 849

might seem particularly significant given that human rights areexpressly based on putative universal values — as the universalDeclaration of Human Rights emphasized with the very first word ofits title 850. Regional human rights regimes enable individuals — tovarying degrees — to seek redress for breach of more or less well-defined human rights by the State. But despite common aims and asimilar legal and philosophical genealogy, the specific content ofhuman rights, and of the mechanisms to enforce those rights, isnuanced between regional regimes — to the point of significant variation 851.426. For example, the african Charter on Human and Peoples’

Rights emphasizes the indivisibility of human rights by substantivelyrecognizing economic, social and cultural rights as well as protectingcivil and political rights 852. It recognizes both the rights of each person and the collective rights of all peoples. Collective rights arerecognized to a far greater extent than in other human rights instru-ments 853. In contrast, the european Convention on Human Rightsonly recognizes civil and political rights of individuals (or, strangely,

244 J. Crawford

use made of these treaties by the International Court in the Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion, ICJ Reports 1996, pp. 248-250 (paras. 58-59) (the existence of these treaties gives rise to an a contrario argument with respect to deployment and testing of weapons else-where).848. There is a proliferation of regional economic development and free trade

areas : e.g., the european Communities, eFTa, CeFTa, naFTa, MeRCOSuR,COMeSa.849. Through the parallel development of regional human rights protection

systems in europe (1950), the americas (1969) and africa (1981).850. Ga res. 217a (III), 10 December 1948.851. See, e.g., american Convention on Human Rights, San José,

21 november 1969 (in force, 18 July 1978), 114 UNTS 213 (aCHR) ; africanCharter of Human and Peoples’ Rights, nairobi, 27 June 1981 (in force, 21 October 1986), 1520 UNTS 217 (aCHPR). See also M. evans and R. Murray(eds.), The African Charter on Human and Peoples’ Rights : The System inPractice, 1986-2000 (Cambridge, CuP, 2nd ed., 2008). 852. aCHPR, op. cit., Preamble (para. 7), Chap. I. Cf. aCHR, op. cit.,

Preamble (para. 4), Chaps. II-III. 853. aCHPR, op. cit., arts. 19-24. See also J. Crawford, “The Rights of

Peoples : ‘Peoples’ or ‘Governments’ ?”, in J. Crawford (ed.), The Rights ofPeoples (Clarendon Press, Oxford, 1988), pp. 55-68 ; J. Crawford, “The Right toSelf-Determination in International Law : Its Development and Future”, in P. alston (ed.), Peoples’ Rights (Collected Courses of the Academy of EuropeanLaw) (Oxford, OuP, 2001), pp. 7-67.

corporations) 854. These differences reflect negotiating priorities linkedto place and time. The european regime, though born from the ashesof war, was established in a region that was more or less accustomedto the language of individual rights as a currency of political debate.The later african regime was established in a region still yearningfor collective and individual emancipation after decolonization. 427. The enforcement machinery for violations of “universal”

human rights also differs between regional regimes. For example,the Inter-american Court of Human Rights has developed an inno-vative judicial penalty of non-pecuniary collective reparations — aremedy which seeks to preserve the name of the victim and to offersome form of recognition to the wider community affected 855.Reparations have included public acts of apology, the broadcastingof judgments on radio and the renaming of streets, all of which seekto educate current and future generations — especially where illiter-acy or oral history traditions prevail — about human rights viola-tions and to deter recurrence 856. In contrast, the european Court ofHuman Rights dispenses penalties that conform to judicial ortho-doxy, including declarations of violation and pecuniary penalties 857.no doubt regional human rights regimes pursue similar “universal”meta-rights but these differ materially in their definition, practicalapplication and implementation.

2. The legal significance of regionalism

428. For its part, the International Court has been reluctant toattribute any legal significance to regional considerations. Since its


854. Convention for the Protection of Human Rights and FundamentalFreedoms, Rome, 4 november 1950 (in force, 3 September 1953), 213 UNTS221 (eCHR). 855. aCHR, op. cit., art. 63 (1) ; Plan de Sánchez Massacre v. Guatemala,

Reparations and Costs, Judgment, Series C, No. 116 (19 november 2004), p. 97(para. 110).856. See, e.g., Yatama v. Nicaragua, Preliminary Objections, Merits, Repa-

rations and Costs, Judgment, Series C, No. 127 (23 June 2005), p. 99 (paras. 252-253) ; Myrna Mack-Chang v. Guatemala, Merits, Reparations andCosts, Judgment, Series C, No. 101 (25 november 2003), p. 130 (para. 286) ; D. Shelton, Remedies in International Human Rigths Law (Oxford, OuP, 2nd ed., 2004), pp. 216-218, 266-276, 285-289, 299-301, 361-364. 857. eCHR, op. cit., art. 41 ; european Court of Human Rights, Rules of

Court (1 September 2012), Rule 75 — Ruling on Just Satisfaction. Cases of DeWilde, Ooms and Versyp (“Vagrancy”) v. Belgium (article 50), Judgment, app.nos. 2832/66, 2835/66 and 2899/66 (10 March 1972), para. 21 ; Shelton, op. cit., pp. 189-192, 194-200, 259-266, 280-285, 294-298.

function, especially in contentious cases, is to pronounce on par-ticular questions of international law arising in disputes betweenStates, it might be expected to be sympathetic to considerationsbased on the regional interests or affiliations of those States. In factit tends to see cases raising regional considerations through theprism of general international law.429. In Asylum (Colombia/Peru), Colombia expressly relied on a

“regional” rule of customary international law : it argued that inaccordance with such a rule it was competent to qualify the offencecommitted by an asylum-seeker, by unilateral and definitive deci-sion, for the purpose of granting him diplomatic asylum 858. TheCourt treated the existence of this “alleged regional or local custompeculiar to Latin-american States” 859 as in effect a bilateral ques-tion :

“The Party which relies on a custom of this kind must provethat this custom is established in such a manner that it hasbecome binding on the other Party. . . . [e]ven if it could besupposed that such a custom existed between certain Latin-american States only, it could not be invoked against Peruwhich, far from having by its attitude adhered to it, has on thecontrary repudiated it by refraining from ratifying theMontevideo Conventions of 1933 and 1939, which were thefirst to include a rule concerning the qualification of theoffence in matters of diplomatic asylum.” 860

246 J. Crawford

858. Asylum (Colombia/Peru), Judgment, ICJ Reports 1950, pp. 271, 274.859. Ibid., p. 276.860. Ibid., pp. 276-278 (emphasis added). The Convention on Political

asylum, Montevideo, 26 December 1933 (in force, 28 March 1935), 152 BFSP231 was adopted with the express purpose of defining certain terms of the earlierConvention fixing the Rules to be observed for the Granting of asylum, Havana,20 February 1928 (in force, 21 May 1929), 132 LNTS 323. The text of theMontevideo Convention formulated general law-making propositions by usingmandatory, express and specific language :

“It shall not be lawful for the States to grant asylum in legations, war-ships, military camps, or airships to those accused of common offenses whomay have been duly prosecuted or who may have been sentenced by ordi-nary courts of justice, nor to deserters of land or sea forces.” (art. 1.)

This provision replaced art. 1 of the Havana Convention, which declared it “notpermissible” for States to engage in such activities. Cf. the softened position inthe Treaty on asylum and Political Refuge, Montevideo, 4 august 1939 (inforce, 29 September 1954), Oea/Ser.X/1, Treaty Series 34, trans. (1943) 37AJIL Sup. 99-103 :

430. It seems clear that the Court, despite its invocation of article 38 (1) (b) of its Statute 861, was applying a stricter standard ofproof than it would have done to a “universal” rule of general inter-national law. In the case of such general rules it is not necessary thatevery State have specifically accepted or adhered to the rule : persis-tent opposition may render the rule inapplicable 862 but must be consistent and clear, and is not manifested by a simple failure to ratify a treaty. In this case systematic elements of general interna-tional law — sovereignty, non-intervention, the regular enforcementof domestic law even against political offenders — overwhelmedconsiderations of “regional” custom or practice. Since the body ofLatin american “regional” custom is by far the most sophisticatedand generalized, it seems safe to infer that the same applies every-where in the world. This is not to imply that regional or local customcan never be relied on, just that it must be proved as between theparticular States parties to the dispute ; it makes no differencewhether the “region” in which the custom exists comprises two ortwenty-two States.431. This point is well illustrated in the Right of Passage case,

where Portugal argued that local custom provided it with unencum-bered passage over a section of Indian territory for civilian and military purposes. The International Court rejected India’s assertionthat a local custom could only be established between three ormore States 863. Having held that a limited local custom (passage for private persons, civil officers and goods) was established throughlong bilateral practice — some 125 years — the Court considered itunnecessary to determine whether a similar result would follow from


“asylum may be granted only in embassies, legations, men-of-war, mili-tary camps or military airplanes, and exclusively to persons pursued forpolitical reasons or offenses, or under circumstances involving concurrentpolitical offenses, which do not legally permit of extradition” :

art. 2.

“asylum shall not be granted to persons accused of political offenses,who shall have been indicted or condemned previously for commonoffenses, by the ordinary tribunals” : art. 3.

861. Asylum (Colombia/Peru), op. cit., pp. 276-277.862. as in the Fisheries (United Kingdom v. Norway), Judgment, ICJ Reports

1951, pp. 131, 138-139. See J. I. Charney, “The Persistent Objector Rule and theDevelopment of Customary International Law” (1985) 56 BYIL. See Chap. II foranalysis on the formation of custom. 863. Right of Passage over Indian Territory (Portugal v. India), Judgment,

ICJ Reports 1960, p. 39.

any rules of general international law 864. The reason was that theparticular practice was intended to be “governing” of the relationsbetween the parties and would thus prevail over any general rule 865. 432. The lessons of Asylum were not lost on norway in Fisheries

(United Kingdom v. Norway) 866. There was no regional practicenorway could rely on to support its particularistic approach to base-lines for maritime delimitation. The gist of its position was that itsapproach was justified under general international law having regardto the responses of States and the requirements of reasonableness. Itwas on this basis that the Court upheld norway’s view, and did notconsider essential the united Kingdom’s assent, which it would haverequired to find that there existed a local custom. The Court held thatall it “can see [in norway’s position] . . . is the application of generalinternational law to the specific case” 867, and it supported norway’sposition that the base-line methodology used was “an adaptation ren-dered necessary by local conditions” 868. The Court held that “certaineconomic interests peculiar to a region, the reality and importance ofwhich are clearly evidenced by a long usage” 869 were relevant to itsapplication of the law to the facts, but not to establish historic title inderogation from international law.433. The International Court has also held in other cases that a

standard first adopted in a particular region has been incorporatedinto general international law, as for example the uti possidetis prin-ciple was endorsed in Frontier Dispute. Uti possidetis can broadlybe defined as maintaining the territorial status quo of a territorialentity after its independence 870. as the Court said, “[T]he obligationto respect pre-existing international frontiers in the event of a Statesuccession” 871, applied as a rule requiring the respect by newlydecolonized african States of “administrative boundaries and fron-tiers established by the colonial powers” 872. The Court held that :

248 J. Crawford

864. ICJ Reports 1960, pp. 40-44.865. Ibid., p. 44.866. Fisheries (United Kingdom v. Norway), Judgment, ICJ Reports 1951,

p. 116.867. Ibid., p. 131.868. Ibid., p. 133.869. Ibid.870. Frontier Dispute (Burkina Faso/Mali), Judgment, ICJ Reports 1986,

p. 565 (para. 22).871. Ibid., p. 566 (para. 24).872. Ibid., p. 565 (para. 21).

“[T]he principle of uti possidetis seems to have been firstinvoked and applied in the Spanish america, inasmuch as thiswas the continent which first witnessed the phenomenon ofdecolonization involving the formation of a number ofsovereign States on territory formerly belonging to a singlemetropolitan State. nevertheless the principle is not a specialrule which pertains solely to one specific system of interna-tional law. It is a general principle, which is logically con-nected with the phenomenon of the obtaining of independence,wherever it occurs . . .” 873

and that the obligation :

“[D]erives from a general rule of international law, whetheror not the rule is expressed in the formula uti possidetis. Hencethe numerous solemn affirmations of the intangibility of thefrontiers existing at the time of the independence of africanStates . . . are evidently declaratory rather than constitutive :they recognize and confirm an existing principle, and do notseek to consecrate a new principle or the extension to africa ofa rule previously applied only in another continent.” 874

434. In other words : “regional” claims can be accepted only ifthey can be accommodated within the ordinary framework of inter-national law ; there is no third, non-universal category in betweengeneral international law and the legal relations of specific Statesbased on consent, acquiescence and recognition. This is not to denythe fact of regional hegemony at different times and places. But tothe extent that such hegemony is not embodied in treaties or other-wise clearly agreed to by the affected States, it will be extralegal incharacter ; it does not detract from the tendency of the internationalsystem towards universal or general principles.435. Furthermore, even where regional hegemony is embodied in

treaties, universal standards relating to the use of force will prevail, onthe basis of Chapter vIII of the united nations Charter (under whichtreaty-based enforcement action requires Security Council approval)875.


873. ICJ Reports 1986, p. 565 (para. 20) (emphasis added).874. Ibid., p. 566 (para. 24). See generally G. nesi, “uti possidetis Doctrine”,

in R. wolfrum (ed.), The Max Planck Encyclopedia of Public International Law(OuP, 2008, online). 875. See B. Simma (ed.), The Charter of the United Nations : A Commentary

(Oxford, OuP, 1994), pp. 679-757.

D. The Future of Universality

436. an early advisory Opinion by the International Court,Reservations to the Genocide Convention 876, illustrates the com-plexity of the relationship between universality and regionalism. Onthe one hand, as in the cases just discussed, the Court did not hesi-tate to assert that alleged regional practice in fact amounted to a ruleof general international law. It relied on a Latin american practice ofpermitting reservations to multilateral treaties, despite the contrarypractice of the League of nations 877. as the dissentients predictedwould happen, the new approach was subsequently adopted morewidely in State practice and codified in the vienna Convention onthe Law of Treaties 878.437. On the other hand, the Court inferred from “the clearly

universal character of the united nations under whose auspices the[Genocide] Convention was concluded” and the near-universal participation in that treaty that certain variations were permissible asbetween some of its States parties 879. This relationship between uni-versality and the capacity for variation might seem paradoxical ; aspurious sort of unity in diversity. But it might also be seen as thenecessary product of an attempt to conceive of and organize a globalsociety of States in the persistent absence of any central authority.More recently, even this apparent qualification to the universality ofinternational law has been brought into question, at least in the context of “universal” human rights, by claims that certain imper-missible reservations — such as those that violate basic guaranteesdesigned to support the attainment of protected rights — can be severed from human rights treaties without depriving those treatiesof effect 880. It appears then, that reservations in human rights law

250 J. Crawford

876. Reservations to the Convention on the Prevention and Punishment of theCrime of Genocide, Advisory Opinion, ICJ Reports 1951.877. Ibid., p. 25 (contrasting the practice of the League and the Organization

of american States). Cf. Reservations to the Convention on the Prevention andPunishment of the Crime of Genocide, Advisory Opinion, ICJ Reports 1951, pp. 34-37 (Judges Guerrero, Mcnair, Read and Hsu Mo, diss. ; Judge alvarez,diss.) who prefer the League practice and explain the Latin american practice asbased on “the prior agreement of the contracting parties” of the Pan-americanunion.878. vienna, 23 May 1969 (in force, 27 January 1980), 1155 UNTS 331

(vCLT), arts. 19-23.879. Reservations, Advisory Opinion, op. cit., p. 20.880. Belilos v. Switzerland, [1988] eCtHR no. 10328/83 91988. Cf. Kennedy

v. Trinidad & Tobago (Admissibility Decision of 2 November 1999), Human

have a different character to reservations in general internationallaw, where objections, among other things, can render purportedreservations inoperable against the objecting State 881. 438. One advantage of the International Court in promoting uni-

versality is that in its contentious jurisdiction it is required to workfrom the particular to the general. It approaches claims betweenneighbouring States against the background of such “universal” prin-ciples as sovereignty, territoriality and consent. at least in casescommenced by special agreement, it also has the advantage that thedecision to refer the matter to it usually entails that the outcome is tobe determinative.439. These advantages are not shared by other institutions that

shape international law, such as the ILC in its “pre-legislative” acti-vity. nonetheless, the ILC’s texts are likely to be acceptable to Statesonly if they can be presented in as general terms as possible. Thiscan be seen from its past successes and failures. The viennaConvention on Succession of States in Respect of State Property,archives and Debts of 1983, for example, has been ratified by onlya handful of States 882. This is partly the result of its focus on the category of “newly independent states” — former colonies, a groupthat is now largely a matter of history — and the controversial diver-gences between its treatment of States within that category and othercategories of successor States, which are not generally accepted ascustom 883. More successful was the ILC’s codification of treatyrules, which culminated in the vienna Convention on the Law of


Rights Committee Comm. no. 845/1999. See Human Rights Committee,General Comment No. 24 : Reservations to the International Covenant on Civiland Political Rights, un doc. CCPR/C/21/Rev.1/add.6, 4 november 1994, para. 11. For criticism see ILC Special Rapporteur (a. Pellet), Second Report onReservations to Treaties, un doc. a/Cn.4/477 & add.1, 10 May and 13 June1996, pp. 71-87. See generally R. Goodman, “Human Rights Treaties, InvalidReservations and State Consent” (2002) 96 AJIL 531-560.881. vCLT, op. cit., art. 21 (3). For analysis of current perspectives see ILC,

Report of the International Law Commission Sixty-third Session (26 april-3 June and 4 July-12 august 2011), Guide to Practice on Reservations toTreaties with Commentaries, un doc. a/66/10/add.1. 882. vienna, 8 april 1983, un doc. a/COnF.117/14. Only seven States have

ratified the treaty : Croatia, estonia, Georgia, Liberia, Slovenia, Macedonia andukraine.883. See, e.g., I. Sinclair, The International Law Commission (Cambridge,

Grotius, 1987), p. 79 ; M. Shaw, International Law (Cambridge, CuP, 6th ed.,2008), p. 986 ; J. Crawford, “The Contribution of Professor D. P. O’Connell tothe Discipline of International Law” (1980) 50 BYIL ; cf. J. Crawford, “StateSuccession and Relations with Federal States : Remarks” (1992) 86 ASILProceedings 15.

Treaties. although it has a reasonably high level of membership,with 113 States parties, particular aspects of the Convention arewidely held to represent custom, even by non-parties, and have beenapplied by them and the International Court as such. Similarly, oneof the best-known moves made by the ILC, masterminded by specialrapporteur Robert ago, was the elevation of its work on Stateresponsibility to a higher level of generality — from “primary” to“secondary” rules 884. This contributed both to the successful conclu-sion of the project and to the widespread acceptance of the result, thearticles on the Responsibility of States for Internationally wrongfulacts of 2001 885. 440. So we are left with a sense that the key institutions that

shape international law, including the ILC, other organs of theunited nations and the International Court, continue to nudge inter-national law in the direction of universality. However, it is becauseof the process of international law that it offers us not “infinite variety” 886, but flexibility, even if within established parameters, toemploy regional instruments and mechanisms where necessary butin due course to express or accommodate these in general terms.These are steps in what seems to be an enduring movement towardsuniversality.

252 J. Crawford

884. See ILC Ybk 1973/II, pp. 169-170 ; ILC Ybk 1974/I, p. 5.885. ILC Ybk 2001/II (2), p. 26. See further J. Crawford, a. Pellet and

S. Olleson (eds.), The Law of International Responsibility (Oxford, OuP, 2010) ;J. Crawford, State Responsibility : The General Part (Cambridge, CuP, 2013).886. R. R. Baxter, “International Law in ‘Her Infinite variety’ ” (1980) 29

ICLQ 549.

international law as a system “international law as an open system”, in j. crawford,...

Documents