35.1  The most frequently cited definition of a treaty is that contained in the Vienna Convention on the Law of Treaties.² It defines the term ‘treaty’ as ‘an international agreement concluded between States in written form and governed by international law, whether embodied in a single instrument or in two or more related instruments and whatever its particular designation’.³ This definition is however expressly for the purposes of the Convention only. It consciously excludes, for example, international agreements concluded between States and international organizations or between international organizations themselves, and equally excludes international agreements not in written form, while providing at the same time that the omission of these cases is without prejudice to their legal force or to the application to them of the rules of international law.⁴ Amongst the leading authorities, McNair uses the term ‘treaty’ to denote ‘a written agreement by which two or more States or international organizations create or intend to create a relation between themselves operating within the sphere of international law’;⁵ while Reuter, at a higher level of generality, speaks of: ‘une manifestation de volontés concordantes imputables à deux ou à plusieurs sujets de droit international et destinée à produire des effets de droit selon les règles de droit international’.⁶

35.2  It is thus apparent that there is no single, all-embracing definition of the term ‘treaty’, and that the Vienna Convention, for all its great authority in the field, does not set out to provide one. The Convention is however now supplemented by a second Vienna Convention, of 1986, equally adopted on the basis of thorough and lengthy preparation by the International Law Commission, and (p. 535) designed to regulate the treaties of international organizations, both those concluded between themselves and by them with States.⁷ With the sole exception of the references to international organizations, the terms used in the 1986 Convention to define ‘treaty’ for its own purposes are substantially identical to those used in the 1969 Convention. It is thus possible to extract the following elements as central to a modern general definition of the term ‘treaty’: an agreement, of a suitably formal character, designed to give rise to legal rights and obligations, operating within the sphere of international law, and concluded between two or more parties possessing legal personality under international law. This should be read together with the fact that, whereas States are automatically endowed with treaty-making capacity, as an aspect of their sovereignty,⁸ the extent to which a particular international organization will be competent to conclude treaties will be a function of its constituent instruments and other rules of the organization, based on the fulfilment of the organization’s functions and purposes.⁹

35.4  It should finally be noted that, so long as it evidences a clear indication of consent by the parties to be bound, it is not necessary for a treaty, even one in writing, to take the form of an instrument signed, and as the case may be ratified, by each of them. The Algiers Accords, which settled a series of questions in dispute between the United States and Iran following the revolution that deposed the Shah of Iran, takes the form of declarations made by the government of Algeria in which it records, as intermediary, ‘the … interdependent commitments … made by the two governments’, and to which were appended the terms of an escrow agreement between the United States government and certain banks.¹⁴ In the case between Qatar and Bahrain on Maritime Delimitation and Territorial Questions,¹⁵ the International Court of Justice found that separate exchanges of letters between the King of Saudi Arabia and the Amirs of Qatar and Bahrain, together with a document headed ‘Minutes’ and signed by the foreign ministers of the three States, were international agreements creating rights and obligations for the parties sufficient to establish the jurisdiction of the Court.

35.5  For practical purposes, therefore, neither agreements not in written form, nor treaties concluded by international organizations, nor treaties cast in a non-conventional shape, need be regarded as a special case when it comes to the application to them of the law of treaties. But several of the central elements set out above warrant further analysis.

35.6  First there is the requirement that the agreement should be ‘governed by international law’.¹⁶ This serves to distinguish a ‘treaty’ from other agreements which, although concluded between States or other subjects of international law, are (p. 537) regulated not by international law but by the national law of one of the parties (or by some other legal system chosen by the parties). An example would be a state contract concluded between the competent bodies of the Ruritanian and Utopian governments, whereby the latter agreed to supply the energy needs of the former through a gas pipeline on the basis of a standard form of contract used in the energy trade. Such a contract would not be a treaty and would not be governed by international law but by the terms of the contract itself supplemented, where necessary, by general principles of law. Other transactions of a private law nature, such as leases of land and buildings and loan agreements, may also be entered into between States; in such cases, it may be difficult to determine whether the parties intended the transaction to be governed by international law, or by general principles of law, or by a particular system of national law.¹⁷ Conversely, it is perfectly possible for transactions between non-state parties, or between a State and a foreign private party, to be governed, by agreement, by international law, either on its own or as a supplement to the application of national law; this is very frequently the case for investment contracts, notably those concluded under the cover of bilateral treaties for the protection and promotion of investments.¹⁸ The Washington Convention of 1965 on the Settlement of Investment Disputes between States and Nationals of Other States, concluded under the auspices of the World Bank, established a Centre, based in Washington, for the settlement of such disputes by conciliation or by arbitration, and provides (in the case of arbitration) that, in the absence of agreement, the tribunal shall apply national law ‘and such rules of international law as may be applicable’.¹⁹

35.7  The requirement also serves moreover to distinguish a further category of international transactions which, although cast in writing and representing the result of an agreement arrived at between one or more governments, are nevertheless not intended to give rise to legal relations, ie to give rise to a set of binding rights and correlative obligations. Lacking this essential element, these instruments cannot properly be classed as treaties.

35.8  Not all instruments of an international character, however important they may be, are intended to be binding in law. For instance, no legal obligations were incurred through the Atlantic Charter of 14 August 1941, in which Franklin Roosevelt (for the United States) and Winston Churchill (for the United Kingdom) declared formally that they deemed it right ‘to make known certain (p. 538) common principles in the national policies of their respective countries on which they base their hopes for a better future for the world’. Similarly, it is plain on the internal evidence that the Final Act of the Conference on Security and Cooperation in Europe²⁰ was not considered to amount to a treaty. The government of Finland is requested to arrange for the circulation of the Final Act as an official document of the United Nations, but it is specifically stated that the Final Act ‘is not eligible for registration under Article 102 of the Charter of the United Nations’, a further indicator of the fact that the States signing the Final Act did not consider that they were subscribing to a treaty instrument.²¹ Another example is the Universal Declaration of Human Rights adopted by the General Assembly of the United Nations on 10 December 1948,²² which was followed by lengthy discussion to turn its provisions into legally binding form in the two International Covenants on human rights.²³

35.9  Modern practice is replete with cases in which the States drawing up an international instrument incorporating points of agreement between them make the deliberate choice not to clothe it in legally binding form.²⁴ Such instruments are very often entitled ‘Memorandum of Understanding’ (MOU), or are referred to as such, as a specific indicator of their non-binding character. The FCO Guidance summarizes the essential character of an MOU as follows: ‘An MOU records international “commitments” but in a form and with wording which expresses an intention that it is not to be legally binding’. Whether the practice of recording commitments in MOUs is desirable or undesirable is a matter of debate, but various reasons may enter into the choice of the non-binding form over the treaty form. One might be the wish to avoid the publicity attendant on registration with the United Nations²⁵ or inclusion in a national treaty series, if for example the agreement deals with sensitive or secret matters, like nuclear technology, weapons production, intelligence cooperation, or joint action against terrorism. Another might be the wish to avoid having to submit the agreement to a process of formal internal approval before the national parliament. Yet another might be (p. 539) the simple preference for a commitment depending for its fulfilment more on political harmony and good faith than on the possibility of bringing formal claims in the event of a future breach.²⁶ Valid as these reasons may be—and indeed are—in particular cases, they can as easily be called into play to shore up less noble motives in other cases.

35.10  British practice in the drawing up of MOUs follows, as far as possible, a set terminology in order to distinguish MOUs from treaties. Standard terminology for use in MOUs can be found in the FCO Guidance: for instance, use ‘governments’ or ‘participants’, not ‘parties’, ‘provisions’ not ‘terms’ and ‘will’ not ‘shall’. MOUs are expressed to ‘come into operation’ or ‘effect’, while treaties ‘come into force’. An MOU, and any exchange of Notes confirming it, should record the ‘understanding’ of the governments concerned in the matter, not their ‘agreement’. If this terminology is followed, it will generally be plain that the intention is not to create legally binding rights and obligations.²⁷

35.11  Many other States are similarly consistent in the use of particular terminology. Confusion has however been caused by the fact that the practice of the United States, in particular, has been less consistent, with the result that the US has regarded some instruments as treaties which other States would regard as MOUs.²⁸ It is of course possible to provide expressly to the effect that the instrument is not legally binding,²⁹ but this is not commonly done, and the absence of such a provision would not on its own be conclusive.

35.12  The fact that an MOU is not legally binding does not however mean that entering into one entails no commitment; as indicated above, the commitment is one of a policy rather than a legal kind, and issues of good faith and reputation are just as much at stake in the one case as in the other. The MOU form can thus in practice be used, not simply for mundane exchanges on technical subjects, but (p. 540) as the vehicle for commitments of considerable importance, for example projects for the joint development or procurement of sophisticated weapons systems involving substantial investment. So MOUs have also been concluded in recent years with a number of countries in the Middle East and North Africa intended to facilitate the deportation or extradition of terrorist suspects by providing assurances about their treatment on return to their home country sufficient to meet the United Kingdom’s human rights obligations. Whether the vehicle is in fact sturdy enough to carry this weight of freight has been cast into doubt in national courts. A national court, faced with a question (arising for example out of legislation on human rights) as to whether another State can be relied upon to uphold its commitments, might well take an adverse view of the fact that the State was not willing to give those commitments legal form. How far this kind of assurance may be relied upon, given conditions in the receiving countries, has been the constant subject of challenge in British courts, with varying results, and the possibility of further recourse to the European Court of Human Rights.³⁰

35.13  Nor does the absence of legal obligation mean that an MOU is devoid of legal consequences. The principle that States must conduct their dealings with each other in good faith may have the effect that a State is estopped from going back on the terms of an MOU where the other State has acted in reliance upon them, though this is still a developing area of international law.

35.14  ‘Terminology’ refers here to the title given by the parties to their treaty, and ‘form’ to its formal structure. Although the question of terminology was once the source of much confusion, it is now well established that the title given to a treaty instrument is entirely a matter of the free choice of the parties, and has no bearing on the legal quality of the instrument or on the application to it of the law of treaties. The Vienna Convention definition cited above³¹ makes this explicit by providing in terms that it applies to all such agreements whatever their particular designation. Conversely, the title given to an instrument, although it may be a strong indicator of the intentions of the parties drawing it up, is not in itself conclusive of the status of the instrument, which would have to fulfil in substance the criteria listed above³² in order to rank as a treaty. Although Article 102 of the United (p. 541) Nations Charter, when laying down the obligation of registration,³³ uses the dual terminology ‘[e]very treaty and every international agreement’, the term ‘treaty’ may be taken as having the broadest generic scope, and is equally apt to cover those written instruments or series of instruments designated as a treaty, convention, agreement, protocol, covenant, charter, statute, act, declaration, concordat, exchange of Notes, exchange of letters, modus vivendi, and others besides. It may also, depending upon the intention of the parties, cover agreed minutes and (subject to paragraphs 35.9–11 above) memoranda of understanding.

35.15  A rationalization of the terminology employed in international practice to describe treaty instruments is difficult to attempt, and hardly seems profitable. So far as trends are discernible at all in modern practice, it might be said that ‘treaty’ and ‘convention’ tend to be invested with the most solemnity, the former being frequently (but by no means exclusively) employed for bilateral engagements and the latter (but again by no means exclusively) for general multilateral engagements; whereas ‘statute’ is most commonly reserved for the constitutive instruments of international organizations. But any statement of this kind is subject to so many exceptions as to let it serve as no more than the most general guidance, and then of a purely descriptive character. It would nevertheless appear that the most frequently utilized designations for treaty instruments are treaty, convention, agreement, exchange of Notes, protocol, and declaration.

35.16  In this chapter, it is proposed to deal with these six types of treaty instrument, and in Chapters 36 and 37 with various other types of treaty instrument which would normally fall within the definition of a treaty: eg Pact, Constitution, Charter, Statute, Regulation, Concordat, Additional Articles, Act, General Act, Final Act, Modus Vivendi, and Compromis (Special Agreement).

35.17  Of those international instruments which are clearly intended to have an obligatory or binding character, many of the most significant bear the title ‘treaty’. The word ‘treaty’ is derived from the French word traiter, which means to negotiate. The comment to Article 1 of the Harvard Draft Convention on the Law of Treaties states that ‘in the literature of diplomacy and international law the term “treaty” is employed in both a general and a restricted sense’. In the general sense, it ‘embraces a great variety of instruments to many of which other names than “treaty” are given, although there is seldom if ever any juridical distinction between them’. (p. 542) The definition of the term ‘treaty’ in the Vienna Convention is of course based on this general sense. But the designation ‘treaty’ in the restricted sense refers to the formal instrument of agreement by which two or more States establish or seek to establish a relation under international law between themselves; and it is this restricted sense of the term ‘treaty’ which must now be considered.

35.18  Generally speaking the designation ‘treaty’ has been reserved for international agreements of a particularly significant character, for example treaties of peace or alliance. The designation ‘treaty’ may also be employed to mark the political importance of the particular international agreement—as for example the North Atlantic Treaty of 4 April 1949,³⁴ and all of the treaties establishing, first, the European Communities and then the European Union,³⁵ as well as the treaties providing for the accession of new Member States to them.³⁶ It was once the case that treaties in this restricted sense were always concluded between heads of state; but modern practice demonstrates that treaties can be concluded in heads of state form, in interstate form, or in intergovernmental form. The choice of form may depend in part on the political importance of the treaty and in part on the constitutional requirements of the Contracting Parties; otherwise it is not possible to discern any general principle governing the choice between these forms, although the following illustrations of practice may be given.

35.19  Treaties between heads of state are still concluded. The European Community and European Union Treaties are cast in this form, as are the accession arrangements for new Member States.³⁷ The form has virtually died out, however, for bilateral treaties; the last example in British practice appears to be the Treaty between the United Kingdom of Great Britain and Northern Ireland and the French Republic concerning the Construction and Operation by Private Concessionaires of a Channel Fixed Link, with an Exchange of Letters, done at Canterbury on 12 February 1986.³⁸

35.20  A treaty concluded in heads of state form can be broken down into the following parts:

1. 1. The preamble, beginning with (a) the names and titles of the High Contracting Parties; (b) a summary of the object and purpose of the treaty; (c) the names (p. 543) and official designations of the Plenipotentiaries appointed by the High Contracting Parties;³⁹ (d) a paragraph stating that the Plenipotentiaries have produced their Full Powers, which were found to be in good and due form, and that they have agreed upon the following articles.

2. 2. The various substantive articles of the treaty, which may begin with a definitions article and will then normally be followed by general provisions, more specific provisions and, finally, any provisions laying down the means of executing the general and specific provisions.⁴⁰

3. 3. A series of articles generally known as the ‘final clauses’ which deal with such matters as the territorial application of the treaty, signature, ratification, accession, entry into force, amendment, denunciation (withdrawal), and duration. In the case of a multilateral treaty, there may also be a reservations article. The provision for ratification usually mentions the place for the exchange or deposit of instruments of ratification, whilst the clause dealing with duration may specify that the treaty shall remain in force indefinitely, or for a fixed number of years, or for successive fixed periods unless notice of termination is given by a specified time.

4. 4. A clause (testimonium)⁴¹ stating ‘In witness whereof … (‘En foi de quoi …’) the respective plenipotentiaries have signed this Treaty’.⁴²

5. 5. Locality and date (‘Done at … the … day of … , …’).

6. 6. Seals and signature.

35.21  Treaties between States (ie in interstate form) are now much more frequent. The North Atlantic Treaty of 4 April 1949, the Treaty of Peace with Italy of 10 February 1947,⁴³ and the Treaty of Peace with Japan of 8 September 1951⁴⁴ were all concluded between States rather than between heads of state. The Treaty on the Non-Proliferation of Nuclear Weapons of 1 July 1968 is another example of an important multilateral treaty concluded in interstate form. Recent bilateral treaties concluded by the United Kingdom in interstate form include the Treaties on Extradition, and on Judicial Assistance in Civil and Commercial Matters, concluded with the United Arab Emirates on 6 and 7 December 2006 respectively.⁴⁵

(p. 544) 35.22  Treaties between States are rather less formal in their terminology than treaties between heads of state.⁴⁶ In place of the term ‘High Contracting Parties’, the expression ‘Contracting Parties’ or ‘Contracting States’ (or sometimes simply ‘the Parties’ or ‘the States Parties’)⁴⁷ is used. The preamble may also be reduced in length by the omission of one or other of the formal elements usual in the preamble of a treaty concluded between heads of state. It is, however, customary, even in the case of a treaty between States, to retain in the preamble a statement of the object and purpose of the treaty.

35.23  Treaties between governments are also less formal. The parties normally refer to themselves as the ‘Contracting Governments’ or the ‘Contracting Parties’, although occasionally one may come across the use of such terms as ‘Signatory Governments’ or ‘Participating Governments’. Again, however, the preamble will usually contain a statement of the object and purpose of the treaty.

35.24  The designation ‘convention’ tends to be utilized for multilateral treaties of a law-making type. Illustrative of this tendency are the various Hague Conventions emanating from the First and Second Hague Peace Conferences (of 1899 and 1907 respectively) and the various conventions concluded since the Second World War as a result of the process of progressive development and codification of international law being pursued within the framework of Article 13 of the United Nations Charter. Examples are the Vienna Convention on Diplomatic Relations of 18 April 1961,⁴⁸ the Vienna Convention on the Law of Treaties of 23 May 1969,⁴⁹ the Convention on the Prevention and Punishment of Crimes against Internationally Protected Persons, including Diplomatic Agents, of 14 December 1973,⁵⁰ the United Nations Convention on the Law of the Sea of 10 December 1982,⁵¹ and the Convention against Torture and Other Cruel, (p. 545) Inhuman or Degrading Treatment or Punishment, of 10 December 1984.⁵² A similar practice had been followed by the League of Nations, as, for example, with the Barcelona Conventions of 1921 and the Slavery Convention of 1926. Law-making or regulatory treaties concluded under the auspices of one or other of the Specialized Agencies of the United Nations will also normally take the form of conventions: examples are the Hague Convention for the Suppression of Unlawful Seizure of Aircraft of 16 December 1970,⁵³ and the Montreal Convention for the Suppression of Unlawful Acts against the Safety of Civil Aviation of 23 September 1971⁵⁴ (both negotiated within the framework of the International Civil Aviation Organization (ICAO)). Treaties which are constitutive of international bodies of a technical character may also take the form of conventions—for example, the Universal Postal Convention, the International Telecommunications Convention, and the various conventions for the protection of intellectual property.

35.25  The Council of Europe may be rare in maintaining a carefully ordered practice in this regard, namely:

Once a “Convention” or an “Agreement” is concluded, they have the same legal effect. The only distinguishing feature is the form in which a State may express its wish to be bound by one or the other. While a European Convention is usually the object of the deposit of an instrument of ratification, acceptance or approval, a European Agreement may be signed with or without reservation as to ratification, acceptance or approval. Moreover, the model final clauses adopted by the Committee of Ministers in 1963 and revised in 1980 contain clauses regarding the procedure whereby non-member States of the Council of Europe may become Parties to European Conventions or Agreements; in principle, these clauses enable them only to accede.⁵⁵

35.26  However, it must not be thought that the designation ‘convention’ is used only for multilateral treaties of a law-making or normative type. The designation is also used for a wide range of bilateral treaties, in the fields, for example, of extradition, double taxation, and judicial cooperation.

35.27  The term ‘agreement’, like the term ‘treaty’ itself, is used in a number of senses. In a generic sense, it covers any meeting of minds—in this case, the minds of two (p. 546) or more international persons. But, as has already been stated, a distinction must always be drawn between agreements intended to have an obligatory character and agreements not intended to have such a character. In its restricted sense, the term ‘agreement’ means an agreement intended to have an obligatory character, but usually of a less formal or significant nature than a treaty or convention. Like treaties and conventions, agreements in this restricted sense may be concluded between heads of state, between States, or between governments.

35.28  The treaty instrument in the form of an ‘agreement’ is, as it were, the workhorse of the treaty-maker. While it can be used for multilateral treaties—for example, the Agreement regarding the Status of Forces of Parties to the North Atlantic Treaty of 19 June 1951⁵⁶—it is more commonly used for bilateral treaties of a fairly routine nature. It may be noted here that bilateral treaties concluded between the United Kingdom and other Commonwealth countries usually take the form of ‘agreements’ and are expressed to be concluded as between governments rather than as between heads of state or as between States.⁵⁷

35.29  Sometimes agreements are concluded between a government department in one country and a government department in another. It depends on the circumstances whether such ‘interdepartmental agreements’ amount to treaties binding under international law or not.⁵⁸

35.30  The term ‘protocol’ derives from the Low-Latin protocollum, itself deriving from a word in the original Greek meaning the ‘first glued-in’ to a book—that is to say, a summary or digest of the contents. In diplomacy, it gradually came to mean the register in which the minutes of a conference were held. It is, of course, also employed in diplomacy to signify the forms to be observed in the official correspondence of the minister for foreign affairs and in the drafting of diplomatic documents such as treaties, full powers, letters of credence and recall, etc. Finally, it has come also to embrace the concept of something which is added to a treaty in order to perfect or complete the treaty. The word ‘protocol’ may accordingly, depending upon the context in which it is used, signify ‘an addition to a treaty, a summary of official proceedings or a technique of the proper method of doing things, including official etiquette’.⁵⁹ In this last very broad sense, the word will, (p. 547) of course, cover such matters as protecting sensitive national dignities on major occasions and the practical arrangement of even informal occasions where there might well be difficulties if no rules existed.⁶⁰

35.31  When used in the sense of a treaty instrument, the designation ‘protocol’ usually refers to a single instrument which amends or supplements an earlier international agreement. The designation ‘protocol’ may also be given to a treaty instrument which is ancillary to a treaty or convention, although concluded simultaneously. Thus there were 30 separate protocols annexed to, but forming an integral part of, the treaties providing for the accession of Denmark, Ireland, Norway, and the United Kingdom to the EEC and EURATOM in 1972, and nine such protocols are annexed to the treaties providing for the accession of the Czech Republic, Estonia, Cyprus, Latvia, Lithuania, Hungary, Malta, Poland, Slovenia, and Slovakia to the EU in 2003.⁶¹

35.32  Compulsory dispute settlement provisions are frequently put into a separate protocol, so that their acceptance can be made optional. Thus there are Optional Protocols on the Compulsory Settlement of Disputes to the Vienna Conventions on Diplomatic Relations and Consular Relations and the Convention on Special Missions.⁶² The right of petition to the European Court of Human Rights was originally optional, and governed by a separate protocol, but has since 1998 been an integral part of the Convention on Human Rights and Fundamental Freedoms, following its amendment by Protocol No 11.⁶³

35.33  The term ‘protocol’ may also be given to a treaty instrument which prolongs the duration of a treaty or convention that is about to expire. Examples are the Protocol for the Continuation in Force of the International Coffee Agreement, 1968, as extended,⁶⁴ and the Protocol of 15 January 1967,⁶⁵ for the extension of the period of validity of the Convention on the Declaration of Death of Missing Persons.

35.34  In its secondary sense of a summary of official proceedings, the term ‘protocol’ has much the same meaning as the minutes of a conference, and has been used as the vehicle to record, on the conclusion of a multilateral treaty or convention, (p. 548) observations, declarations, and ancillary agreements relating to the text or to its interpretation.⁶⁶

35.35  Agreements between governments in regard to some particular matter are sometimes styled ‘protocols’, though differing in no other respect from other agreements. The form has been used to conclude an armistice (Protocol between the United States and Spain of 12 August 1898;⁶⁷ and between Poland and Lithuania of 29 November 1920);⁶⁸ to provide for the delimitation of a boundary (Protocol concerning the delimitation and marking of the Argentine-Uruguayan Boundary Line in the River Uruguay of 16 October 1968);⁶⁹ to re-establish diplomatic relations (Protocol between the Netherlands and Venezuela of 20 August 1894);⁷⁰ to regulate the status of international military headquarters (Protocol of 28 August 1952, between the Parties to the North Atlantic Treaty);⁷¹ and to regulate the exercise of criminal jurisdiction over UN armed forces (Protocol of 26 October 1953, between Japan, the United States, Australia, Canada, and the United Kingdom).⁷² There is a Protocol on Arbitration Clauses, signed at Geneva on 24 September 1923, concerning the recognition and enforcement of certain arbitration agreements and arbitral awards.⁷³ Finally, mention should be made of the important Geneva Protocol of 17 June 1925 for the Prohibition of the Use of Asphyxiating, Poisonous or Other Gases, and of Bacteriological Methods of Warfare.⁷⁴

35.36  It is, however, correct to say that the protocol is now used principally as an instrument subsidiary to a treaty or convention, usually (but not necessarily) drawn up by the same negotiators, and dealing with ancillary or incidental matters such as the interpretation of particular articles of the main treaty or any supplementary provision of a minor character. Ratification of the treaty or convention⁷⁵ will normally ipso facto involve ratification of any supplementary or additional protocol of this nature. In 1993 the English Divisional Court held, with reference to (p. 549) the Maastricht Treaty, that ratification of the Treaty ‘will automatically involve’ ratification of the Protocols to it.⁷⁶ Where the protocol is concluded independently at a later date (as, for example, an amending protocol), it will of course be subject to independent ratification.

35.37  Alternatively, the protocol form may be used as the means for subsequently adding to or expanding, sometimes in a significant way, the terms or scope of an earlier treaty, as in the case of the three Additional Protocols to the Geneva Conventions of 12 August 1949 on the protection of victims of armed conflict (Red Cross Conventions).⁷⁷

35.38  An interesting special case is the Convention on Prohibitions or Restrictions on the Use of Certain Conventional Weapons which may be Deemed to be Excessively Injurious or to have Indiscriminate Effects of 10 October 1980,⁷⁸ which is drawn up as a framework convention for the separate regulation of particular weapons through individual protocols. Three such protocols are annexed to the Convention itself, and two have been adopted since, one on blinding laser weapons,⁷⁹ and one on explosive remnants of war.⁸⁰

35.39  Oppenheim places declarations, whether made by a single State or several States, under a general heading of ‘unilateral acts’, since any legal significance they may possess is not dependent on the response of an addressee. Sometimes however a declaration, ‘perhaps for reasons of dignity’, as Oppenheim puts it, is just a treaty or convention by another name. A particularly good example is furnished by the Joint Declaration of China and the United Kingdom on the Question of Hong Kong of 1984, where it may be that the studiously neutral title was chosen in order to bridge the fundamental gap between the legal views of the two parties as to British authority over Hong Kong. Although its terms were largely written into the Basic Law of the Hong Kong Special Administrative Region on the territory’s restoration to China, the Joint Declaration undoubtedly establishes legal rights and obligations, sounding in international law, between the (p. 550) two States, and was registered as such with the United Nations Secretariat on 12 June 1985. Similar considerations may however apply to multilateral instruments as well, like the Declaration of Paris of 1856,⁸¹ which aimed at defining the rules of international law relating to blockade and contraband. Other law-making (or law-declaring) declarations are the Declaration of St Petersburg of 1868 (explosive bullets),⁸² the (unratified) Declaration of London of 1909 (blockade and contraband),⁸³ and the Barcelona Declaration of 1921 recognizing the Right to a Flag of States having no Sea-Coast.⁸⁴ But declarations are not confined to law-making treaties; the Declaration on the Neutrality of Laos signed at Geneva on 23 July 1962,⁸⁵ contains a number of specific undertakings by the signatory governments and, in accordance with its terms, is to be ‘regarded as constituting an international agreement’.

35.40  It may sometimes be difficult to determine whether an instrument which is styled a declaration amounts to a treaty rather than a pronouncement of policy expressive of a shared or common purpose. The Declaration by the United Nations of 1 January 1942⁸⁶ should almost certainly be regarded as a treaty, since acceptance of the Declaration was a condition of original membership of the United Nations. On the other hand the type of instrument which became increasingly common during the Second World War (and has persisted to this day), namely, a declaration published after a conference of three or four heads of state,⁸⁷ presents rather special features. The contents of a declaration of this kind may be partly agreements to do or not to do something, and partly statements of common policy; accordingly, it is unsafe to generalize on how far declarations of this somewhat unorthodox nature may constitute treaty instruments.⁸⁸

35.41  Secondly, under the general heading of unilateral acts, unilateral declarations are undoubtedly capable of creating rights and duties for other States.⁸⁹ One would previously have listed among these: declarations of war, declarations on the part of belligerents concerning the goods they will condemn as contraband, declarations at the outbreak of war on the part of third States that they will remain neutral. Present-day practice is less amenable to a listing of this kind. The International Law Commission had the subject under study from 1996 to (p. 551) 2006 and reported in terms of general ‘Conclusions’, to the effect that ‘formal declarations formulated by a State with the intent to produce obligations under international law’, and publicly made, may have the effect of creating legal obligations, the binding character of which is based on good faith.⁹⁰ The Commission went on to say that a unilateral declaration entails obligations for the formulating State only if it is expressed in clear and specific terms, and that, in the case of doubt as to the scope of the obligations resulting from such a declaration, such obligations must be interpreted in a restrictive manner’.⁹¹

35.42  Thirdly, the term ‘declaration’ is used to describe the action taken when States communicate to other States, or urbi et orbi, an explanation and justification of their policy or conduct, or an explanation of their views and intentions.⁹²

35.43  Declarations of these second or third types, whether they give rise to legal obligations or not, do not constitute treaties as such.

35.44  In the Ambatielos case⁹³ the International Court of Justice held that a Declaration annexed to the Treaty of Commerce and Navigation between Great Britain and Greece of 1926, and separately signed by the parties, was a ‘provision’ of the Treaty to which it was annexed, within the meaning of Article 29 of the Treaty, and consequently founded the jurisdiction of the Court to determine a dispute relating to the interpretation of the Declaration. In reaching this conclusion, the Court drew on the wording of the instruments of ratification of the Treaty, the fact that the Plenipotentiaries had included the Treaty and the Declaration in a single document, as had the United Kingdom government in publishing them in the Treaty Series, and that both countries had registered them with the League of Nations under a single number.⁹⁴

35.45  The style ‘declaration’ is also occasionally given to agreements between governments regarding some relatively minor matter. A modern example is the Declaration on the Construction of Main International Traffic Arteries of 16 September 1950.⁹⁵

35.46  The treaty concluded in the form of an exchange of Notes or letters is, in modern times, the most frequently used device for formally recording the agreement of two governments upon all kinds of transactions. It takes the form not of a single instrument but of an ordinary exchange of correspondence between the ambassador of one State and the minister for foreign affairs of the State to which he is accredited. The content of the agreement to be recorded by the exchange of Notes will of course have been agreed in advance between the two governments. The initiating Note sets out the provisions of the proposed agreement and goes on to suggest that if the proposals are acceptable to the other government the initiating Note and the other government’s reply to that effect should constitute an agreement in the matter. The exchange is then concluded by way of a Note in reply accepting the proposal in the initiating Note.

The fact that exchanges of Notes do not normally require ratification, and accordingly enter into force as soon as the exchange has been effected, makes them a very convenient and flexible instrument for treaty-makers. Although the form is used for recording agreement between two governments on a wide variety of routine matters, it has also been used to regulate transactions of considerable political importance. Thus, exchanges of Notes have dealt with such matters as the transfer and lease of military bases, the maintenance of armed forces and military missions on foreign soil, settlement of war claims, settlement of boundary disputes, delimitation of territorial waters, and jurisdiction over foreign armed forces for criminal offences.⁹⁷ More recently, the device of an exchange of Notes (p. 553) was used to record the interim agreement between the United Kingdom and Iceland in connection with the fisheries dispute occasioned by the Icelandic claim to a 50-mile exclusive fisheries limit.⁹⁸

35.48  Normally the Notes exchanged recording the agreement bear the same date, in which case, unless they provide otherwise, the agreement has effect from that date. If they bear different dates, that of the last Note, or at any rate the date of its receipt, is the governing date (unless otherwise provided), since it is the acceptance of the initiating proposal that constitutes the agreement on both sides. It is also now commonplace for the coming into force of an exchange of Notes to be postponed until the necessary internal requirements (short of ratification) have been completed on both sides. In that case the exchange of Notes will provide for each side to notify the other formally when its internal requirements have been met.

35.49  Agreements embodied in exchanges of Notes are, as a general rule, bilateral. Exceptionally, there may be more than two States concerned in an exchange of Notes or letters. The agreement between the Bank for International Settlements, on the one hand, and the United Kingdom, United States, and French governments, on the other hand, for the return to those governments of gold looted by Germany was constituted by a letter addressed to representatives of the three governments by the chairman of the Bank and by a letter in reply signed by the representatives of the three governments.⁹⁹ A somewhat unorthodox example of a ‘multilateral’ exchange is afforded by the exchange of letters on monetary questions annexed to the Treaty concerning the Accession of Denmark, Ireland, Norway, and the United Kingdom to the EEC and EURATOM of 22 January 1972. The initiating letter was addressed by Mr Geoffrey Rippon (the minister in charge of the United Kingdom delegation for the entry negotiations) to M. Thorn (Minister for Foreign Affairs of Luxembourg) in his capacity as president of the EEC Council of Ministers, and sought confirmation from ‘the Governments of the Member States of the Community and of the Governments of the Kingdom of Denmark, Ireland and the Kingdom of Norway’ that they agreed to a declaration made by Mr Rippon in the course of the negotiations. The letter in reply was signed by M. Thorn and countersigned by the eight other foreign ministers.¹⁰⁰

35.50  The ‘single-instrument’ type of treaty (whether designated as treaty, convention, agreement, protocol, or declaration) is normally concluded, and the authenticity of its text established, by means of the signatures of the negotiators, who will, as a general rule, have to produce Full Powers for the purpose.¹⁰¹ It will depend on the circumstances, but primarily on the intention of the parties as signified in the treaty, whether signature alone is sufficient to bring the treaty into force or whether some further step, such as ratification, is necessary.¹⁰²

35.51  Sometimes, however, and particularly when there may be a significant interval between the conclusion of the negotiations and the signature of a treaty (for example, in order to enable the governments concerned to satisfy themselves as to the acceptability or otherwise of the agreement as a whole), the negotiators append to it their initials ne varietur as a guarantee of the authenticity of the text. Even when it may be intended to sign a treaty immediately upon the conclusion of the negotiations, the separate pages of the text are sometimes initialled prior to its reproduction in a form suitable for signature (this practice applies to bilateral rather than multilateral treaties). The act of initialling a treaty can of itself constitute a signature of the treaty when it is established that the negotiating States have so agreed.¹⁰³ The International Law Commission point out that in practice initialling, especially by a head of state, prime minister, or foreign minister, is not infrequently intended as the equivalent of full signature.¹⁰⁴ But initialling by lesser dignitaries may equally have the effect of full signature if this is the clear intention of the parties. Thus, the Memorandum of Understanding of 5 October 1954, between the Governments of the United Kingdom, the United States of America, Italy and Yugoslavia about the Free Territory of Trieste¹⁰⁵ provides that certain things were to be done ‘as soon as this Memorandum of understanding has been initialled’ or within a specified period from the date of initialling.

35.52  The preparation for signature of a bilateral ‘single-instrument’ treaty is undertaken by the country in whose capital the treaty is to be signed. The style and (p. 555) presentation of the instrument will accord with the practices of that country. Bilateral treaties are prepared for signature in duplicate, in order that each party may retain a signed original version of the treaty instrument. Each of the two countries is entitled to precedence in the original retained by it—that is to say, its language will (where the treaty text is printed on facing pages or in parallel columns) appear in the first or left-hand page or column; its title (or the title of its sovereign and representative) will be named first in the preamble; and the signature of its representative will appear above, or to the left of, the signature of the representative of the other party.

35.53  If a treaty covers more than a single sheet of paper, the sheets will be united by ribbon or fancy cord and the ends of the ribbon will be sealed. A cover will normally be supplied by the host government.

35.54  Personal seals (ie a hand-seal or signet-ring showing crest, emblem, or other device, or the initial letters of the owner’s names) are used only when the treaty is drawn up in heads of state form and when the testimonium requires that the Plenipotentiaries should affix their seals to the treaty. Where a negotiator has no personal seal—as is nowadays usually the case—an official wax seal or wafer seal (which, in United Kingdom practice, bears the Royal Arms and the name of the diplomatic mission or the Foreign and Commonwealth Office as the case may be) is used.

35.55  A bilateral treaty between two countries which share the same language will be drawn up in that language.

35.56  Sometimes, however, a treaty may be drawn up in a language or languages which are not those of the Contracting Parties. The Treaty of Peace between Japan and Russia of 23 August/5 September 1905¹⁰⁶ was drawn up in English and French. Again a treaty may be drawn up not only in the languages of the Contracting Parties, but also in a third language, which prevails in the event of any divergence of interpretation.¹⁰⁷ Although this practice had fallen out of favour, it has now re-emerged in bilateral treaties for the promotion and protection of investments, which in several recent instances make an English-language version prevail even though it is not the language of either treaty party—presumably on the basis that English is regarded as the language of modern commerce and finance.

35.57  Where a treaty is drawn up in two or more languages, great care has to be taken to ensure the closest correspondence possible between the texts. But this may sometimes be difficult, more especially when the languages differ widely in character. The faith of negotiators that they have achieved concordance between differing language texts of a treaty may sometimes be sadly misplaced. In the important instance of the Four-Power Agreement on Berlin of 1971–72, significant nuances of difference between the English and French texts, on the one hand, and the Russian text, on the other, emerged at a very late stage immediately before signature and were never resolved. An attempt to mediate the practical consequences of this through the production of an agreed German version between the East and West German authorities who would have to implement its terms was equally unsuccessful, but the potential for future difficulty disappeared with the unification of Germany in 1990. And the 2007 Arbitration between the British and French ministers of transport and their respective concessionaires over the Concession Agreement for the Channel Tunnel revealed, in the words of the Tribunal, ‘many discrepancies between its two equal and authoritative texts’, which the Tribunal resolved by reference to the parent Treaty between the United Kingdom and France.¹⁰⁸

35.58  Desirable as it may be in theory to pre-empt problems of this kind by specifying in the treaty that one language text is to be regarded as authoritative,¹⁰⁹ national sensitivities and the circumstances of negotiation will nowadays normally make this wholly unrealistic.¹¹⁰

35.59  The number of language texts in a multilateral treaty may vary from a single text to as many as six (Arabic, Chinese, English, French, Russian, and Spanish) for treaties concluded under the auspices of the United Nations, or 23 for the treaties providing for the accession of Romania and Bulgaria to the European Union in 2005, all being equally authentic.¹¹¹ But a multilateral treaty may be drawn up in (p. 557) two or three widely spoken language texts only (for example, in English, French, and Spanish), provision being made for ‘agreed translations (or official texts) of the treaty to be established in other languages also)’.¹¹²

35.60  It is sometimes argued that the language (or languages) in which a multilateral treaty was negotiated should be given more weight in its interpretation than the other language texts. In the case of the United Nations Charter, for example, the working languages of the San Francisco Conference of 1945 were English and French, and the text of the Charter finally approved by the Coordination Committee—and from which the translations into other languages were made—was in English. There is however no evidence of international tribunals approaching questions of interpretation in this way, even though in practice it is likely that more judges and arbitrators work in English or French (or Spanish) than in other languages, and it is also true that the official and working languages of the International Court of Justice are English and French only.

35.61  The rule laid down in Article 33 of the Vienna Convention on the Law of Treaties is that ‘the terms of the treaty are presumed to have the same meaning in each authentic text’. It then goes on to deal with the case where a comparison of the authentic texts discloses a difference of meaning which the application of accepted rules of treaty interpretation does not remove, and lays down that, leaving aside the case where the treaty has provided or the parties have agreed that, in case of divergence, a particular text shall prevail, the meaning is to be adopted ‘which best reconciles the texts, having regard to the object and purpose of the treaty’.¹¹³

Even if that approach was justified by the particular question at issue in the case, the International Law Commission did not endorse it as a general rule laying down a presumption in favour of restrictive interpretation in the case of an ambiguity in plurilingual texts.¹¹⁵ Nor is there any sign of the Mavrommatis (p. 558) presumption being followed by the International Court of Justice or by arbitral tribunals. In the LaGrand case¹¹⁶ the International Court, finding itself ‘faced with two texts [of its own Statute] which are not in total harmony’, resolved the issue, in application of the rule in Article 33 of the Vienna Convention, by reference to the object and purpose of the Statute, together with the context of the provision in question.¹¹⁷ Some years previously, the tribunal in the Young Loan Arbitration¹¹⁸ had expressly rejected a submission that it should give special weight to the English text, as the language of negotiation, and had similarly applied the Vienna Convention rules by referring to other articles in the treaty, to its object and purpose, and to the negotiating history. In the Taba Arbitration of 1988, on the other hand, the Tribunal applied an English language translation (out of an Arabic translation) of a treaty in the Turkish language, on the basis that it was this translation the parties had relied upon in practice.¹¹⁹