THE INTERACTION OF TREATY AND CUSTOM IN THE CONCEPT OF OFFSHORE ARCHIPELAGOS

Abstract

The law of the sea has long been a rich source of examples of the interplay, and occasional entanglement, of treaty and custom. This article discusses whether claims to close off the waters of ‘offshore archipelagos' by non-archipelagic States are consistent with international law against the background of this perennial issue. Analysis of the 1982 Law of the Sea Convention (LOSC) demonstrates quite clearly that there is no basis for such claims. ‘Going beyond the LOSC’ the article examines whether the matter remains subject to customary international law; whether subsequent practice may have established the agreement of the parties that the relevant provisions of the LOSC are to be interpreted as allowing their invocation by non-archipelagic States with offshore archipelagos; and whether there is ‘supervening custom’ that may have emerged since the adoption of the LOSC and that permits such claims by non-archipelagic States. Identifying and critically assessing the current state of international law on these fundamental questions of the relationship between treaty and custom, it is concluded that there is no basis for arguing that non-archipelagic States are able to claim any sort of special status for ‘offshore archipelagos’.

I. INTRODUCTION

The United Nations (UN) Convention on the Law of the Sea (LOSC), adopted in 1982,¹ was the first international instrument to define the term ‘archipelagos’ and to introduce the concept of the ‘archipelagic State’ in international law.² In its Part IV, the Convention establishes a special regime for so-called ‘archipelagic States’ and allows such States to enlarge significantly the maritime zones they are entitled to by drawing straight ‘archipelagic’ baselines around the outer edges of the outer islands comprising them, claiming the waters landward of those archipelagic baselines as ‘archipelagic waters’ subject to their sovereignty (albeit with some limitations), and measuring further maritime zones seaward of the archipelagic baselines rather than seaward of the normal (or even straight, if allowed) baselines around the relevant islands.

The implications of such a regulation in the LOSC are evidently far reaching both for the States that can invoke it and for other States, whether neighbours or maritime powers. Large areas of sea can be ‘closed off’ by the archipelagic State and be subjected to its sovereignty. It is no doubt for this reason that the Convention adopts a definition of archipelagic State that immediately limits the number of States that can take advantage of the far-reaching regime of Part IV: such archipelagic States must be wholly constituted by one or more archipelagos, defined in turn as

a group of islands, including parts of islands, interconnecting waters and other natural features which are so closely interrelated that such islands, waters and other natural features form an intrinsic geographical, economic and political entity, or which historically have been regarded as such.³

In that, the LOSC establishes a ‘juridical’ archipelagic State, which is not in line with the potential geographical or geomorphological content of the homonymous term beyond the law—much like it does with other terms, such as ‘bay’,⁴ for example, or ‘continental shelf’.⁵ In particular, it clearly precludes any State with continental territory from claiming the status of, and thus the rights that accrue to, an archipelagic State.

All this should be rather straightforward; and yet there have been claims by States that do not fall within the definition of archipelagic State under the LOSC to apply rules similar to those applicable to archipelagic States in order to close off the waters around so-called ‘offshore’, ‘mid-ocean’ or ‘outlying’ archipelagos⁶ over which these States have sovereignty (‘offshore archipelagos'). Such claims have been the subject of protest, and recently a number of States have set forth their general understanding of the Convention as precluding such claims.⁷

The purpose of this article is to discuss whether such claims to close off the waters of offshore archipelagos by non-archipelagic States are consistent with international law. In order to do this, it revisits the general issue of the relationship (and potential entanglement) between treaty and custom: if the LOSC already regulates who is and who is not an archipelagic State (and thus enjoys or does not enjoy the rights that accrue to such a State), is it possible to argue that non-archipelagic States might have similar rights with respect to offshore archipelagos under their sovereignty? And, if so, would that be on the basis of customary law? How would such customary law relate to the treaty? Or would such a claim be on the basis of practice that has somehow led to agreement as to a particular interpretation of the LOSC (which might even be contra legem)? Or would a claim be, even, on the basis of practice that has led to a modification of the LOSC?

The discussion is pursued in the following steps: First, it considers the terms ‘archipelagos’ and ‘archipelagic State’ in international law, and traces their development and status pre-LOSC (Section II). Next, it focuses on the regulation of archipelagic States, and their rights and obligations, in the LOSC, against the background of the travaux preparatoires—ie the discussion in the Third UN Conference on the Law of the Sea (UNCLOS)⁸ (Section III). It is argued that the outcome of this analysis demonstrates quite clearly that there is no basis for claiming rights accruing to an archipelagic State for non-archipelagic States that have sovereignty over one or more offshore archipelagos.

This sets the stage for a discussion of possible arguments for ‘going beyond the LOSC’ and seeking to make an argument on the basis of customary international law, or, more timidly, subsequent treaty practice. Hence, Section IV addresses the relationship between treaty and custom and the argument that the question at hand is actually not regulated by the LOSC and remains subject to customary international law. Section V seeks to determine whether there is an argument that subsequent practice under the LOSC may have established the agreement of the parties that its relevant provisions are to be interpreted as allowing their invocation by non-archipelagic States with offshore archipelagos. Finally, Section VI considers a potential argument regarding ‘supervening custom’, ie customary law that may have emerged since the adoption of the LOSC and that permits such claims by non-archipelagic States. Section VII concludes.

II. DEVELOPMENT OF TERMS AND STATUS PRE-UN CONVENTION ON THE LAW OF THE SEA

An archipelago is, according to the Oxford English Dictionary, originally a term used for the Aegean Sea, dotted as it is with many islands, and ‘hence, any sea or sheet of water in which there are numerous islands; and transferred a group of islands’.⁹ Its etymology, though somewhat complicated in terms of history, simply comes from the Greek words archi- (meaning ‘chief’ or ‘principal’) and pelagos (meaning ‘sea’).¹⁰ Based on this generic definition of the term, any more or less circumscribed sea with numerous islands, or for that matter any group of islands in some sort of proximity or otherwise close connection to one another, could be characterised as an ‘archipelago’. Accordingly, any State that possesses such an archipelago (or even more than one archipelago) could be characterised as an archipelagic State. However, as already noted above, the legal meaning of a term—especially in the law of the sea—is often (and sometimes far) removed from its meaning in ordinary speech or in other disciplines.

The issue of archipelagos and their legal regulation emerged early on¹¹ in discussions relating to the codification of the law of the sea.¹² However, neither the 1930 Hague Codification Conference nor the International Law Commission (ILC) in 1956,¹³ nor UNCLOS I and UNCLOS II in 1958 and 1960, respectively, were able to overcome the difficulties involved in the ‘group of islands’ issue. This was despite discussions gaining new momentum in 1951 through the International Court of Justice (ICJ) decision in the Fisheries Case.¹⁴ Until then the principal focus of States and in the academic literature was on the extent and delimitation of territorial waters where the coastal State possessed scattered or complex geographical features including fringing islands constituting a coastal archipelago.¹⁵ The 1951 Fisheries Case sparked some debate about whether its principles could be applied to independent mid-ocean archipelagic States or dependent non-coastal archipelagos of continental States.¹⁶

While the physical characteristics of archipelagos vary widely, and clearly posed challenges for legal regulation of the issue pre-UNCLOS III, three broad geographic categories are commonly identified in the literature. One is this concept of coastal archipelagos, which consist of a group of islands that are ‘situated so close to a mainland that they may reasonably be considered part and parcel thereof, forming more or less an outer coastline’.¹⁷ Such continental land masses with fringing islands are catered for with the drawing of straight baselines first recognised in the 1951 Fisheries Case ¹⁸ and reflected in Article 4 of the 1958 Geneva Convention on the Territorial Sea and Contiguous Zone and in Article 7 of the LOSC.¹⁹ Even prior to UNCLOS III, ‘there appeared to be a consistent body of practice which treated coastal archipelagos as a unit forming an outer coastline [ie a baseline] from which to measure the territorial sea’.²⁰

A second category is mid-ocean or outlying archipelagos which are ‘groups of islands situated out in the ocean at such a distance from the coasts of firm land as to be considered as an independent whole rather than forming part or outer coastline of the mainland’.²¹ These may comprise archipelagos forming the whole territory of States, such as the Maldives, Fiji, Indonesia and the Philippines. In contrast with coastal archipelagos, early State practice revealed a ‘profusion of different views and approaches with regard to the delimitation of the territorial waters of outlying archipelagos’.²² In the preparatory work undertaken by the ILC (1949–1956) for UNCLOS I in 1958, Special Rapporteur François had introduced the concept of ‘groups of islands’ or archipelagic status in his 1953 Report.²³ However, the widely varying physical characteristics of archipelagos was one of several reasons cited by the ILC for the omission from its 1956 Draft Articles of any provision on ‘groups of islands’, acknowledging like The Hague Conference before it an inability to surmount the difficulties involved.²⁴ Nor was the issue resolved at UNCLOS I in 1958, with mid-ocean archipelagos a ‘conspicuous gap’ in the resulting 1958 Conventions,²⁵ one which remained after the (inconclusive) UNCLOS II in 1960.²⁶

A third category are dependent archipelagos belonging to continental States, such as the Faeroe Islands (Denmark), the Galapagos Islands (Ecuador) and the Andaman and Nicobar Islands (India).²⁷ States, such as the archetypal archipelagic example of Greece, which combine continental territories with significant island formations offshore, were also interested in achieving some special status or beneficial regulation at UNCLOS III, but, as it shall be seen below, were unsuccessful in that endeavour.

Before UNCLOS III then, while the concept of the coastal archipelagos was well established (even if uncertainties would persist over the drawing of straight baselines), no customary international law on ‘offshore’ archipelagos appears to have emerged.²⁸ Given the unsuccessful attempts at UNCLOS I and II to introduce the concept of a ‘group of islands’ or ‘archipelagic State’ into the treaty text—at a time when ‘most archipelagos belonged to colonial maritime powers which were mostly interested in the freedom of navigation on the high seas’—there was then a concerted campaign by Indonesia and the Philippines in particular to ensure that UNCLOS III resulted in recognition of archipelagic status for newly independent archipelagic States.²⁹

III. UNCLOS III AND THE ARCHIPELAGIC STATE REGIME OF THE LOSC

The UN Convention on the Law of the Sea is the archetypal ‘package deal’.³⁰ The negotiations in UNCLOS III demonstrated beyond any doubt that only such an approach would be capable of yielding a coherent instrument that would not unravel post-adoption. This is because of the different groups of States, with different interests, pursuing different objectives in the different areas of regulation covered by such an expansive instrument, which proposes to regulate (at least at overview level) almost every aspect of the sea and its uses.

The negotiation of a regime for ‘archipelagic’ States was no different.³¹ The Philippines and Indonesia³² unsurprisingly took the lead in trying to introduce relevant regulation that would benefit them and other States in a similar position.³³ States like Greece and others who combined continental territories with significant island formations offshore were also interested in achieving some special status or beneficial regulation. Also, maritime powers had all sorts of reasons to be worried about such regulation, which would have the effect of subjecting potentially large expanses of the seas to the sovereignty or jurisdiction of whoever was to be characterised as an archipelagic State. Indeed, as noted above, it was differences of view over the status of the waters within ‘groups of islands’ which had been a reason for their deletion from the ILC's 1956 draft and omission from the 1958 treaty texts.

A crucial threshold issue was thus eligibility to claim archipelagic status and what this might entail, mindful of the geographical variety among archipelagos, including the distinction between continental and mid-ocean archipelagos, and the potential for vastly increased claims restricting navigational freedoms. During the 1973 sessions of the Seabed Committee preparing for UNCLOS III, Fiji, Indonesia, Mauritius and the Philippines introduced principles which might govern an archipelagic regime.³⁴ Following acceptance by a major maritime power—the United Kingdom—of the archipelagic State concept ‘subject to objective criteria regarding the identification of archipelagic States, and safeguards with respect to navigational freedoms’,³⁵ these principles resulted in ‘Draft Articles on Archipelagos’ being prepared that formed the basis for formal proposals eventually put to UNCLOS III in 1974.³⁶ A breakthrough in UNCLOS negotiations then occurred with Bahamas’ introduction in 1975 of ‘18 Principles for Inclusion in Archipelagic Articles’.³⁷ The hopes of Greece and others were dashed by 1976 when negotiations narrowed on the concept of the mid-ocean archipelago alone, and not those archipelagos associated with a continental State, in what would become Part IV of the final convention text.³⁸

The archipelagic regime finally adopted is thus characteristic of the Convention as a whole, including its ‘package deal’ nature: there are some significantly beneficial provisions for archipelagic States, balanced out with some serious concessions in favour of freedom of navigation, which was the primary concern of maritime powers. And so, the matter of offshore archipelagos was resolved by the Convention: there are certain requirements for States to benefit from what is quite clearly a new and special ‘archipelagic State’ status, and States with continental territory and one or more offshore archipelagos cannot do so. Today, 22 States claim³⁹ archipelagic status,⁴⁰ with an estimated 35 States entitled to do so consistently with the Convention.⁴¹

The first and most important limitation of the archipelagic State regime comes in the definition of an archipelagic State. According to Article 46(a) of the Convention, ‘“archipelagic State” means a State constituted wholly by one or more archipelagos and may include other islands’.⁴² Article 46(b) defines an archipelago as:

a group of islands, including parts of islands, interconnecting waters and other natural features which are so closely interrelated that such islands, waters and other natural features form an intrinsic geographical, economic and political entity, or which historically have been regarded as such.

Though the legal definition of archipelago provided is quite close to the ‘ordinary’, for lack of a better term, meaning of the term discussed in Section II above, the legal concept of the ‘archipelagic State’ immediately excludes any State with continental as well as insular territory. It does this by requiring that the archipelagic State be ‘wholly constituted’ by one or more archipelagos as defined in the Convention, along with, possibly, other islands. The effect of this definition is clear: any State with continental territory cannot be considered an archipelagic State and thus benefit from the special regime established in the Convention, irrespective of how many archipelagos it may actually have in addition to its continental territory.⁴³ Even the archetypical archipelagic State, the one whose archipelago provided the term itself (Greece), is not actually an archipelagic State under the LOSC and cannot benefit from the relevant special regime.⁴⁴

This may be the starkest, but it is not the only limitation when it comes to invoking the beneficial regulation of Part IV of the LOSC. According to Article 47, being ‘archipelagic’ entitles a State to draw ‘straight archipelagic baselines’ joining the outermost points of the outermost islands and drying reefs of the (or: each) archipelago.⁴⁵ In the same breath, however, the provision requires that ‘within such baselines are included the main islands and an area in which the ratio of the area of the water to the area of the land, including atolls, is between 1 to 1 and 9 to 1’.⁴⁶ Paragraph 2 of Article 47 compounds the situation by limiting the length of straight archipelagic baselines to no more than 100 nautical miles, ‘except that up to 3 per cent of the total number of baselines enclosing any archipelago may exceed that length, up to a maximum length of 125 nautical miles’. Notably, this is the only provision in the Convention that sets out a maximum length of baseline; the position regarding the maximum length of straight baselines under Article 7 of the LOSC is still disputed.⁴⁷ In that sense, the provision also highlights that the ‘straight’ baselines of Article 7 and the ‘straight archipelagic baselines’ of Article 47 are not the same type of baseline—a consideration that will be important later on.

These conditions, ie the water-to-land ratio and the maximum length of baseline requirement, are quite important: though the definition of the Convention may allow States such as Japan or Iceland to claim archipelagic State status, this would be of little help to them. The conditions for drawing archipelagic baselines, and thus of establishing a zone of archipelagic waters—the main benefit of being characterised as an archipelagic State—would not be fulfilled in their cases, and thus the claim of status would be a rather empty gesture.⁴⁸ There are further conditions for the drawing of straight archipelagic baselines under Article 47,⁴⁹ which are quite similar to those for drawing straight baselines under Article 7. In common with other technical provisions in the LOSC such as Article 76 reflecting geographic (and geomorphological) circumstances,⁵⁰ Article 47 resulted from ‘extensive consultations with the principal aspiring archipelagic States’.⁵¹ This has resulted in provisions regarding which there has been substantial compliance by archipelagic States claiming this status; where departures have occurred, these have either been remedied through subsequent adjustment in response to protest or are relatively minor in character.⁵²

The establishment of archipelagic waters through the drawing of straight archipelagic baselines around the outermost points of the archipelago is, as noted, the main benefit to which an archipelagic State may lay claim. This is a zone of sovereignty,⁵³ very similar in status to the territorial sea. The sovereignty extends to airspace and seabed and subsoil, along with all natural resources contained therein.⁵⁴ There are, however, significant limitations regarding existing agreements, traditional fishing rights and existing submarine cables,⁵⁵ as well as regarding innocent passage⁵⁶ and a special type of passage called ‘archipelagic sea lanes passage’ that bears more similarity to transit passage through international straits.⁵⁷

All these provisions seek to circumscribe the benefits that accrue to archipelagic States from their status as such States. However, they also have the effect of demonstrating that this is a wholly new and special regime, a regime that has been created and exists exclusively under the LOSC. This is particularly evident, for example, in the provision of Article 51 of the LOSC regarding ‘existing’ agreements, traditional fishing rights and ‘existing’ submarine cables. A new regime is being created, but certain pre-existing rights are preserved.⁵⁸

A further restriction arises from the requirement that an archipelagic State be ‘constituted wholly by one or more archipelagos’, thereby excluding from the scope of the archipelagic regime mainland, coastal or continental States having outlying or dependent archipelagos from the archipelagic claim (eg Denmark, Ecuador, Portugal, Spain, and Norway) as well as coastal archipelagos lying close to the continental States which are governed by the straight baselines regime of Article 7 (eg the Norwegian Skjaergaard, the Hebrides (United Kingdom) or the Frisian Islands (Germany)).⁵⁹

The status of dependent offshore archipelagos was raised at the second session of UNCLOS III in 1974 where nine continental States submitted a working paper arguing for the extension of the mid-ocean archipelago regime to continental States.⁶⁰ In the words of Portugal:

[T]he arguments in favour of the establishment of a special régime for archipelagic States were also valid for archipelagos forming part of the territory of a coastal State, particularly with regard to the security and economic interests of such States. Application of a different régime to the latter would mean that the archipelagic part of the territory of mixed States would be regarded as second class territory.⁶¹

However, this was met by objections from a number of States with concerns that it would lead to a proliferation of unfounded claims and erosion of the freedom of navigation,⁶² and the proposal was omitted from the revised single negotiating text.⁶³ The travaux thus demonstrate that the issue was discussed and rejected during UNCLOS III, leading to the conclusion that offshore archipelagos were not excluded tout court from the Convention (ie from its scope), but rather were regulated negatively in the sense that they did not benefit from the provisions regarding archipelagic baselines and archipelagic waters.⁶⁴

Subsequently, several of the continental States with offshore groups of islands not meeting the juridical definition of archipelago in Article 46 of the LOSC have used straight baselines to enclose such islands in a manner described as ‘simulating an archipelago’.⁶⁵ Examples include Denmark (the Faroe Islands), Norway (Svalbard), and Ecuador (Galapagos Islands).⁶⁶ It is not always made clear whether these baselines are claimed as Article 7 straight baselines or as archipelagic baselines under Article 47 (which requires, it will be recalled, satisfaction of the definition of archipelagic State in Article 46 and the proclamation of such status).⁶⁷ Moreover, baselines enacted by continental States, including Ecuador's baselines around the Galapagos Islands⁶⁸ and China's baselines around the Senkaku Islands and the Paracel Islands,⁶⁹ have met with protest.⁷⁰ This has particular implications for arguments regarding subsequent practice under the LOSC, discussed in Section V below, and for supervening custom, discussed in Section VI.⁷¹ As Churchill saliently observes in connection with State practice and Article 7, non-conforming and diverse practice by a minority of States which has met with protest from ‘at least eight different States and the EU … leads to the conclusion that practice relating to the drawing of straight baselines does not amount (yet) either to an agreed interpretation of the Convention or a new rule of customary international law’.⁷²

IV. LOSC AND CUSTOMARY LAW ON OFFSHORE ARCHIPELAGOS

What then is the relationship of the LOSC and customary international law regarding the regulation of offshore archipelagos? As a well understood general proposition, in the relationship between two or more States, a treaty regulating a particular matter takes precedence over customary law on the same matter.⁷³ The position is clear and the case law robust. This is simply because the treaty rule is specifically negotiated between the parties, not to mention that it is overwhelmingly likely to be far more precise than any customary rule could ever be—being written and all.⁷⁴ Specific consent must take precedence over general consent. According to the ICJ in North Sea Continental Shelf, if there is a treaty binding for all the parties in the case, ‘then the provisions of [that treaty] will prevail in the relations between the Parties, and would take precedence of any rules having a more general character, or derived from another source’.⁷⁵

Equally settled, at least since Nicaragua,⁷⁶ if not long before, is the understanding that treaty rules and customary rules on the same matter retain their separate existence and parallel application, and one does not subsume or extinguish the other in any way, even when they have identical content. To use the rather poetic expression of Roucounas, ‘contrary to domestic law, where custom fades [away once it comes] into the arms of written law’, in international law the existence and autonomous applicability of each source continues.⁷⁷

The LOSC devotes Part IV to dealing with archipelagic States. It excludes from the definition of archipelagic States any States that have sovereignty over continental (as opposed to exclusively insular) territory. It then establishes a particular regime that is applicable to archipelagic States. All this has already been discussed above. As it emerged from that discussion, the States quite deliberately excluded certain States from qualifying as archipelagic States, and as such from being able to invoke and apply the regime that is applicable to archipelagic States. Accordingly, the matter should be clear: as between LOSC parties, at the very least, the provisions of the treaty prevail and operate to the exclusion of any rules of customary law (if it is assumed that customary law has different content). However, an argument can be (and has been)⁷⁸ made that the LOSC does not deal at all with the question of offshore archipelagos, whose legal status falls to be determined by customary law.

A. When is a Matter (Exhaustively) Regulated by Treaty?

The extent to which a matter is regulated by the LOSC is a matter of treaty interpretation.⁷⁹ The argument that the LOSC does not deal with offshore archipelagos would mean that these remain regulated by customary law that was in existence at the time of negotiation and conclusion of the LOSC, and that they were merely ignored or ‘let be’ by the drafters of the treaty. This argument relies in part on the fact that the last clause of the Convention's preamble states that ‘matters not regulated by this Convention continue to be governed by the rules and principles of general international law’.⁸⁰

First, this provision merely reiterates the general position under international law: even if it were not present, any matters not regulated by the Convention would fall to be governed by general international law (ie customary law and general principles of law). Why was it then included? Well, that can be seen as general practice in preambulatory clauses⁸¹ and is perhaps due to an abundance of caution on the part of the drafters. However, the reality is that it cannot, in itself, form the basis of any argument that something is not regulated by the treaty. By contrast, the real question is precisely what is regulated by the treaty, and this is a matter of determining the scope of the treaty.

Secondly, such argument presumes there is pre-existing customary law that regulates offshore archipelagos, allowing the drawing of straight baselines around their outermost points and subjecting the waters included within those baselines to a special regime. It has been argued above that such a claim cannot be sustained, as there is not sufficient State practice and opinio juris to sustain it, nor could there be any approximating the detailed regulation/establishment of the archipelagic regime achieved in the LOSC.

Thirdly, and more importantly, for the argument to work it would be necessary to show that the matter of offshore archipelagos is somehow not regulated by the LOSC. The authors do not think this is the case, and indeed for a number of reasons. These are taken up in turn below.

1. The travaux

As it has also been seen, the travaux, discussed in Section III above, indicate that the matter of offshore archipelagos was raised and discussed and that their inclusion in the special regime was rejected rather than ignored by the drafters. This means that offshore archipelagos do fall within the scope of regulation of the Convention, and are merely negatively regulated by being excluded from the special regime (thus being subject to the general regime of the Convention regarding islands, baselines and the like). When something is proposed and not accepted, can one really argue that it has thus not been regulated?

2. The nature of the LOSC as a ‘package deal’

The point regarding the travaux, however, is not limited to the travaux. It also accords with the nature of the LOSC as a ‘package deal’ and with its purpose of regulating ‘all issues relating to the law of the sea’ in accordance with the very first preambulatory clause. This clause, by the way, is much more potent than the final preambulatory clause: not only is it specific to the Convention rather than a generic restatement of a standard position under general international law, but it also permits reading the final preambulatory clause in its proper light: this Convention aims to regulate all issues relating to the law of the sea—if perhaps any issues have been completely disregarded or emerge subsequently (ie were unknown at the time of conclusion), these would fall, as usual, to be regulated by customary law.⁸² However, issues that have been discussed and rejected are not such issues.

The nature of the Convention as an all-encompassing package deal is indicated not only by the first preambulatory clause, but much more importantly by Article 309, according to which ‘[n]o reservations or exceptions may be made to this Convention unless expressly permitted by other articles of this Convention’. Neither Part IV nor most other substantive parts of the Convention expressly permit reservations (though Part XV, notably a non-substantive part, does permit States to make reservations and exceptions regarding dispute settlement). The reason for that is simple: this enormous Convention of 320 articles and nine annexes, which took almost a decade to negotiate, is the result of significant concessions-in-return-for-concessions in different substantive areas of regulation.

For Indonesia and the Philippines to achieve the archipelagic State regime in Part IV, they will have made concessions not only in the context of that regime, but in other parts of the Convention as well. For Greece to be appeased not having been included in the archipelagic State regime, it will have achieved or extracted concessions in other parts of the Convention—and so forth. Allowing reservations in the context of such a multi-layered complex deal would risk the deal unravelling before the ink on the paper had even managed to dry: States would come in and try to regain by means of reservations or exceptions that which they had conceded in order to attain other benefits. However, if everyone does that, the whole thing falls apart.

To argue that offshore archipelagos, though raised and discussed in the negotiation of the Convention, and ultimately excluded from the special archipelagic regime that it established, remain ‘unregulated’ by the Convention and thus subject to customary law (assuming that any such existed) would precisely run counter to this concept of the ‘package deal’. In effect it would amount to allowing ‘silent’ reservations or exceptions to a Convention that clearly aspires to be all-encompassing.

There is further support for this line of argument: in those instances where the Convention drafters have considered that an issue could conceivably be understood as falling within the scope of the Convention, but they did not wish it to be so, they made this abundantly clear. In particular, this can be seen in savings clauses such as that of Article 32:

With such exceptions as are contained in subsection A [of Section 3 of Part II] and in articles 30 and 31, nothing in this Convention affects the immunities of warships and other government ships operated for non-commercial purposes.

3. The ‘same subject-matter’ and conflict of norms

Beyond the points made above, how can one determine whether something is or is not regulated by a particular treaty? Theoretically, any definition within a treaty which gives rise to a special regulatory regime for that which is defined excludes that which is not within the scope of the definition. Does this mean that the treaty excludes that which is not within the scope of the definition from regulation altogether? In order not to make a mockery of every attempt to delimit scope by such a contrario arguments, it would have to be determined when something is of a different subject-matter to something else. If things are of a different subject-matter, that which regulates one cannot be seen as regulating the other.

But when are things of a different subject-matter? In fact, there is something that can be of help in answering this question, or rather its inverse: when are things of the same subject-matter? This is a question under Article 30 of the Vienna Convention on the Law of Treaties (VCLT), which deals with ‘application of treaties relating to the same subject-matter’.⁸³ The only way to determine whether two sets of rules relate to the same subject-matter is in reality to try and apply them to the same set of facts, and see whether they lead to incompatible outcomes.⁸⁴ This incompatibility, this conflict, demonstrates that the two sets of rules regulate, in effect, the same subject-matter. Accordingly, one of them must be chosen over the other in order to avoid conflicting outcomes, and this is exactly what ‘conflict rules’, such as lex posterior or lex specialis, achieve.

If this same logic is applied to archipelagic States under Part IV of the LOSC and offshore archipelagos allegedly left to be regulated under customary law, it results in a rather peculiar—and indeed conflicting, if not quite simply absurd—outcome: those States that fall within the scope of Part IV of the LOSC and qualify as archipelagic States would have fewer rights compared to States that have offshore archipelagos: the archipelagic baselines of the former would be subject to water-to-land ratios and maximum length, whereas presumably those of the latter would not be, or not to the same extent (given that customary law is hardly ever as precise as stipulating ratios and maximum lengths). As such, it would be better for an archipelagic State not to claim that status, in order to benefit from the more relaxed rules of the other regime. This is even without going into questions of rights of passage of other States through the enclosed waters and so forth—a rather cumbersome regime for archipelagic States, but presumably at best subject to innocent passage in the case of offshore archipelagos. The point is made.

4. The principle of effectiveness

Additionally, it is worth mentioning the principle of effectiveness, or ut res magis valeat quam pereat. According to this principle, when there are two possible interpretations of a treaty provision (or a set of provisions), that interpretation should be preferred which makes the provisions effective, over an interpretation that would have the effect of rendering them redundant.⁸⁵ This is simply because it should not be presumed that States negotiate and conclude treaties which include provisions without effects (if there is an alternative interpretation which endows them with effect). However, that would be precisely the outcome of interpreting the archipelagic State provisions as not covering offshore archipelagos: the whole regime would be rendered redundant—archipelagic States would be far better off just opting for the (alleged) customary law regime, so much more relaxed and vague as it will be.

This, if anything, confirms that—travaux and all the rest of it aside—the provisions of Part IV of the LOSC must be read as covering the field, ie as regulating any and all aspects of archipelagos, including offshore archipelagos. The latter are simply negatively regulated in the sense that they are excluded from the regime and cannot benefit from it. In other words, they are simply subject to the normal rules regarding maritime features.

5. The understanding of States

Finally, it should be noted that States themselves consider the Convention to be covering the field and exhaustively dealing with all matters regarding the law of the sea to which it refers. This is reiterated every year in the UN General Assembly ‘Omnibus Resolution’ on ‘Oceans and the Law of the Sea’. In its latest iteration, adopted by the UN General Assembly (UNGA) on 30 December 2022, the General Assembly ‘[e]mphasis[es] the universal and unified character of the Convention, and reaffirm[s] that the Convention sets out the legal framework within which all activities in the oceans and seas must be carried out’.⁸⁶

V. SUBSEQUENT PRACTICE ON LOSC PROVISIONS REGARDING ARCHIPELAGIC STATES

Could the subsequent practice of the parties to the LOSC have established their agreement as to a rather peculiar interpretation of provisions of the LOSC allowing non-archipelagic States to apply part of the archipelagic regime to offshore archipelagos? If it is conceded that the LOSC does govern the matter of offshore archipelagos, at least negatively, ie by excluding them from the relevant special regime regarding archipelagic States, it could still be possible to argue that the subsequent practice of LOSC parties has established their agreement over an interpretation⁸⁷ of the relevant provisions of the LOSC along the lines of allowing the application of (some parts of) the archipelagic regime to offshore archipelagos, which were originally denied that regime.

The first point to raise in this connection is that the relevant subsequent practice would have to be ‘in the application of the treaty’.⁸⁸ However, as appears from the overview of the practice that exists regarding offshore archipelagos, any drawing of baselines does not seem to be based on an interpretation or application of the LOSC. This is evident from the fact that States engaging in such practice are unclear as to whether they are drawing straight baselines on the basis of Article 7 or Article 47 of the LOSC.⁸⁹ Even more importantly, such practice is unclear as to the interpretation or application of other Convention provisions regarding the status of waters enclosed by the straight baselines drawn. For example, are they considered internal waters? Or are they internal waters subject to the regime of innocent passage? Or are they a special zone subject to the regime of transit or archipelagic sealanes passage?

At the very best, any such ‘subsequent practice’ is not ‘in the application’ of the LOSC, but rather on its non-application, signifying that the matter of offshore archipelagos is not considered as falling within the scope of the Convention and is regulated by customary law. It has already been pointed out in Section IV why this would make little, if any, sense. However, if it is considered that in principle something like that would be possible, there is an even more serious stumbling block which such an argument would have to overcome.

This is that to qualify as relevant practice, the subsequent practice must ‘establish[…] the agreement of the parties’ as to the interpretation of the treaty.⁹⁰ This of course does not require all the parties to the treaty to engage in the relevant practice.⁹¹ However, it does require at the very least their acquiescence to such practice that is known to them and calls for reaction.⁹² As already discussed in this respect above, the attempts of States to draw straight baselines around offshore archipelagos have indeed been protested, and this is enough—at least according to the ILC—to preclude the practice from establishing any agreement as to the interpretation of the Convention.⁹³ Moreover, in addition to protesting particular actions, some States have been at pains generally to confirm their understanding that there is no legal basis for continental States to claim archipelagic status nor to draw straight archipelagic baselines under the LOSC.⁹⁴

It is worth mentioning, finally, that any such subsequent practice that would either exclude the offshore archipelago issue from the coverage of the Convention or apply the archipelagic State regime to offshore archipelagos (even that is unclear, but let it be accepted that it is not) would effectively amount to a modification of the LOSC. According to the ILC, such modification cannot take place through subsequent practice under Article 31(3)(b) of the VCLT.⁹⁵ It is not necessary to belabour this point, as the argument based on subsequent practice is not even capable of establishing any agreement. However, it is mentioned here, without taking a position, for reasons of completeness.

The remaining possibility is that subsequent practice could be relevant under Article 32 of the VLCT.⁹⁶ This does not require ‘establishing agreement’—it could be merely the practice of one or a few parties, even though not accepted by others.⁹⁷ However, any such practice would constitute only a supplementary means of interpretation, whose sole effect would be either to confirm interpretation under Article 31 of the VCLT, or to deal with a situation where interpretation under Article 31 of the VCLT yields a result that is unreasonable or absurd.⁹⁸ Neither of these situations are here present: interpretation in accordance with the general rule of Article 31 of the VCLT leads to an understanding that the archipelagic State regime excludes offshore archipelagos from the special regulation, as discussed above, meaning that the relevant provisions of the LOSC negatively regulate this question. This outcome would be sought to be overridden by ‘subsequent practice’ under Article 32 of the VCLT, which is not possible. Further, this outcome is neither unreasonable nor absurd—quite the opposite. It is rather the position that would allow the application of such a special regime to offshore archipelagos that appears absurd against the background of the Convention.

VI. SUPERVENING CUSTOM?

Having established that the matter of offshore archipelagos is in fact regulated by the LOSC, and that any subsequent practice has not established the agreement of the parties regarding an interpretation of the LOSC that extends (aspects of) the regime of archipelagic States to non-archipelagic States with offshore archipelagos, the only potential remaining argument is that new (supervening) customary law may have emerged that now somehow supersedes the LOSC regulation and achieves the same result of extending (aspects) of the regime as described.⁹⁹

This argument must also begin by conceding (correctly) that the LOSC regulation does cover the question of ‘offshore archipelagos’, at the very least negatively. It thus denies the regime of archipelagic States to continental States with offshore archipelagos. However, the argument goes, new customary law has emerged since the negotiation and conclusion of the LOSC, and that new customary law regulates the question of offshore archipelagos differently to the LOSC. That supervening custom should take precedence over the LOSC regulation.¹⁰⁰ After all, the law must be allowed to progress and evolve.

This customary law must have developed post-conclusion of LOSC if it is to ‘supervene’ it and escape the standard relationship of pre- and co-existing custom with treaties already discussed in Section IV above.¹⁰¹ However, there are difficulties with this argument, some of which are by-and-large similar to those encountered by the argument of subsequent practice discussed in Section V above.

The first such difficulty relates to the practice required for the establishment of a customary rule of international law. With respect to subsequent practice, uncertainty was noted as to whether States claiming the right to draw straight baselines around offshore archipelagos were seeking to do so under Article 7 or Article 47 of the LOSC. In that case, this meant that the practice was not really ‘in the application of the treaty’. With respect to the emergence of supervening custom, the difficulty lies in the fact that it is not clear what the legal claim accompanying the practice is.

Even if it were accepted that there is a general practice that allows States to draw straight baselines around offshore archipelagos—and it has already been seen that this is not even nearly so—this practice would also need to be accompanied by some legal claim that they are permitted to do so—and that other States are obligated to recognise this. This requirement for practice to be coupled with a claim in law is commonly referred to as opinio juris. But what would be the opinio juris in this instance? For example, would it be that States are entitled to draw straight baselines around offshore archipelagos? If so, then what kind of lines? Would they be akin to those under Article 7 of the LOSC? Would they be possibly akin to those under Article 47 of the LOSC, ie with more significant limitations? Or would they be straight baselines unencumbered by any of these limitations? What would be the status of waters enclosed by such baselines (a question that follows from the nature of the baselines as being akin to those under Article 7 or under Article 47 of the LOSC)? What would be the rights of other States in those waters?

The second difficulty is one of logic, and one that that has already been discussed at some length in previous sections: the absurdity that would ensue by having a treaty regime that is rather circumscribed for archipelagic States, and what must surely be a far more relaxed regime under ‘supervening’ customary international law. This would effectively relegate the relevant provisions of the LOSC into the dustbin of history: why would an archipelagic State not claim the much more liberal regime under supervening customary law?

This relates to the third difficulty with respect to a ‘supervening custom’ argument, in fact a safety valve built into that theory. According to most of the literature on the subject¹⁰²—and it is a subject on which there is rather a dearth of case law, perhaps not without reason—supervening custom does not actually automatically entitle parties to the treaty (168 in the case of the LOSC currently) simply to disregard allegedly superseded provisions of the latter and apply the later-in-time customary law.¹⁰³ In the limited cases where it has arisen, the ICJ's approach has been to recognise that there may be a need to reassess treaty rights and obligations in the light of developments in customary international law but that any treaty modifications should be introduced by negotiation between (and the consent of) the parties.¹⁰⁴ Furthermore, there is evidence of judicial self-restraint in taking account of supervening customary law in treaty interpretation where this might lead to modification of treaty provisions. The law of the sea furnishes an example: in the Guinea-Bissau/Senegal arbitration, the Tribunal refused to interpret a 1960 agreement between the parties reflective of the maritime zones of that time, delimiting their territorial seas, contiguous zones and continental shelves, as extending to the delimitation of newly emergent (and proclaimed) exclusive economic and fisheries zones as this ‘would involve a real modification of [the agreement's] text and … it is the duty of a court to interpret treaties, not to revise them’.¹⁰⁵ To be sure, this is a reflection of proper judicial function, which is to interpret and not to modify treaties,¹⁰⁶ but it is also indicative of the safety valve at work.

At best, what supervening custom has the effect of doing is to establish a good faith obligation between the parties to the treaty to enter negotiations with a view to amending the treaty to bring it in line with new customary international law.¹⁰⁷ What supervening customary law clearly does not do is to prevail over the treaty rule, leading to the observation that ‘a claim that a treaty is incompatible with more recent customary rules may be of limited practical value’.¹⁰⁸

In any event, it is worth noting here again, as above,¹⁰⁹ that States themselves have not only widely ratified the LOSC, but also annually emphasise ‘the universal and unified character of the Convention’ and reaffirm ‘that the Convention sets out the legal framework within which all activities in the oceans and seas must be carried out’ in the ‘Omnibus Resolution’ on ‘Oceans and the Law of the Sea’. This resolution is voted on and passes by a large majority every year (eg in 2022, it received 159 votes in favour, one against (Turkey) and three abstentions (Colombia, El Salvador, and Syria)). This would further stress how high the bar for demonstrating that supervening customary international law has developed would be—a bar that, in accordance with the discussion in this section, is nowhere near passed.

VII. CONCLUSION

Part IV of the LOSC succeeded where previous codification attempts had failed, in establishing a new archipelagic regime. Negotiated as a package deal,¹¹⁰ it was an inevitability that there would be winners and losers under the LOSC. The archipelagic regime was a ‘win’ for mid-oceanic archipelagic States and major maritime powers and a ‘loss’ for continental States with outlying archipelagos, though only in so far as archipelagic status is concerned. Such islands will of course enjoy the maritime zones to which Article 121 entitles them, and may even be tethered to the mainland through the drawing of straight baselines where consistent with Article 7. However, this binary characterisation of winners and losers does not fully capture the delicate balancing between archipelagic States’ interests and those of major maritime powers contained in the provisions of Part IV. It is this balance which is crucial to an understanding of how continental archipelagos were considered and excluded from the new archipelagic regime (so-called ‘negative regulation’).

This balance in Part IV has been vigilantly maintained, with attempts to disrupt it met with protest. So, for example, when the Philippines declared¹¹¹ the status of its archipelagic waters and this was met with objections by neighbouring States and maritime powers on the basis of its incompatibility with the Convention, the Philippines clarified its position.¹¹² Also, as has been seen, protests have been made regarding attempts by some continental States seemingly seeking to ‘simulate archipelagic status’ through the use of baselines (with a lack of clarity as to whether Article 7 or Article 47 is being invoked) with regard to their outlying archipelagos.

This lack of clarity and the protests by other States also have implications when ‘going beyond the LOSC’, where it is concluded that supervening customary law has not emerged (and even if it had, it would not automatically supersede the relevant treaty provisions but rather would give rise to, at most, a good faith obligation to reconsider the treaty provisions). Nor has subsequent practice been established reflecting the agreement (demonstrably absent) of the parties that the relevant provisions of the LOSC are to be interpreted as allowing their invocation by non-archipelagic States with offshore archipelagos. With the 1994 Implementation Agreement,¹¹³ the 1995 Fish Stocks Agreement amplifying, inter alia, Articles 63 and 64 of the LOSC,¹¹⁴ and the recently concluded internationally legally binding instrument to address biodiversity beyond national jurisdiction¹¹⁵—a genuine lacuna in the Convention¹¹⁶—some have argued that ‘the problem’ of offshore archipelagos should be addressed through a further implementing agreement or similar.¹¹⁷ As should be abundantly clear from the law and practice that has been explored in this article, the likely success of such is vanishingly small. This is not the first time, nor likely the last, that States have sought unsuccessfully either to re-open the package, or to recontour its shape, to suit particular national interests.