1. Introduction

It is a truth universally acknowledged that a State with a coast can claim maritime zones. This principle has been famously stated by the International Court of Justice (ICJ) in its 1969 North Sea Continental shelf judgment, where the Court expressed the principle that ‘the land dominates the sea’.[1] Obviously, the land must have a coast: coastal land thus becomes the starting point for any claim to maritime areas; or, as the ICJ has stated ‘the attribution of maritime areas to the territory of a State, which, by its nature, is destined to be permanent, is a legal process based solely on the possession by the territory concerned of a coastline’.[2] The land territory and the maritime zone are therefore in a reciprocal relationship: if there is (coastal) land territory then there is the right to have maritime zones, and, conversely, there is no right to maritime zones without land territory.
While the second part of this relationship is pacifically accepted, the first part came to be questioned during the Third United Nations Conference on the Law of the Sea. When it became evident that claims to maritime zones beyond the territorial sea would ultimately be retained, States started realising that land – any land – was the key to gain access to significant marine resources, be they fish stocks or oil deposits. As a consequence, States began considering their land territory not for what it could provide itself, but for the maritime areas it could generate. Islands, in particular, had the potential to generate maritime zones that would greatly surpass the extension of the land territory of the island itself.
Increasing awareness of this fact led some States to question the capacity of islands to generate maritime zones. Eventually, States agreed on the adoption of what is today Article 121 United Nations Convention on the Law of the Sea (UNCLOS),[3] with the title ‘regime of islands’. This provision first confirms that all (natural) islands generate maritime zones[4] and then provides that ‘rocks which cannot sustain human habitation or economic life of their own shall have no exclusive economic zone or continental shelf’.[5] While Article 121 UNCLOS has often been invoked by States, in particular during dispute settlement proceedings, it was only in 2016 that its content formed the object of an in-depth judicial scrutiny, when the Annex VII Arbitral Tribunal in the South China Sea Arbitration issued its award on the merits of the case (South China Sea Award).[6]
The South China Sea Award has already formed the object of much scholarly commentary.[7] It is not the purpose of this brief article to restate the content of the Award or to analyse in depth its many interesting findings. Rather, it just purports to look into one specific aspect by examining how the relationship between land and sea is handled by the Arbitral Tribunal. To this end, the article will first briefly present the legal and historical context for the Award and will then analyse the Tribunal’s findings on how islands do (not) generate some maritime areas, with a view towards advancing some conclusions on the impact of the award.


1. The land dominates the sea?

Traditionally, the principle according to which ‘the land dominates the sea’ has been considered applicable to islands, as well as to the mainland.[8] The ICJ and other international tribunals addressing disputes between states have regularly recalled this principle also in cases where the maritime zones of islands were at issue.[9]
Notably, in the Qatar/Bahrain case, the ICJ expressly stated that ‘islands, regardless of their size, in this respect enjoy the same status, and therefore generate the same maritime rights, as other land territory’ and that Article 121(2) UNCLOS codifies customary international law.[10] This statement is the more important since, in that particular case, the ICJ was examining, among other issues, the nature of Qit’at Jaradah, a tiny feature located in the maritime space between the parties.[11]
This broad invocation of the principle that ‘the land dominates the sea’ also in the case of islands (and, to echo the ICJ, ‘regardless of their size’) calls for two considerations. In the first place, it would suggest a certain relationship between the second and third paragraph of Article 121 UNCLOS. Since it is a well-established principle that islands do have full entitlement to maritime areas, any rule that would suggest that (at least some) islands have less or no entitlement should be considered an exception. Accordingly, first, such rule should be construed narrowly, and, second, it should be up to the party that invokes this rule to prove that the conditions on the ground actually mandate its application.
In the second place, a more general consideration relates to the object of the dispute which has formed the basis for the decisions mentioning the principle that ‘the land dominates the sea’. These decisions – from the 1969 North Sea Continental Shelf case up to the 2014 Bangladesh/India case – were all adopted following a request to delimit maritime boundaries. The determination of the nature of one or more features and their entitlement to maritime zones, therefore, was not to be the outcome of the case, but just one step in the structured process that leads to the determination of a maritime boundary.
This has a significant practical consequence. An initial, more generous, acceptance of the principle that even tiny islets are entitled to all maritime zones can then be checked when determining the special/relevant circumstances that may have a bearing on the final course of the boundary.[12]
This is, indeed, what happened in the Qatar/Bahrain case mentioned above. After recalling the general principle, the ICJ went on to examine the nature of Qit’at Jaradah and determined that it was an ‘island’:

‘The Court recalls that the legal definition of an island is ‘a naturally formed area of land, surrounded by water, which is above water at high tide’ (1958 Convention on the Territorial Sea and Contiguous Zone, Art. 10, para. 1; 1982 Convention on the Law of the Sea, Art. 121, para. 1). The Court has carefully analysed the evidence submitted by the Parties and weighed the conclusions of the experts referred to above, in particular the fact that the experts appointed by Qatar did not themselves maintain that it was scientifically proven that Qit’at Jaradah is a low-tide elevation. On these bases, the Court concludes that the maritime feature of Qit’at Jaradah satisfies the above-mentioned criteria and that it is an island which should as such be taken into consideration for the drawing of the equidistance line.’[13]

The Court thus seemed to suggest that Qit’at Jaradah would be given full effect in generating maritime zones. However, the final outcome is quite different, as eventually Qit’at Jaradah did not even have effect for the delimitation of the territorial sea. In fact, according to the Court,

‘Qit’at Jaradah is a very small island, uninhabited and without any vegetation. This tiny island … situated about midway between the main island of Bahrain and the Qatar peninsula. Consequently, if its low-water line were to be used for determining a basepoint in the construction of the equidistance line, and this line taken as the delimitation line, a disproportionate effect would be given to an insignificant maritime feature … The Court thus finds that there is a special circumstance in this case warranting the choice of a delimitation line passing immediately to the east of Qit’at Jaradah.’[14]

The same methodology, which initially acknowledges full entitlement of an island but successively limits this effect when the provisional equidistance line is corrected in light of the special circumstances of the case, was adopted in the other cases mentioned above. An instance is provided in Romania/Ukraine, where the ICJ disregarded the small, uninhabited Serpent’s Island when drawing the boundary between the exclusive economic zone and continental shelves of the two States[15] without having to decide whether this feature was an ‘island’ or a ‘rock’.[16]
In some cases, the final outcome has even been the opposite from what would stem from the application of the principle that ‘the land dominates the sea’, when international judges have entirely disregarded sizeable, inhabited islands. For example, the International Tribunal for the Law of the Sea decided to give no effect to St. Martin’s Island, which has ‘a surface area of some 8 square kilometres … sustains a permanent population of about 7,000 people … [and] receives more than 360,000 tourists every year’[17] on the rather summary consideration that otherwise this ‘would result in a line blocking the seaward projection from Myanmar’s coast in a manner that would cause an unwarranted distortion of the delimitation line’.[18]
In conclusion, when the entitlement of islands to maritime zones is discussed in the context of a case involving the delimitation of maritime boundaries, there seems to be a general presumption that – in principle – any island will generate not only its own territorial sea, but also its exclusive economic zone and continental shelf (although this entitlement may have no bearing – in practice – on the final course of the maritime boundary).
Partial or no generation of maritime areas, therefore, is accepted only as a practical consequence of the application of the ‘equitable solution’ requirement of Articles 74(1) and 83(1) UNCLOS to the specific circumstances of the case. Construing lack of entitlement as a practical, case-specific consequence, therefore, strengthens, rather than weakening, the general principle according to which land – including islands – generate maritime zones.


3. The sea dominates the land?

In contrast to the cases mentioned so far, the South China Sea arbitration was the first case in which an international tribunal was called upon to determine the nature of certain maritime features, without at the same time being entrusted with the task of drawing the maritime boundary between the parties to the dispute. This situation left no alternative to the Arbitral Tribunal but to consider the merits of Article 121 UNCLOS itself and has led to a careful, in-depth assessment, which makes use of the different methods provided by Articles 31 and 32 of the Vienna Convention on the Law of Treaties in order to interpret this provision and clarify its content.[19]
As has already been noted by commentators, the Arbitral Tribunal eventually set a high threshold for evaluating whether a certain feature is an ‘island’ or just a ‘rock’. While it has accepted that the assessment should concern the ‘objective capacity of the feature to sustain human habitation or economic life’, the Tribunal has also added the evaluation of both ‘time’ and ‘qualitative’ elements, thus requiring that ‘[h]abitation and economic life must be able to extend over a certain duration and occur to an adequate standard’.[20]
Furthermore, although the Tribunal openly rejects historical evidence as decisive,[21] it nonetheless puts a great emphasis on the historical record. Notably, it considers that ‘evidence of physical conditions will ordinarily suffice only to classify features that clearly fall within one category or the other’[22] while in borderline cases ‘the most reliable evidence of the capacity of a feature will usually be the historical use to which it has been put’, in particular ‘evidence … that predates the creation of the exclusive economic zones’.[23]
It is not to be wondered that the Arbitral Tribunal in the South China Sea does not recall the principle that ‘the land dominates the sea’. On the opposite, the impression is that in the South China Sea, the sea dominates the land. Rather than looking at the land in the case and determining its entitlement to maritime areas, the Tribunal seems to consider what maritime areas could potentially be claimed, and whether such claims would be acceptable, in order to determine whether a feature is a rock or an island.
The Tribunal openly discloses the reasons at the basis of this restrictive approach. It considers that Article 121(3) UNCLOS ‘is a provision of limitation’ which was adopted with ‘the object and purpose of preventing encroachment on the international seabed reserved for the common heritage of mankind and of avoiding the inequitable distribution of maritime spaces under national jurisdiction’.[24] Significantly, it adds that:

‘[t]he introduction of the exclusive economic zone was not intended to grant extensive maritime entitlements to small features whose historical contribution to human settlement is as slight as that. Nor was the exclusive economic zone intended to encourage States to establish artificial populations in the hope of making expansive claims, precisely what has now occurred in the South China Sea. On the contrary, Article 121(3) was intended to prevent such developments and to forestall a provocative and counterproductive effort to manufacture entitlements’.[25]

One can certainly feel very sympathetic with the Arbitral Tribunal’s last remark, in particular if one considers the extent to which States are going to ‘manufacture’ entitlements, in the South China Sea as well as in other parts of the world’s oceans and the negative impact these activities have on international peace and security.
All the same, the statement of the Tribunal is not proof against criticism. Two aspects will be mentioned here. First, the Tribunal always refers to the exclusive economic zone alone; nonetheless, Article 121(3) excludes the entitlement of rocks to both the exclusive economic zone and the continental shelf. One may only surmise why the continental shelf is never mentioned, but this reference to one part only of the final part of Article 121(3) appears to weaken the arguments advanced in the award.
Second, the emphasis on the historical record – evidenced, among others, by the reference to the features’ ‘historical contribution to human settlement’ – generates the implied presumption that an island which has not been inhabited in the past will not be inhabited in the future. Historical evidence is certainly determinative of the capacity of an island to sustain human habitation. The opposite might not however be always true, as islands that had never been inhabited before started hosting human groups only later in time. One may actually wonder whether the Tribunal’s interpretation may not achieve the result of ‘freezing’ the relevant rules in the past. This feeling is strengthened if one takes into account the outright rejection of technology[26] and the marginalisation of the role that successive State practice could play in the interpretation and the evolution of the content of the UNCLOS, including Article 121.[27]


4. Concluding remarks

The arbitral award in the South China Sea case has provided the first occasion for an extensive discussion of Article 121 UNCLOS, and in particular for determining under what conditions dry land emerging during high tide can be considered as an ‘island’ or a ‘rock’. This is a much needed clarification, coming at a time when States, in an effort to consolidate maritime claims, generate situations that pose a threat to international peace and security.
Reactions by States have been varying and, in any case, it is necessary to keep in mind that the award is binding only on the States parties to the case. One element of international practice that might be relevant is the 2018 ICJ judgment in the Costa Rica/Nicaragua case. In this judgment, the ICJ had to determine, among other things, whether Nicaragua’s Corn Islands generated maritime zones. The Court briefly noted that these islands ‘have a significant number of inhabitants and sustain economic life’ before concluding that they ‘amply satisfy the requirements’ of Article 121(2) UNCLOS.[28] The Court did neither mention the principle ‘the land dominates the sea’ nor the South China Sea award analysis.[29]
It is yet early to determine the impact of the award on the traditional principle that ‘the land dominates the sea’ and its application to islands. The detailed analysis of this award is certainly likely to be referred to by scholars addressing the issue of islands and rocks. At the same time, the analysis might eventually prove to be more tied to the current geopolitical characteristics of the South China Sea than the award would appear to concede, and therefore more difficult to apply to other contexts.



* Associate Professor of International Law, University of Milano-Bicocca.

[1] North Sea Continental Shelf (Federal Republic of Germany v Denmark) (Merits) [1969] ICJ Rep 3 para 96.

[2] Maritime Delimitation in the Area between Greenland and Jan Mayen (Denmark v Norway) (Merits) [1993] ICJ Rep 38 para 80; see also Aegean Sea Continental Shelf (Greece v Turkey) (Jurisdiction of the Court) [1978] ICJ Rep 3 para 86, according to which rights over maritime zones ‘are legally both an emanation from and an automatic adjunct of the territorial sovereignty of the coastal State’.

[3] United Nations Convention on the Law of the Sea (adopted 10 December 1982, entered into force 16 November 1994) 1833 UNTS 397 (UNCLOS).

[4] Art 121(2) UNCLOS.

[5] Art 121(3) UNCLOS.

[6] South China Sea Arbitration (The Republic of the Philippines v The People’s Republic of China) PCA case no 2013-19 (Award of 12 July 2016).

[7] In addition to W Gullett, ‘The South China Sea Arbitration’s Contribution to the Concept of Juridical Islands’ in this issue, see also S Faccio, ‘The South China Sea Arbitration Award of July 12, 2016: The Unbearable Lightness of Being a Rock’ (2017) 72 La Comunità Internazionale 623; JH Paik, ‘South China Sea Arbitral Awards: Main Findings and Assessment’ (2017) 20 Max Planck YB United Nations L 367; Y Tanaka, ‘Reflections on the Interpretation and Application of Article 121(3) in the South China Sea Arbitration (Merits)’ (2017) 48 Ocean Development Intl L 365; M Loja, ‘The Spratly Islands as a Single Unit Under International Law: A Commentary on the Final Award in Philippines/China Arbitration’ (2016) 47 Ocean Development Intl L 309; N Oral, ‘“Rocks” or “Islands”? Sailing Towards Legal Clarity in the Turbulent South China Sea’ (2016) 110 AJIL Unbound 279; AO Elferink, ‘The South China Sea Arbitration’s Interpretation of Article 121(3) of the LOSC: A Disquieting First’ (2016) The JCLOS Blog <http://site.uit.no/jclos/2016/09/07/the-south-china-sea-arbitrations-interpretation-of-article-1213-of-the-losc-a-disquieting-first>.

[8] For example, in Libya/Malta, both parties to the case ‘agree[d] that the entitlement to continental shelf is the same for an island as for mainland’ with Libya only arguing that ‘while the entitlement is the same, an island may be treated in a particular way in the actual delimitation’ (Continental Shelf (Libyan Arab Jamahiriya v Malta) (Merits) [1985] ICJ Rep 13 para 52). See also Jan Mayen (n 2) para 80, according to which Denmark, while referring to the language of art 121 UNCLOS, did not challenge the entitlement of the Norwegian island of Jan Mayen to (residual) maritime zones beyond the territorial sea.

[9] See for example Delimitation of the Maritime Boundary in the Gulf of Maine Area (Canada v United States of America) (Merits) [1984] ICJ Rep 246 para 157; Aegean Sea (n 2) para 86; Maritime Delimitation and Territorial Questions between Qatar and Bahrain (Qatar v Bahrain) (Merits) [2001] ICJ Rep 97 para. 185; Territorial and Maritime Dispute between Nicaragua and Honduras in the Caribbean Sea (Nicaragua v Honduras) (Merits) [2007] ICJ Rep 659 paras 113 and 126; Maritime Delimitation in the Black Sea (Romania v Ukraine) (Merits) [2009] ICJ Rep 61 para 77; The Bay of Bengal Maritime Boundary Arbitration (Bangladesh v India) (PCA, Award of 7 July 2014) para 279; Continental Shelf (Tunisia/Libyan Arab Jamahiriya) (Merits) [1982] ICJ Rep 18 para 73; Dispute concerning delimitation of the maritime boundary between Bangladesh and Myanmar in the Bay of Bengal (Bangladesh v Myanmar) (ITLOS case no 16, Judgment of 14 March 2012) para 185.

[10] Qatar/Bahrain (n 9) para 185, recalled in Nicaragua/Honduras (n 9) para 113.

[11] According to the ICJ’s description, ‘at high tide its length and breadth are about 12 by 4 metres, whereas at low tide they are 600 and 75 metres. At high tide, its altitude is approximately 0.4 metres’ (Qatar/Bahrain (n 9) para 197).

[12] Following the well-established practice that delimitation first requires the drawing of a provisional equidistance line and then consideration of whether special/relevant circumstances requiring an adjustment of the line exist; see, for the most recent statement of this procedure, Maritime Delimitation in the Caribbean Sea and the Pacific Ocean (Costa Rica v Nicaragua) and Land Boundary in the Northern Part of Isla Portillos (Costa Rica v Nicaragua) (ICJ Judgment of 2 February 2018, nyr) para 98 (for the territorial sea) and para 135 (for the exclusive economic zone and continental shelf, where a third step is also provided).

[13] Qatar/Bahrain (n 9) para 195.

[14] ibid para 219.

[15] Romania/Ukraine (n 9) para 185, where it is noted that ‘the Court may on occasion decide not to take account of very small islands or decide not to give them their full potential entitlement to maritime zones, should such an approach have a disproportionate effect on the delimitation line under consideration’.

[16] ibid para 187.

[17] Bangladesh/Myanmar (n 9) para 143.

[18] ibid para 318.

[19] South China Sea Award (n 6) paras 473-553.

[20] ibid para 504. The application of this standard seems to raise it even more than its abstract expression might indicate; for example, according to the Tribunal, ‘the criterion of human habitation is not met by the temporary inhabitation of the Spratly Islands by fishermen, even for extended periods’ (ibid para 618, emphasis added).

[21] ibid para 483.

[22] ibid para 548.

[23] ibid para 549.

[24] ibid para 535.

[25] ibid para 621.

[26] ibid para 511. For some critical comments see the article by Gullett in this issue.

[27] Compare for example the lengthy discussion of the travaux preparatoires in South China Sea award, paras 521-538, with the quick disposal of State practice in South China Sea award (n 6) para 553. The different treatment becomes even more significant if one considers that while successive State practice forms part of the general rule on interpretation (art 31(3)(b) VCLT), the travaux preparatoires are relevant only as supplementary means of interpretation (art 32 VCLT).

[28] Costa Rica/Nicaragua (n 12) para 140.

[29] This might however be due to the traditional reticence of the ICJ to refer to the jurisprudence of other courts.