Islands, archipelagos and water: Insights from New Guinea

Most dictionaries agree that an island is defined as a piece of land, typically larger than a rock but smaller than a continent, that is surrounded by water, even at high tide. As definitions go, it is a rather strange one, akin to defining the human body as a series of different organs surrounded by skin.

We can quibble with the various parts of this definition, but it is the word ‘surrounding’ – by and large, a word that is, with its synonyms, and in all languages, ubiquitous in definitions of islands – that concerns us most in this paper. The word ‘surround’ suggests various consequences and implications about the land–sea dynamic that are partial at best, and even false. For example: –

that the sea is a hostile force, like an invading army or enemy, surrounding the vulnerable islanders on their island, besieging them and threatening their future;

It is ironic and confusing that the word ‘surround’ should be used to define the relationship between an(y) (is)land and water when, at the same time, the island is defined by the encircling water. Moreover, is water a benign force, a source of protein, the bedrock of the tourism industry, the originator and definer of so many aspects of island life; and yet a dark and evil entity, all out to destroy and overwhelm the (fragile) island and islanders it surrounds? We know that water – itself a medium that is hard to define – can be both these things, and even at the same time.

We should acknowledge that there are no hard edges between an(y) (is)land and its surrounding waters. Hard borders are only found on stylised maps, and all maps lie (Monmonier 2018). Hard borders are tricks of magnification. Hard borders are part of the territorial fantasy of states, whose governments and administrators purport to convince uneasy citizens that they maintain full control over what happens ‘inside’ their borders. A clear land–sea edge is a figment of the imagination, a convenient design, an act of legal fiction and an example of geographical falsehood. Ask the marine life that thrives at the intertidal zone; ask the island folk (usually women) who scavenge rock pools at low tide; look at the impact of erosion or accretion of beaches or cliffs: a long series of actions that involve the eliding of land and sea. And try circumambulating an island by walking strictly along its edge: quite an unfeasible task, since “it is impossible to precisely measure the length of any coastline” (Stoa 2019). Confronted by a sinking Venice, by a Majuro threatened by sea level rise, and by new islands arising out of the ocean in Japan, most islanders instinctively know that their reality is dialectically terraqueous (DeLoughrey 2016; Grancher & Serruys 2021). There is no such thing as land surrounded by sea. And so, there must be other, better, ways of defining an island other than as land and water, the latter operatively surrounding the former.

This problem of treating land and sea as distinct, and with land as the ascendant variable, recurs in international law. We can observe how different unfolding protocols send contradictory messages about the alleged durability of islands and their borders. On the one hand, there is the United Nations Convention on the Law of the Sea, which has led to a sanctioned territorialisation of the oceans, by means of which national sovereignty is projected outwards from the coast and its baselines over water. However, this stance is facing a series of challenges, resulting from the fact that shores, coasts and borders are not cast in stone; alas, even when they are, that stone can be eroded or submerged (eds. Kennedy & de Sousa Moreira 2022) – to the extent that some island states, such as Fiji and Seychelles, have resorted to baselines that are simply coordinates in the ocean (Grima 2024). The indeterminate distinctiveness of land and sea is stark on small islands and atolls where sea water intrusion can damage the water table (and compromise drinking water supplies), and where the ocean spray and salt can easily cross over the entire narrow land from ocean to lagoon (e.g. White & Falkland 2010).

On the other hand, there is solid evidence in favour of islands presenting hard geopolitical borders: islands make single jurisdictions. The number of inhabited island jurisdictions that are fragmented polities can be counted on two hands; and they are the exception that justifies the rule.

Let us review these oppositional dynamics below.

The United Nations Convention on the Law of the Sea (UNCLOS), which celebrated thirty years of coming into force in 2024, sought to deploy a hard and rigid approach to ocean governance via a territorialisation of ocean spaces:

This means that the range and extent of coastal, island and archipelagic states’ exclusive economic zones and other delineated spaces over water – often determined from baselines – depend crucially on their distance from discrete islands or rocks, and not just from mainland coasts. The problems arise when, given the active geo-dynamism of the ocean, such islands and rocks subside, erode or even disappear beneath the surface of the waves (while other islands may appear), causing havoc on charts, maps and other instruments that may have been deposited with the appropriate international bodies and accepted as national positions and claims by others, including neighbouring states. It is a tempting idea to tame the oceans and assume that they shapeshift as occasionally as land does. But this pursuit is a delusion:

Sea level change – which consists mainly in sea level rise; but note that it can also fall – is proving to be a powerful independent variable that could very well threaten the bedrock of the UNCLOS regime. Four small island [actually, archipelagic] states / large ocean states – Kiribati, Maldives, Marshall Islands and Tuvalu – are looking at stark ‘submarine futures’ and they have been pushing a conversation in recent decades that seeks to force an answer to the question: what happens when the complete territory of a sovereign state gets submerged? (Blue 2020; Kelman 2018; Pelling & Uitto 2001). Surely, the parameters of state borders and zoning need a sober reassessment?

This argument favours some fluidity, flexibility and humility, obliging a search for alternative marker regimes, such as fixed lines of longitude and latitude; or resorting to seamounts, rocks, islets and islands, irrespective of whether they are, or will remain, above water or not (Grima 2024).

At the same time, when discussing borders, it is quite clear that some of the most enduring borders are those that deal with island jurisdictions. The manner in which islands are surrounded by water means that their political confines and frontiers often align so perfectly with their geographic space. The likelihood of quarrels with neighbours over fronters and borders is vastly diminished because, frankly, there are no immediate neighbours.

This is where the archipelago variable becomes significant. Only five sovereign, populated island states are actually composed of single island units: Barbados, Cyprus, Dominica, Nauru and St Lucia. The remaining island states – thirty-seven, or thirty-eight if one counts Taiwan – are all archipelagic. This means that their territory is fragmented and dispersed; so, parts of it may be more liable to invasion and/or counterclaims than others. ⁽¹⁾ And, of course, various non-island but coastal states have island units. So, when territorial conflicts erupt, these may involve quibbles over particular small and unpopulated islands within an archipelago: as with the standoff between China (and Taiwan) and Japan over the Diaoyu Dao / Senkaku Islands; and the complex games being waged by multiple state actors in the South China Sea / West Philippine Sea (Baldacchino 2017; Raine 2017).

In any case, island states are spared from the possibility of neighbours crossing land borders and occupying and claiming part of their land: Ukraine, a non-island state, is an apt example. Such intrusions rarely happen with islands: crossing over a land border with an army is much easier, and can benefit from an element of surprise, while invading an island requires a navy apart from an army and needs to be more carefully planned, giving enough time to alert the islanders. Furthermore, if an enemy successfully invades a (small) island, what usually happens is that the complete territory falls into someone else’s hands (Royle 2001, pp. 57–8). Thus, the political integrity of the island as a single unit of governance is rarely affected. ⁽²⁾ As with the events of 1066 and their aftermath in Britain, only the ruler and the regime changes. And so, the borders of countries such as, say, the United Kingdom – a large island, with outlying small islands, plus a sizable chunk of the island of Ireland – have not changed much over recent centuries; and most of the recent changes, and tensions, have dealt with ‘the chunk’: the land border on the island of Ireland, between the Republic of Ireland (and the European Union) and Northern Ireland, which is part of the UK (and, so, out of the EU following the conclusion of the Brexit process). ⁽³⁾ Meanwhile the borders of France, Germany and Italy have changed multiple times; Poland’s borders changed five times in the 20th century alone. Let us take note that, out of some 85,000 populated islands in the world today, just ten are shared between two or more countries. And most of these ten borders are uneasy, having been subject to violence and armed conflict: these include Ireland, Cyprus, Tierra del Fuego and Timor Leste (Baldacchino 2013).

Thus, islands present themselves as controversial candidates. They have become significant gamechangers in light of the UNCLOS regime, exacerbating a rush by coastal and island states to consolidate their island possessions, not so much for the mineral resources that they hold but for the potential swathes of ocean space, and all that these may contain, above and below the seabed, that the islands provide title to. And yet, the very same UNCLOS regime is finding itself embarrassingly wanting in the face of geo-dynamism. Territorialising the ocean may require some radical fine-tuning. Meanwhile, island borders are especially rigid and stand the test of time. Island invasions are rare, successful island invasions are rarer; and, when the latter does happen, it usually means a total regime change, rather than splitting an island territory.

These arguments align with the notion of the ‘territorial trap’. In a seminal 1994 paper, Agnew argues that binding state to territory – a core tenet of mainstream international relations theory – was wrong and even delusional in at least three ways. The first is the manner in which the association between a state’s territory and state sovereignty both limits and legitimates that state’s sovereignty. Simply by ascribing a territory to a state makes that state appear ‘in control’, and to do so rightfully. Maps thus ‘contain’ and legitimate state power within borders, and they also purport to suggest that the state maintains full and complete control over those borders, and what happens within them. This is doubly not the case: states practise governmentality and venture to exercise control beyond their borders (for example, the projection of power by the United States and France around the world; or the FATF’s attempts to prevent local taxpayers from spiriting away their monetary deposits into offshore, low tax jurisdictions); while failed states lack the capacity to even govern their own territory (for example, Haiti, Mali, Somalia).

The second is to see the state as a single territorial actor in a unidimensional world that is interpreted only in terms of state–state competition or cooperation. This is a naïve and simplistic understanding of international relations, since so much sway and influence is exercised by non-state actors (such as MNCs, NGOs or regional organisations such as the European Commission), as well as different factions within states (such as the members of coalition governments, and their shifting negotiating clout).

The third and final explanation is to mistake the territorial state as a ‘container’ of society. Rather, society is socially constructed and reconfigured through the multiple and continuous actions of so many members, within, across and outside the state’s territory, and including the diaspora (Agnew 1994, 2015).

The triple tenets of the territorial trap are now applied in this paper to the island of New Guinea and its surrounding archipelago. The exercise (1) exposes the fallacy of cartography, especially of the straight-lined borders so enamoured by the partition exercises of the 19th and early 20th centuries; (2) identifies various cross-border movements – such as secessionism and shifting settlements, which puncture the aura of effective border management and permanence of states; and (3) foregrounds the agency of water, and how this is impacted by environmental change, which includes global warming and sea level rise, but also human-induced activity. The exercise serves to illustrate the deep impact that water, and its accompanying qualities, brings to bear on the management of the world’s second largest island, and the world’s largest shared island jurisdiction. Such observations offer a sharp but necessary contrast to the fixity and stability that are purported to effuse from straight lines on maps.

New Guinea (whose native names are Niugini, or Papua) is the largest of just eight populated islands in the world that are shared between two countries. ⁽⁴⁾ It is also unique in being the only island shared between two continents: its west, also known as Irian Jaya, being part of Indonesia and, therefore, Asia; its east, forming most of the territory of the country of Papua New Guinea, self-defined as a Melanesian nation, lying in Oceania. The Indonesia portion now contains six provinces. Papua New Guinea is divided into four regions and twenty-two province-level divisions. Five of these provinces cover the outer islands: Manus, New Britain (two provinces), New Ireland and Bougainville; the latter has special autonomy arrangements. The remaining seventeen provinces are located in the PNG part of the island of New Guinea.

Between 1884 and 1919, the island was divided between three European colonial powers: the Netherlands – at the time, governing the Dutch East Indies, which eventually achieved independence as Indonesia in 1945 and was handed over to the Indonesian authorities in the 1960s; Germany’s Kaiser-Wilhelmsland – to the north-east of what is today ‘mainland’ PNG; and British New Guinea – to the south-east. The W–E German–British boundary line broadly followed a mountain range; while the N–S boundary line sliced through the mountain range. Germany lost its PNG province as part of the settlement after the First World War. Straight lines were then drawn, evidence of how the evolution of maps is “only loosely grounded upon the lie of the land” (Ballard 1999). See Figure 1.

Figure 1.

The administrative division of the Island of New Guinea during 1884–1919: the Dutch West, the German North-East and the British South-East

Source: Wikimedia Commons 2024

The delineated frontier between Indonesia and Papua New Guinea on New Guinea was drawn up between Indonesia and Australia in 1973: at the time, the latter was responsible for the administration of PNG, before its independence in 1975. The border here consists of two unequivocal straight lines, running from north to south, slicing the island into two, mainly along Longitude 141⁰ East. But, what to do with a 60km stretch where there is a pesky and meandering river, around the centre of this straight line? This is the stretch where the parties agreed to stick to previous agreements, whereby the frontier shall follow the thalweg (the centre of the river course) of the Fly River. Hence, we have a combination of a hard border, specified by longitude and latitude, and a boundary, which is liable to change and shapeshift, as rivers are wont to do.

Thalwegs respond to the dynamics of river ecosystems. Thus, the Indonesia–PNG border changes here, occasionally forming oxbow lakes, and displacing the international border to and fro, in a non-linear and unpredictable manner. ⁽⁵⁾ Here, maps and satellite or aerial photographs, drawn or taken at a suitable magnification, and produced at regular intervals to capture the state of flow of the river, would be faithful to the river course, but at particular points in time. See Figure 2.

Figure 2.

Geographical map of contemporary New Guinea, showing the international land border between Indonesia and Papua New Guinea

Source: Freeworldmaps.net 2024

Meanwhile, the 750km-long hard border follows a line of longitude. But, despite its visually unproblematic straightness, it is just a long line symbolically drawn over mountains, valleys, swamps and generally challenging terrain. Meanwhile, on the ground, it is often impossible to determine whether one is in Irian Jaya, and so in one of the six New Guinea provinces of Indonesia, or in one of the seventeen mainland provinces of Papua New Guinea:

The border is poorly defined. Until the 1980s, there were only fourteen markers along its entire length. Since 2014, eighteen additional Transnational Border Posts have been built.

The border is a product of arbitrary decisions taken by past colonial regimes. No locals were consulted. The straight line on the map is clear evidence of this. Language groups and traditional rights to land, as well as relations of kin and of trade, extend seamlessly across the border. Border surveys during the 1960s established that it ran right through the middle of at least one village and that several villages that had been administered by the Dutch (that is, in what became Indonesia) were in fact in the Australian territory (which became PNG). As recently as 1980, a village included in Papua New Guinea’s National Census was found to be inside the Indonesian province of Irian Jaya [which, in 2000, Indonesian President Wahid renamed Papua]. The situation is made more complex for administering authorities by the tendency, among the shifting cultivators and hunter-gatherers who inhabit these areas, for whole villages to shift, re-form and disappear over time.

There are other reasons for border crossings. A small number of Irianese nationalists have sought political asylum in PNG. Some of them have been allowed to resettle in Papua New Guinea but, increasingly, in the 1980s, those granted refugee status were passed on to third countries. As a result of military activity in Irian Jaya, groups of Irianese villagers have crossed over into PNG from time to time, seeking temporary refuge, often with kin or extended families. And the OPM [Organisasi Papua Merdeka, or Free Papua Movement] guerrillas operating in the border area have occasionally crossed over into PNG, seeking refuge from Indonesian military patrols. Jakarta has accused Port Moresby of not devoting adequate resources to the task of ‘sanitising’ the border.

On the PNG side of the border, what development there has been – a little basic infrastructure (schools, aid posts, minor roads) – is largely the result of the attention the border area has received during tense periods of OPM-Indonesian military confrontation; whereas, on the Indonesian side, Jakarta has invested in highway construction, airstrips, health and education services, industrial and agricultural developments, and the establishment of trading centres to improve living conditions – thus accelerating the acculturation process that seeks to better incorporate the West Papuan population into the Indonesian nation-state (Gietzelt 1989). Nevertheless, recent flare-ups between West Papuans and Indonesian security forces, combined with steady international support for the West Papuan struggle, and the emergence of the separatist United Liberation Movement for West Papua (ULMWP), foreshadow a looming regional diplomatic situation, which may yet turn violent (Blades 2020; Taliawo 2022; Wangge & Lawson 2023).

Meanwhile, beyond the border on the island of New Guinea, the two neighbouring states have additional concerns that speak to the volatility and fluidity of their respective territories, caused by actions that involve water. For Indonesia, thousands of its small islands will be submerged, and millions of residents in coastal areas across Indonesia will be displaced, due to rising sea levels caused by climate change. By 2050, a third of the capital city Jakarta, described as the world’s fastest sinking city, could be submerged. Land subsidence, primarily caused by excessive groundwater extraction, damages infrastructure and buildings, and contributes to worsening flood events and tidal inundation (Colven 2020). A new capital city, Nusuntara, planned to be the home of seven million people, is taking shape, not just to take over from Jakarta but also to serve as a more central capital to the world’s most populated, sprawling archipelago. In the Indonesian part of New Guinea, the glaciers near Puncak Jaya, the highest island mountain peak on Earth, are in rapid retreat (ed. Bender 2019). The extensive coastal mangrove swamps of southern Irian Jaya have also been retreating as a result of rising sea levels (Ellison 1994). The significant coral reefs at Ayau Island are experiencing “massive” climate change-induced losses (Setiawati et al. 2024). The islands of Liki, Bepondi and Miossu, which serve to mark the extreme boundaries of Indonesia’s easternmost territory, are susceptible to an estimated sea level rise of around 40cm in the next hundred years (Wirasatriya et al. 2020).

For Papua New Guinea, the stark consequences of environmental change have been impacting several of its islands, such as Toruar (Fainu 2021) and the emblematic Carteret (Connell 2018), where flooding and coastal erosion have been the result not of a climate change-induced sea level rise but of tectonic changes, seismic events, ENSO-related tidal and storm surges, cyclones, wind-driven waves, and local actions. Meanwhile, Bougainville remains a reluctant member of the PNG state and has been clamouring for secession and eventual independence: a widespread popular sentiment energised after a convincing 2019 referendum result (Regan, Baker & Oppermann 2022).

The turbulence and turmoil of the sea is palpable. From melting glaciers and meandering rivers to advancing sea levels, coastal states struggle to mitigate, and increasingly adapt to, the effects of these major environmental changes, in which the sea is implicated. Island and archipelagic states are more dramatically and intimately confronted by such dynamics, and these help to lend credence to the naïve rendering of islands as defined by being ‘surrounded’ by water. UNCLOS has its benefits; but deciding to territorialise the oceans looks now, in retrospect, like wishing to make a problem go away … which it will not.

Oceanic geo-dynamism is a powerful force. It needs to be added to other, political forces – such as aspirations for secession, illegal migration, the actions of non-state actors and economic mobility – which collectively challenge the hard definitions of international borders, bolstered by state narratives and cartographic fictions. This is especially the case on islands which, in their visual representations on maps, evoke false understandings of being ‘natural’ jurisdictions, so neatly and totally surrounded by water.