Introduction: Kashmir as a debatable land

‘The Debatable Land’ was a term first coined in the medieval period to describe the borderland between the kingdoms of England and Scotland. Over the centuries it passed back and forth between the two realms without any formal line of demarcation being established. Its inhabitants variously identified themselves as English or as Scots (but also simply as Border people), and a rough-and-ready system of law – the Law of the Marches – emerged to resolve their disputes. Although a clear borderline was finally established in 1552, it only served as an international frontier for about 50 years, until the two kingdoms were united under James VI and I. To this day, the people of the Anglo-Scottish borders retain a distinct identity and a sense of divided loyalty, recently expressed in a short essay on the implications of Scottish independence by the Conservative MP for Penrith and The Border, Rory Stewart (Stewart Reference Stewart2012).

Taking the original Debatable Land as a prototype, I shall understand debatable lands in general to be territories that have the following features in common. First, they stand between at least two rival nation-states, each of which takes its ‘homeland’ to include all or part of the disputed territory. Second, the inhabitants of the debatable land identify themselves in different, more or less complex, ways, with some typically bearing allegiance wholly or partly to nation A, and others likewise to nation B, etc. Third, the debatable land itself functions as a social unit and a source of identity, although its occupants may also find themselves in violent conflict with one another from time to time. This identity is likely to reflect the fact that in some past era the debatable land formed an independent political unit, even if not a fully fledged state. Given this background, we may expect to find some of those who live there wanting the territory to be wholly or partly absorbed by A, others wanting it affiliated to B, and yet others preferring it to become largely independent of both.

Debatable lands, therefore, are a sub-class of disputed territories. Even though secessionist movements, for example, will always make territorial claims that the state they are trying to leave is likely to dispute, the lands in question may or may not be debatable according to the above criteria. I single them out for discussion because, as will shortly become apparent, they pose particular difficulties for nationalist theories of territory. Moreover, there is no shortage of cases, past and present, that qualify as debatable lands (regions that were once debatable may cease to be so if their inhabitants’ identities align over time with one or other of the adjoining nations). Thus Alsace-Lorraine, South Tyrol, Transylvania, Bosnia, and Kurdistan are all examples of this phenomenon (for further examples outside of Europe, see Anderson Reference Anderson1996). However, the contemporary case that I will focus on here is Kashmir, which exemplifies the intractable problems posed by debatable lands in their starkest form (in what follows I have relied on Malik Reference Malik2002; Bose Reference Bose2003; Zutshi Reference Zutshi2004; Bose Reference Bose2007; Kaul Reference Kaul2011; Chandhoke Reference Chandhoke2012).

To sketch in the main features of this case, Kashmir (or to give it its full title ‘Jammu and Kashmir’) is territory claimed by both India and Pakistan, and currently divided between them along a militarized Line Of Control (LOC) that has been in place since 1972. It had originally been a princely state, whose Hindu ruler opted to accede to India when the two countries split in 1947. Its population, however, was 77% Muslim, and military conflict immediately erupted between India and Pakistan, leading to de facto partition in 1949.Footnote ⁵ Since then, neither the inhabitants of Jammu and Kashmir themselves nor the two national governments have been able to agree on a resolution to the dispute, which has led to ongoing episodes of terrorist violence in Indian-held Kashmir especially. From the Pakistani side, Kashmir is regarded as the country’s ‘jugular vein’, while Indians regard it as an ‘integral part’ of Indian territory as whole. Opinions inside the region reflect the fact that it is an ethnic patchwork, with Sunni and Shia Muslims, Hindus, Buddhists, and Sikhs all forming majorities in different areas, so there is support for inclusion in India, for inclusion in Pakistan, and for an independent Kashmir, as well as for more complex solutions that involve dividing the territory along ethno-religious lines. Thus it fully meets the criteria set out above for debatable status – two contending nation-states with territorial claims, a previous history of semi-autonomy as a political unit, and a population with divided loyalties, alongside an independent sense of themselves as Kashmiris – together with the added complication of deep religious and linguistic divisions, which were not present in the original Debatable Land (see Malik Reference Malik2002: Chs 1, 2).

Liberal nationalism and territorial rights

I approach the problems posed by debatable lands from a liberal nationalist perspective, the main pillars of which are taken for granted here. That is, I assume that nations, as intergenerational communities of people who identify with each other and with their forebears, are real, and also that there are sound reasons, of justice and democracy, for states to have a national basis wherever possible; I also take for granted that the system of states is a defensible arrangement. Equally, I acknowledge that nation-states are rarely, if ever culturally, homogenous, and that their political arrangements should respond to this fact, for example by implementing policies that grant recognition to cultural minorities, and by devolving power to territorially concentrated sub-groups within their borders. From this perspective debatable lands are a hard case. They challenge the assumption that the world can be organized into national units with external and internal boundaries that can be defended on normative grounds. To understand the problem I need first to sketch in the nationalist theory of territory to see how far it can deal with the issue of territorial disputes in general (I have defended it in Miller Reference Miller1998; Miller Reference Miller2007: Ch. 8; and most fully in Miller Reference Miller2012; other important defences include Moore Reference Moore2001: Ch. 7 and Meisels Reference Meisels2009).

The theory holds that states can claim determinate territorial rights only as representatives of peoples who are their primary possessors. To qualify as a ‘people’ in the relevant sense, a group must be culturally integrated and continuous over time, and in the modern world the main examples of such groups are nations. They acquire territorial rights by occupying the land in question and endowing it with material and symbolic value – material value in the form of improvements (cultivated land, buildings, transport systems, etc.) that make the territory better able to support human life, symbolic value in the form of the meaning that is attached to places in the territory as a result of the group’s culture, and historical memory. If two states dispute rights to a piece of territory, therefore, the nationalist theory invites us to ask, first, who each state can claim to represent, and then which people or peoples have added most value to the land in either of the ways outlined above. In standard cases, territorial rights should be awarded to the state that represents the people whose value claim is strongest.Footnote ⁶

In many cases the nationalist theory will yield a reasonably clear-cut solution to such disputes, so provided its premises – the importance of land to the people who inhabit it over time, and the collective responsibility they have to improve and maintain it – are accepted, it can serve as a valuable tool in political debate and international law. Thus, a state that merely takes command of land against the wishes of its current inhabitants, or (again without consent) implants its own citizens so that they eventually outnumber the natives, cannot by doing so acquire territorial rights, even if it performs normal state functions such as maintaining law and order in the territory. If a community of people have occupied the land over time and added value to it, only a state that represents them legitimately can claim such rights. Equally, the nationalist theory discounts what might be called ‘purely historical’ claims, where states assert territorial rights simply on the grounds they have exercised sovereignty over the area at some point in the past.Footnote ⁷ To defeat the rival claims of present occupants, it must be shown that the state represents people who in the past have imbued the land with greater value. The Argentinian claim to the Falklands/Malvinas, resting as it does on a papal grant and short-lived colonization far in the past, is of this historical kind. It is therefore, in my view, defeated by the will of the present inhabitants, who have created almost all of the infrastructure that makes the islands fit for human habitation, to remain associated with Britain.Footnote ⁸ If instead Argentinian settlers had occupied the islands for a considerable stretch of time, building houses, establishing farms and so both – and had then been displaced – the nationalist theory would now tell in favour of Argentina.

A more complicated case, however, is a secessionist claim made by a group within the state that over time has engaged in value-adding activities of its own. For what is then at issue is the identity of the relevant group for purposes of the nationalist theory. Are the would-be secessionists a separate nation who have imprinted the territory they are now claiming with their own distinctive culture; or do they have a long history of interaction and cultural interchange with their neighbours, such that there also exists a larger nation of which they form a component? For example, as I have argued elsewhere (Miller Reference Miller2000), it is not clear that peoples such as the Scots or the Catalans can claim full rights in the territories they occupy to the exclusion of the larger nations to which their members also belong. The historical development of these territories has been a collaborative process, even though they may hold special symbolic significance for the smaller nations. Here, then, the nationalist theory does not deliver an unequivocal verdict in favour of either the existing state or the new state that the secessionists aim to establish. Instead, it acknowledges both claims and recommends a division of territorial rights, through devolution or a federal arrangement, that respects both the smaller nation and the larger nation within which it is ‘nested’.

When we come to debatable lands proper, though, it may seem that the nationalist theory is of no use at all. For what seems to be missing here is a collective of the right kind to be a holder of territorial rights. There is no ‘nation’ in the debatable land, not even a minority nation that forms part of something larger. Instead, we find intermingled populations who variously identify themselves nationally with either of the neighbouring states involved in the dispute while also feeling that they belong to a separate community. In the case of Kashmir, for example, some residents will regard India as their larger homeland, others will think the same about Pakistan, while yet others will regard Kashmir itself as their main home.Footnote ⁹ So, if we try to apply the arguments outlined above, which are meant to tell us who holds the stronger claim to disputed territory, we immediately run into difficulties. Consider the symbolic value argument first. This invests groups with territorial rights in places that for religious, cultural, or historic reasons have come to acquire a special significance as features of the group’s homeland. But in the case of debatable lands, the identity of the homeland is disputed. As noted above, both Indian-identified Kashmiris and Indians outside of Kashmir regard that territory as an integral part of India. This is partly a matter of the physical beauties of the Kashmir valley (lyrically celebrated by Nehru among others: see Kaul Reference Kaul2011, 10) and partly because, as the only Muslim-majority Indian state, Kashmir symbolized the idea of India as a secular regime within which different religious identities could peacefully co-exist. On the Pakistani side we find an equally strong mixture of historical celebration (e.g. a poetic literature in Persian dating back to the fourteenth century: see Zutshi Reference Zutshi2004: Ch. 1; Kaul Reference Kaul2011: Ch. 2) and contemporary identity considerations. Since the underlying rationale for Pakistan’s existence in the first place was to provide a gathering point for the Muslim communities on the sub-continent, it appears as an anomaly if a region whose population is more than three-quarters Muslim is not included within it. If on the other hand we consider the religious dimension of symbolic identification, we find that Kashmir contains both Hindu and Muslim shrines – in fact some sites are recognized as being sacred to both religions (see Malik Reference Malik2002: Ch. 1). So the argument again points us in opposite directions.

The argument from material value is no more decisive. Assuming that there has been no large-scale immigration into or emigration from a debatable land, the value embodied in the land will be the work of the generations from which the people now living there are descended. But how are these generations to be characterized from the point of view of national identity? As we have seen, they are likely to be the bearers of rival identities and/or the identity of the debatable land itself. There is no single cohesive group that we can credit with developing the land. Nor is it likely that credit is owed to one or other of the nations on whose border the land sits. Because of the marginal status of debatable lands, they are unlikely to enjoy high levels of investment from the centre. In the case of the larger, Indian-held section of Kashmir, for example, expenditure on the military has dominated industrial development in the planning policies of the Indian government, and the region remains one of the poorest in India, with high levels of unemployment (Prakash Reference Prakash2000).

So we appear to have reached an impasse. The nationalist theory of territorial rights, which gives clear-cut answers to some territorial disputes, and at least substantial guidance in the case of others (such as secessionist claims), appears to be of no help in the case of debatable lands. But these are not relatively trivial cases, like the uninhabited islands referred to above. The governance of debatable lands is a matter of some urgency. Living in such territories was never particularly comfortable, as an examination of the original Debatable Land reveals, but in a world in which the territorial state has become the dominant mode of political organization, being caught in limbo between several states is likely to mean suffering a condition of open or latent conflict, indeed of war in Hobbes’ sense.Footnote ¹⁰ Nor does the Hobbesian solution to the problem – assign territorial rights to whichever power is able to impose order by means of the sword – seem plausible in the face of the conflicting political identities at stake. What other solutions might we contemplate?

Plebiscites and partitions

There are two obvious possibilities. One is to hold a plebiscite within the debatable land to allow its inhabitants to determine their own future. The other is to partition the land between the contending states along some agreed-upon line. Would liberal nationalist principles support either of these?

At first sight, the plebiscitary solution might seem to be a way of respecting the idea of territorial self-determination: the people on the ground, or at least a majority of them, will decide what their own future status should be. But on closer inspection, difficulties and objections quickly multiply. The first is that the decision to be taken is not a binary one. Even if we leave aside more complex arrangements (I will return to them later), there are at least three simple outcomes to choose between: incorporation into state A, incorporation into state B, and independence. We can presume that each inhabitant will rank order these options in a certain way, and although some orderings may be more likely than others, we have no reason to exclude any of the resulting six possibilities. So what form will the plebiscite take? If each voter is asked just to nominate their first preference, then a lot of relevant information is being discarded. For instance, it may be that incorporation into state A gets most first preferences, but is bitterly opposed by a significant minority, whereas independence would at least be regarded as acceptable by the vast majority of inhabitants. In that case a plebiscite will fail to produce a workable outcome. Of course, a more elaborate voting system (such as a Borda count) might be used instead, but as social choice theory has taught us, there is no ‘correct’ procedure for reaching decisions in cases such as this (in the sense of one that is guaranteed to meet some quite minimal conditions of collective rationality: see Arrow Reference Arrow1967).

Perhaps a bigger problem still concerns the constituency within which the plebiscite is to be held. Let us grant for the sake of argument that the debatable land has clearly identifiable borders of its own (unlike the original Debatable Land). Then the plebiscite would presumably be offered to all those currently residing within those borders. But this would exclude non-residents who might have a strong interest in the outcome – people in the neighbouring states who either for reasons of identity or because they would wish to return to the territory in the event of a successful (from their point of view) outcome have a claim to be included. It would also not deal with divisions within the debatable land itself – the existence of specific locally based communities who might almost unanimously prefer one or other solution. Kashmir can again serve as our example here. At the time of partition India and Pakistan both announced that they were willing to allow the future status of Jammu and Kashmir to be determined by a plebiscite carried out under UN supervision, but India has subsequently retreated from this position, and Pakistan’s continued formal support for a referendum does not extend to allowing ‘independence’ as a third option on the ballot.Footnote ¹¹ Support for a three-way plebiscite is thus confined to pro-independence parties within Kashmir. But how might this work out in practice? According to Bose:

The plebiscite is a democratic device, and as such it cannot escape the general problem that afflicts all democratic proposals, namely that of finding the appropriate ‘domain’ in which to implement the proposal (for discussion see Whelan Reference Whelan1983; Arrhenius Reference Arrhenius2005; Goodin Reference Goodin2007; Miller Reference Miller2009; Näsström Reference Näsström2011). The normative force that a democratic procedure, such as voting in a referendum, carries depends on that problem having been resolved satisfactorily. But it does not seem that a debatable land forms a suitable domain for democracy, at least in its simple, majoritarian form: its inhabitants, fractured in the way that Bose describes in the Kashmiri case, do not qualify as a demos whose collective decisions about its future status would carry moral authority for its individual members. It would only appear as such to those who already believed in the idea of Kashmiriyat, that is believed that Kashmiris already formed a distinct nation with its own identity and culture (see Malik Reference Malik2002: Ch. 1; for a more critical line on ‘Kashmiriyat’, see Zutshi Reference Zutshi2004, Introduction). But this of course is precisely what is at issue, here and in other debatable lands.Footnote ¹²

Suppose then we look in the other direction, towards partition. The debatable land would be divided between the two nations involved (or between all if there are more than two) and then political debate within each newly created unit would decide on the precise form of incorporation – whether the territory would become simply an administrative district within a unitary state, or whether it would enjoy partial autonomy with its own legislature, and so forth. The presumption here is that the two territories would be contiguous with the nations they are going to join, so that at the end of the carving-up we would have two states with continuous borders. This by itself will not uniquely determine the line of partition, however. There are then two main possibilities: the ‘status quo’ solution where the dividing line follows an existing line of political control, such as the military LOC in the Kashmir case, or the line of truce where a military conflict has ended, and the ‘ideal’ solution where questions of identity, allegiance, and preference are introduced in order to determine the division that mostly closely reflects the profile of the inhabitants of each section (or possibly other factors, such as symbolic or historical claims: different normative principles may produce different outcomes).

Setting aside the line of partition issue for a moment, how does this solution look in general if we are following nationalist principles? It has two obvious merits. We end up with consolidated territories within which national self-determination can be practised, and neither nation need feel that it has had to abandon its territorial claim entirely. On the other hand, the outcome is clearly a compromise: part of what may be seen as the historic ‘homeland’ has been consigned, irretrievably for practical purposes, to a nation that is likely to be seen as ‘the other’ if not ‘the enemy’. The section that is given up may include sites of particular symbolic importance even if they are not now inhabited mainly by fellow nationals. Equally important, partition does nothing to meet the aspirations of all those whose primary identity is the debatable land itself – those who believe that they form a separate nation that transcends the ethnic or religious divisions within the territory. They are likely in fact to regard partition as nothing short of a disaster, since it appears to frustrate their nation-building aims in perpetuity. So it cannot be regarded as a satisfactory answer by anyone who subscribes to the principle of nationality, and at best it might provide a pragmatic solution to a conflict that was threatening to turn into civil war.

There is also the serious issue of how the line of division should be settled. If an existing administrative or military boundary is used, there is no guarantee that this will place a majority of the inhabitants within their preferred state. In the case of Kashmir, for example, if partition followed the current LOC, India would be awarded a territory whose population is about two-thirds Muslim, most of whom, it is reasonable to assume, would prefer to belong to Pakistan if that were the only other choice open to them. So is there a better approach? We cannot assume that the population of a debatable land slowly and steadily changes its cultural complexion as one moves from the territory of state A towards the territory of state B, so that a mid-way line can easily be drawn. In Transylvania, for example, ethnic Hungarians inside Romania live not only in areas adjoining present-day Hungary itself, but also in enclaves further to the east, forming majorities in some parts. Returning to our main example, Kashmir, a good number of partitionist plans have been proposed, some simply redrawing the existing boundary between India and Pakistan, others involving the creation of an independent Kashmir alongside areas attached to the two large states, yet others attempting still more intricate solutions in an attempt to capture the complex ethnic and religious make-up of the territory (these are described in detail, and their shortcomings noted, in Bose Reference Bose2003, 178–93). Each appeals to certain groups within Kashmir and outside, but none commands widespread support across the region as a whole.

The underlying problem with partition is that it aims to carve debatable lands into separate states, and although this follows the general logic of the modern political world, it has particular drawbacks given the way that the territorial issue has arisen in these places. The context, we may assume, is some degree of historic and continuing hostility between states A and B. This means that where partition also involves the creation of devolved forms of government, this is likely to be more problematic than in other cases in which devolution may provide the best answer to nationalist conflicts. Central governments may worry that giving too much power to the devolved administration may provide a springboard for attempts to move the region within the orbit of the rival state. India, for example, has found itself unable to grant the region of Kashmir it controls the autonomy that it was formally promised under Article 370 of its Constitution. So the reassurance that might otherwise be provided to groups who find themselves living on the wrong side of the line of partition by being given some say in the government of the territory is not likely to be forthcoming.

Governing debatable lands

A solution to the debatable lands problem that respects the claims of nationality has to address both the material value question and the symbolic value question. It requires finding a structure of government that ensures that the people who actually live there can enjoy the fruits of their own labour and the labour of their ancestors, while at the same time giving symbolic recognition to the conflicting ‘homeland’ claims of people both within and outside the territory. So the structure has to be more complex than either the plebiscitary or partitionist solutions envisage. It must involve at least two and possibly three levels of government. Within the territory what is needed is a power-sharing form of government with significant powers in the field of economic and social policy, especially, so that development within the region is controlled by the people living there. Power sharing, for example, in the form of proportional representation in the legislature and reserving offices within the executive for the members of different ethnic or national groups, is needed to ensure that members of each group are treated fairly when resources are distributed, and also to enable them to be involved in self-determination.Footnote ¹³ Depending on the ethnic geography, it may be necessary, in addition, to have a further devolution of powers to areas in which a particular group forms a majority. Meanwhile, at a higher level, the larger states must agree to divide powers over the territory between them. That is to say, they must either exercise them jointly, by forming a combined authority to supervise policing or military defence, or they must exercise them side-by-side, with two national flags flying over public buildings, two national television stations broadcasting throughout the region, and so forth. In this way both states maintain not only some measure of control but also an important symbolic presence in the disputed territory.

An immediate objection is that this presumes a willingness on the part of the contending states to co-operate in this way, whereas I have been underlining the mutual hostility that debatable lands tend to generate. There may be a chicken-and-egg problem here if the root cause of antagonism between the two states is their rival claim to the debatable land itself, since the solution, once it is achieved, may be accepted as representing a fair compromise. From a practical point of view, this suggests that third parties may have an important role to play in propelling the two sides to make concessions that once made may produce a reasonably stable outcome. Nonetheless, it remains true that if the relationship between the two states is deeply hostile – neither perhaps recognizes the other as a legitimate authority – then the kind of solution sketched above is not going to be possible. The history of the original Debatable Land reveals how the possibilities for practical co-operation between different groups within the territory ebb and flow as the relationship between the two bordering states deteriorates or improves (see Neville Reference Neville1998). To take another example, it was been widely noted that the implementation of the Good Friday agreement in Northern Ireland, which moves somewhat in the direction suggested above,Footnote ¹⁴ only became feasible because relations between the Republic of Ireland and the United Kingdom had improved in general, making co-operation in areas such as policing imaginable, where they had not been before. The difficulty here is a real one, but it is important to see that it is not an objection to the solution itself, understood as an application of the nationalist theory of territorial justice. The point of that theory is to tell us what entitles a group to control territory; it does so by providing criteria by which to measure the strength of rival claims. In the case of debatable lands, we find that the criteria do not point unequivocally to one group as the proper holders of territorial rights, but suggest rather that these rights need to be divided up in the way that best reflects the nature of the underlying claims. The fact that at a particular time and place national hostilities prevent the just outcome from being reached does not disqualify the theory.

A more principled objection might be that nationalist demands are always claims for sovereignty, which implies exclusive control of territory: thus, the proposed solution cannot meet these demands, on any side, whereas the solutions I have rejected (the plebiscite and partition) do preserve sovereignty, even though the area over which it is exercised will fall short of the full aspirations of the peoples in question. But although nationalist leaders frequently claim that territorial sovereignty is their objective, it is far from clear that sovereignty is always necessary in order to achieve national self-determination. The underlying aim is for a people to be in charge of its own destiny, and as part of that objective to preserve and enjoy the value that they have added to land. For that purpose they do need to have some rights of jurisdiction, but these do not have to be unqualified (any more than it follows, e.g., that national self-determination is lost when a nation decides to join a larger union, sacrificing some powers in the process: this may prove to be the best way to safeguard self-determination for the future).

It is also important to recall that in the case of debatable lands, national identity is a complex phenomenon. Take a Kashmiri who is committed to the idea of Kashmiriyat. Who does he see as his fellow nationals? He cannot restrict the club just to those who share his commitment to Kashmiriyat as an exclusive identity. He has to recognize that there are also many others who in varying degrees bear Indian or Pakistani identities as well as belonging to Kashmir by virtue of occupancy over time. He cannot dismiss these others as simply misguided, even though he is entitled to believe that his is the truer version and to try to persuade them to accept it. Meanwhile he must acknowledge that any territorial expression of Kashmiri identity in its present plural form requires that some recognition should be given to these other components. Thus, he cannot object in principle to a two-level solution in which the Indian and Pakistani states are both given some rights, though he will doubtless argue for enhancing the relative scope and powers of the Kashmiri government itself.

Questions may then be asked about the status in international law of the debatable land under the proposed arrangement. Does it qualify as an independent state? Would it be given a seat in the United Nations, be entitled to compete in the soccer World Cup, have its own flag, etc.? These questions reflect the statist bias of the current international order, notwithstanding the fact that the entities recognized as independent states are hugely diverse, ranging from nuclear-armed superpowers to tiny principalities like Lichtenstein, which has no armed forces at all. Depending on the precise division of powers arrived at, debatable lands governed in the way proposed here might be recognized as states for some purposes but not others (it might be useful to draw comparisons with those minority nations in Europe such as Catalonia and Scotland that have a substantial degree of political autonomy and operate for some purposes as independent political (and sporting!) units).

Another question that might be raised here is about the drawing of boundaries: a debatable land may or may not already have formal boundaries, in the sense of being recognized as a sub-division of one or more existing states. How does the approach taken here give guidance if the positioning of the boundaries is in dispute – for instance, if a partial autonomy scheme of the kind described above is being mooted, but one or both of the rival states wants to absorb part of the debatable land into its own territory outside of the scheme? The solution I am defending aims to lessen the significance of boundaries by giving expression to interests and identities that run across as well as within them, but this does not mean that the boundary issue can be wholly ignored. The nationality theory can again give some guidance here. If within the existing area of the debatable land, there are smaller areas whose population with few exceptions identifies strongly with one of other neighbouring state, then it may be a consensual matter to redraw the border so that these areas are included within it. Of course, the effect of redrawing on the political balance within the debatable land itself needs also to be considered, since the aim is to create a balanced consociational arrangement between the different groups who live there. It is sometimes argued that all boundary redrawing should be considered illegitimate under the so-called Uti Possidetis rule of international law, but in fact that rule has not been rigidly applied in practice, nor is it supported by decisive reasons.Footnote ¹⁵ It is better regarded as the default rule that should be followed unless there are widely accepted reasons in favour of border changes, such as reuniting cohesive groups who have been split asunder by the outcome of a military conflict. The Dayton agreement that ended the Bosnian conflict included some redrawing of the line between the Serb-held and Muslim/Croat-held areas so that each group ended up as a majority group in several cantons of the Bosnian state.