Frozen Conflicts in International Law

II.Time Factor

Time plays a crucial role in frozen conflicts. They have frequently been defined by their longevity and their persistence for over three decades.⁵³² However, whether three decades is a long or short duration for a conflict is relative. Furthermore, the legal content of such temporal descriptions is limited. The essential issue is not the duration of the conflicts per se but the various legal questions that emerge over time. The following chapters discuss three time-related — S. 77 —questions that arise in frozen conflicts. First, reflecting on the duration of a conflict prompts the question of when a war truly ends. Second, the question arises whether international law recognizes any time frame to settle a dispute or whether the parties to a conflict can defer the status question ad infinitum. Finally, a long-lasting conflict has consequences for both the title over territory and the de facto state. Discussing these points will show how the time factor contributes to the frozenness of the conflicts in many ways.

A.Short Wars and Long Conflicts

When looking at the conflicts in South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh from the perspective of the passage of time, two initial observations can be made. First, effective control over the four regions was established by the separatists, with support by their patron states, in relatively short armed conflicts that lasted for some weeks or months at the beginning of the 1990s.⁵³³ In contrast, the de facto states that were established after these short wars and the ensuing disputes over the territories’ status have persisted for the relatively long period of more than three decades.⁵³⁴ Commenting on the duration of the conflicts involves grappling with the question of when a war or conflict truly terminates: can we say definitely when an armed conflict no longer exists under international law?⁵³⁵ In Orend’s words, ‘who can say, around the dawn, exactly where night ends and day begins?’.⁵³⁶ According to the ILA Report on the Definition of Armed Conflict, ‘international law contains no rule … as to how long the cessation needs to last for an armed conflict to be considered legally at an end’.⁵³⁷ Similarly to defining the existence of an armed conflict, defining its end involves considering a number of factors. These include considerations on the maintenance of battle positions, the withdrawal of forces, the risk of renewal of fighting at any time, the resumption of peacetime activities such as trade, commerce, agriculture, and manufacturing, and the return of refugees and displaced people.⁵³⁸

The most obvious and ideal termination of a war is the conclusion of a peace treaty between the warring parties.⁵³⁹ Such a treaty is the counterpart to the formal declaration of war that marks the beginning of a war and decides the temporal scope of the application of the laws of war. Yet, in none of the frozen conflicts has a peace treaty been concluded. In fact, neither peace treaties nor declarations of war have been common since the Second World War.⁵⁴⁰ The states of war and peace have both been deformalized, and today it is the factual situation that determines the application of the laws of war, rather than formal criteria of declarations. By consequence, peace treaties have mostly been replaced by armistices or ceasefire agreements.⁵⁴¹ Traditionally, ceasefires and armistices were intended to be of short duration until a peace treaty is concluded. However, they can also be prolonged for decades, for instance in the case of North and South Korea.⁵⁴² The practice of agreeing on ceasefires developed decisively after the end of the Cold War in cases of non-international armed conflicts.⁵⁴³ Normally, ceasefire agreements provide for the physical separation of forces, including the definition of ceasefire lines and buffer zones.⁵⁴⁴

Given the quasi-replacement of peace treaties by ceasefire agreements, the question arises whether the ceasefire agreements that have been concluded in frozen conflicts have terminated the conflicts. On the one hand, they have successfully ended recurring armed hostilities, not only in the early 1990s but in 2008 for South Ossetia and Abkhazia and in 2020 for Nagorno-Karabakh as well. They effectively put an end to the armed conflicts, and the laws of war no longer applied. On the other hand, the underlying disputes have not been resolved. The established case law of the ICJ defines a dispute as a ‘disagreement on a point of law or fact, a conflict of legal views or of interests between two persons’⁵⁴⁵ and notes ‘that the claim of one party — S. 79 —is positively opposed by the other’.⁵⁴⁶,⁵⁴⁷ Moreover, the factual situation on the ground presents a permanent breach of international law concerning the territorial integrity of the parent state. Thus, despite the growing importance of ceasefire agreements, it must be stressed that they do not ‘introduce peace in the full sense of that term’.⁵⁴⁸ Although there is no official legal definition of ‘ceasefire’, it has been described as a suspension of acts of violence by military and paramilitary forces, usually resulting from the intervention of a third party.⁵⁴⁹ It indicates a temporary or permanent cessation of hostilities, but without normalization of the relations between the parties.⁵⁵⁰ Thus, Grotius might still have a point in saying that in a ceasefire, a war is ‘not dead, but only sleeping’.⁵⁵¹

This holds true particularly for frozen conflicts. Although large-scale hostilities and full-scale wars have been terminated by ceasefire agreements, relations between the parties have not been normalized, and the question of the status of the break-away regions has not been settled. The status question has instead been postponed, and violence still breaks out from time to time, except in Transnistria, where hardly any violence has occurred since the war in 1992. There is a permanent risk that the parent state might seek to reincorporate the break-away region by military force, resulting in a permanent state of readiness for war.⁵⁵² Therefore, it is accurate to refer to the frozen conflicts as ongoing, long-lasting, protracted, and so forth. However, such time-related labels do not carry a substantial value on their own. Likewise, the notions of ‘prolonged occupation’, as opposed to ‘occupation’, and ‘protracted armed conflict’ (as in article 8(2) (f) Rome Statute), as opposed to ‘armed conflict’, have been found to be descriptive — S. 80 —and not to carry any distinct legal significance.⁵⁵³ Similarly, the ICJ held that the fact that Israel’s occupation of the Palestinian territory ‘is prolonged does not change its legal status under IHL’.⁵⁵⁴ It is therefore valuable to further explore the time factor beyond merely describing frozen conflicts as long-lasting.

B.Playing for Time

Considering the persistence and long-lasting nature of frozen conflicts, the question why they endure emerges. The parties’ maximalist legal positions have already been identified as a key factor leading to indefinite deadlock. The parties prefer to uphold the status quo rather than achieve a consensus through compromise.⁵⁵⁵ Similarly, Borgen stated that the conflict in Transnistria ‘has been frozen not so much because there are no other options under domestic and international law besides secession, but because the separatists gained by making the conflict seem intractable’.⁵⁵⁶ Considering the time factor, this strategy can be regarded as playing for time. Wordings such as ‘frozen agreement’⁵⁵⁷ and ‘deferral of the issue’⁵⁵⁸ support the view that playing for time is one of the parties’ tactics. This tactic is facilitated by the fact that international law defines no time frame in which to settle disputes. ‘Long-term diplomatic negotiations on the settlement of territorial disputes with de facto regimes have no time frames and their success depends on the parties’ and international organisations’ political considerations.’⁵⁵⁹ Negotiations on the final status of the break-away regions in question have been unsuccessful for decades without any final deadline being set for status talks.⁵⁶⁰ Under international law, the parties have no binding obligation to further status negotia- — S. 81 —tions;⁵⁶¹ however, on the basis of their prior political commitments, they arguably have the duty to conduct negotiations in good faith.⁵⁶²

The UN Charter has enshrined the pacific settlement of disputes in Chapter VI. According to article 33, the parties to any dispute shall seek a solution by negotiation, enquiry, mediation, conciliation, arbitration, judicial settlement, resorting to regional agencies or arrangements, or other peaceful means of their own choice. The rule has crystallized as customary law and applies not only to UN member states but also to third parties and to all other entities enjoying the protection of the ban on the use of force, including de facto states.⁵⁶³ The rule does not explicitly require the dispute to be of an international character, but purely domestic disputes are clearly excluded from its scope.⁵⁶⁴ It does not set any time frame, for instance by adding the words ‘in a timely manner’; neither does the UNGA Friendly Relations Declaration, which entails the obligation to refrain from the use of force in territorial dispute settlements.⁵⁶⁵

Furthermore, the parties’ playing for time is facilitated because international organizations and third states do not act vis à vis frozen conflicts in a timely manner. In none of the cases examined here has the UNSC exercised its powers under Chapter VII of the UN Charter to regulate post-conflict political transformation. As a result, no peace-building mission or international territorial administration has been mandated for the purpose of de-occupation transition.⁵⁶⁶ Article 41(1) of the ILC’s Draft Articles on Responsibility of States for Internationally Wrongful Acts⁵⁶⁷ holds that states shall cooperate to bring to an end through lawful means any serious breach of an obligation arising under a peremptory norm of general international law. According to the commentary, such cooperation must be by lawful means, the choice of — S. 82 —which will depend on the circumstances of the situation.⁵⁶⁸ Berkes discussed this obligation to cooperate by international organizations and third states to end the unlawful territorial situation in frozen conflicts.⁵⁶⁹ He emphasized that this rule, again, does not explicitly set any time frame for the obligation to cooperate, and the speed of any collective action is at the states’ discretion.⁵⁷⁰ However, he further argued that according to a progressive interpretation of the notion of ‘cooperation’, it has to occur in a timely manner in order to eliminate the wrongful situation as fast as possible.⁵⁷¹ He bolstered this argument with the rationale of solidarity, the duty to cooperate, and with the nature of the breach of the peremptory norm that has to be terminated: especially when a breach of the peremptory norm is likely to cause irreparable damage to individuals, collective action by the international community should be taken in a timely manner.⁵⁷² Examples that require collective action ‘in a timely and decisive manner’ include the breach of a ceasefire agreement in eastern Ukraine,⁵⁷³ the shooting down of a civilian aircraft over eastern Ukraine,⁵⁷⁴ and the four most serious international crimes—genocide, war crimes, ethnic cleansing, and crimes against humanity.⁵⁷⁵

Yet, in frozen conflicts, Berkes considered it more problematic to argue for the duty of timely cooperation of international organizations and third states. He stressed that the general problem is that most of the time there is no breach of jus cogens to provoke an emergency situation, nor ongoing armed hostilities.⁵⁷⁶ In other words, frozen conflicts are not sufficiently disruptive so as to deserve decisive intervention by third states.⁵⁷⁷ Therefore, the international community and third states do not cooperate to take timely and decisive actions.⁵⁷⁸ Hence, frozen conflicts are addressed instead through dialogue — S. 83 —and the facilitation of endless, time-consuming talks, in which the parties restate their maximalist positions, rather than through rapid emergency measures.⁵⁷⁹ Considering the distinction between conflict management and conflict settlement, Berkes found that the international community holds itself bound to act urgently only in conflict management but not so much in conflict settlement.⁵⁸⁰ The fact that frozen conflicts can persist for more than three decades surely does not support their perception as emergency situations. Such situations are described by Orwell: ‘But when war becomes literally continuous, it also ceases to be dangerous. When war is continuous there is no such thing as military necessity.’⁵⁸¹ Therefore, as long as frozen conflicts do not escalate into emergency situations or actual armed conflicts, neither involved nor uninvolved parties take rapid action, which results in the continuation of the status quo—possibly ad infinitum.

C.Consequences of Protracted Conflicts

1.Territorial Title

The discussion above has shown two points: first, that frozen conflicts are indeed persistent and of a long duration and that ceasefire agreements have not succeeded in terminating the underlying conflicts and normalizing relations between the parties; second, that international law facilitates this protraction, because there is neither a time frame for territorial dispute negotiations nor an emergency situation that requires decisive and timely action by the international community. This section now examines the consequences of these protracted conflicts on the territorial title.

The uti possidetis principle has already been introduced. In addition, the temporal element⁵⁸² of uti possidetis must be considered. In territorial disputes, there is a critical date at which the principle is applied. Normally, that critical date is the moment the new state declares its independence, with the effect that from this moment the territorial title is frozen.⁵⁸³ Accordingly, for Georgia, Moldova, and Azerbaijan, the critical dates were the moments they — S. 84 —each declared independence from the USSR. On these specific dates, the territorial titles over South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh were fixed, assigning each to its respective parent state, and became immutable.

The process of fixing territorial titles at the critical date is explained by the ICJ in the Frontier Dispute Case: the delimitation is made as if by taking a photograph of the territorial situation at that specific date, ‘it freezes the territorial title; stops the clock’, no matter if the facts change later.⁵⁸⁴ Interestingly, the ICJ used the metaphor of ‘freezing’ to describe the fixation of the territorial title at one specific point in time. The description also stresses the resistance of the title to all subsequent factual changes of boundaries with the metaphor of a photograph. The critical date thus signifies a break in the sequence of state actions related to establishing sovereignty over a disputed territory.⁵⁸⁵ The uti possidetis principle and the critical date doctrine are blind to dynamism or changes over time, no matter how long such changes persist, in pursuit of the overall aim of stability. In the Minquiers and Ecrehos case (France v UK) before the ICJ, the UK’s legal advisor stated in an oral submission that the critical date theory

means that, whatever was the position at the date determined to be the critical date, such is still the position now. Whatever were the rights of the Parties then, those are still the rights of the Parties now. If one of them then had sovereignty, it has it now, or is deemed to have it. If neither had it, then neither has it now. And if both did—that is to say, if there was some sort of joint régime, or condominium—then that régime is still deemed to exist and to govern the rights of the Parties to-day. The whole point, the whole raison d’être, of the critical date rule is, in effect, that time is deemed to stop at that date. Nothing that happens afterwards can operate to change the situation that then existed. Whatever that situation was, it is deemed in law still to exist; and the rights of the Parties are governed by it.⁵⁸⁶

The quotation makes explicit how change to territorial titles is deemed to stop at the critical date. Time is not an agent of law but one of its dimensions; it neither establishes nor extinguishes territorial titles.⁵⁸⁷ Only states can — S. 85 —change territorial titles by their relevant acts over time, such as treaties.⁵⁸⁸ Thus, whoever holds enduring factual control over territory in a frozen conflict, the formal territorial titles remain fixed as they were at the critical date. Neither the passage of time nor the persistence of frozen conflicts over three decades have consequences for the territorial title of the break-away regions. In other words, the passage of time does not affect the de jure title over territory. However, it does affect the de facto situation on the ground, which becomes more consolidated over time.

2.Consolidation of the De Facto States

Examining the consequences of protracted conflicts requires a closer look at the de facto states that have emerged from these situations. First, the persistence of a frozen conflict has distinct effects on both the parent state and the de facto state. For the former, the prolonged conflict exacerbates the injustice due to the ongoing loss of territory, while the latter enjoys continued de facto control over the territory and tends to consolidate it.⁵⁸⁹ Expressed succinctly, time works for the de facto states.⁵⁹⁰ The passage of time normalizes and entrenches the status quo and thus has a consolidating effect on the de facto states.⁵⁹¹ Consolidation or stabilization of the de facto states means strengthening the ‘effectiveness’ and the degree of fulfilment of the factual elements of statehood⁵⁹² according to article 1 of the Montevideo Convention of 1933.⁵⁹³ However, this consolidation is relative, as the de facto states can only exist with the patron state’s support and do not qualify as states. To explain the idea of consolidation, each statehood criterion and its consolidation over time will be briefly outlined following Berkes’s overview.⁵⁹⁴

The first criterion for statehood is a permanent population. Berkes points out that despite the populations’ decrease during or shortly after the wars in South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh, in — S. 86 —more quiet times they have stabilized and sometimes even grown.⁵⁹⁵ There is no threshold for a minimal population; what is important is that the individuals have a shared allegiance.⁵⁹⁶ The regions affected by frozen conflicts have been populated throughout the duration of these conflicts. Moreover, their persistent isolation from the parent states has sustained and deepened the distinction between their and the parent states’ populations and shaped the collective identity of the de facto states.⁵⁹⁷ For example, according to the OSCE, a distinct feeling of Transnistrian identity has developed in Transnistria.⁵⁹⁸ Berkes also argues that international tribunals applying human rights treaties to the break-away regions help normalize the de facto situation and render the status quo more acceptable to the local population.⁵⁹⁹

The second criterion for statehood concerns effective control over a defined territory. The regions in question, with the exception of Nagorno-Karabakh since 2023, are under effective control of the separatists supported by Russia. In fact, exercising effective control over a territory is the key characteristic of a de facto state. As explained above, factual control over the territory has no power to change the formal title, even if it has lasted for decades. Where a title of sovereignty exists, effectivités contra legem cannot prevail over it.⁶⁰⁰ In other words, even if consolidation of a de facto state occurs, it does not go as far as to alter the territorial title, nor does it render the situation lawful. Nevertheless, Berkes argued that in the long term, international dispute settlement mechanisms have a consolidating effect on the ceasefire lines and, consequently, the de facto states’ control over a defined territory.⁶⁰¹ The border management and peacekeeping missions that followed the ceasefire agreements in the frozen conflicts have stabilized the lines between the parent states and the de facto states and thereby de facto consolidated the latter’s territory.⁶⁰² Instances include the EU Border Assistance Mission to Moldova — S. 87 —and Ukraine in Transnistria,⁶⁰³ the OSCE Mission monitoring the contact line⁶⁰⁴ and the Russian peacekeeping mission since 2020 in Nagorno-Karabakh,⁶⁰⁵ the UN Observer Mission deployed in Abkhazia between 1993 and 2009,⁶⁰⁶ and the EU Monitoring Mission patrolling the ceasefire lines following the Six-Point Ceasefire Agreement in Georgia.⁶⁰⁷ To prevent further escalation, these missions aim to stabilize the situations. Although stabilization does not legitimize the status quo, it results in a consolidation of the effective territorial control of the de facto authorities.⁶⁰⁸

The third criterion for statehood is independent government. Both factual and legal evaluations are required to determine the competence to govern a specific territory.⁶⁰⁹ Although, like for the other criteria, the passage of time cannot affect the legal claim to this competence, it can strengthen the factual capacity to govern: Berkes highlighted how through the programs of international organizations and donor states, the international community has even helped build institutions and develop public services in de facto states.⁶¹⁰ He pointed to the instances of aid provided to various projects in Transnistria through the European Neighbourhood Instrument for the Republic of Moldova,⁶¹¹ USAID Moldova,⁶¹² USAID-financed programmes in Nagorno-Karabakh,⁶¹³ and the Peace and Development Programme of the UNDP Georgia in Abkhazia and South Ossetia.⁶¹⁴ Over time, these programs have contributed to economic recovery, the improvement of the humanitarian — S. 88 —situation, and institution-building, all of which strengthen the de facto authorities’ capacity to govern.⁶¹⁵ Government capabilities in the wider sense also include the judicial branch: certain actions by the de facto authorities can gain validity under international law. This was held for instance by the ECtHR regarding the decisions of de facto courts in Transnistria,⁶¹⁶ which constitutes an exception to the principle that ex injuria jus non oritur.⁶¹⁷

The fourth criterion for statehood is the capacity to enter international relations. The most obvious indication of established international relations would be the recognition of statehood by other states. Remarkably, some de facto states have international networks comparable to the diplomatic relations of regular states.⁶¹⁸ Berkes cited Taiwan and Palestine as prime examples of gradual statehood development over time.⁶¹⁹ According to Frowein, the longer a de facto state exists, the more likely it is to be recognized by other states.⁶²⁰ However, the development of Taiwan and Palestine is not only a matter of long-term policy but also hinges on the international community’s willingness to support their independence.⁶²¹ Therefore, Berkes concluded that the passage of time can contribute to the effectiveness of de facto states but does not alone suffice.⁶²² In the cases of South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh, the international community with very few exceptions has always followed the duty of nonrecognition and has therefore not fostered relations with the de facto states.

It is important to note that the criteria for statehood are not met by the de facto states. In particular, the criteria of independent government and the capacity to enter international relations are not fulfilled. An independent government would stand in contrast with the actual dependence of the de facto states on their patron states: Russia in the cases of South Ossetia, Abkhazia, and Transnistria, and Armenia in the case of Nagorno-Karabakh before 2023. However, the key point is that, generally, time helps strengthen de facto states.

3.De Facto States’ Protection under International Law

The passage of time has further consequences for de facto states. Although the existence of de facto states implies a constant violation of the territorial integrity of their parent states, in some respects de facto states are granted a legal status comparable to states: the prohibition of the use of force or annexation applies equally to de facto states.⁶²³ The rationale is to prevent spirals of violence and support stability. Therefore, a state is not allowed to use force against a de facto state, even if it is de jure within its territory.

Azerbaijan’s use of force in 2023 to reconquer Nagorno-Karabakh thus violated the prohibition of the use of force.⁶²⁴ Ruys and Rodrigues Silvestre comprehensively discussed the unlawfulness of Azerbaijan’s use of military force in alleged self-defence in the second Karabakh war in 2020.⁶²⁵ They found that Azerbaijan had no right to the use of force for self-defence, because the factual territorial status of Nagorno-Karabakh had already been established for such a prolonged period that an armed attack from the separatists was no longer ongoing.⁶²⁶ In other words, Azerbaijan lost any right it may have had to act in self-defence because the status quo of Nagorno-Karabakh had lasted for almost three decades.⁶²⁷ Even in cases of occupation, where the immediacy requirement of self-defence according to article 51 UN Charter should be construed flexibly, the ‘lapse of time between the initial attack and the invocation of self-defence cannot be extended indefinitely’.⁶²⁸ This assessment underscores how time works for the de facto states. However, instances where states have used armed force to challenge the existing territorial status have been rare, and even fewer instances exist where they have justified such actions by invoking the right to self-defence.⁶²⁹

In contrast to this perspective, Akande and Tzanakopoulos argued that Azerbaijan still had the right to self-defence against Nagorno-Karabakh in 2020 irrespective of the passage of time.⁶³⁰ They noted that the occupation of Nagorno-Karabakh by Armenia and the separatists was a direct consequence of the unlawful armed attack of 1994.⁶³¹ The situation in 2020 therefore could not but be an integral part of that initial attack, and as a consequence, Azerbaijan had the right to self-defence for as long as the occupation continued.⁶³² They bolstered the argument by referring to article 3(a) UNGA Resolution on the Definition of Aggression, which states that a military occupation resulting from an invasion or armed attack qualifies as an act of aggression.⁶³³ The view that the prohibition of the use of force protects a de facto state vis à vis its parent state is controversial: although the IIFFCG Report clearly affirmed the application of article 2(4) UN Charter to de facto states when assessing Georgia’s attack on Tskhinvali in August 2008, not viewing it as problematic at all,⁶³⁴ Henderson and Green criticised this approach vehemently.⁶³⁵ Furthermore, Waters generally stressed that there is simply no prohibition in international law on a state using force within its own territory to supress insurrections.⁶³⁶

Lieblich analysed both Ruys and Rodrigues Silvestre’s restrictive view and Akande and Tzanakopoulos’s permissive view of Azerbaijan’s war of recovery in 2020.⁶³⁷ Analysing Lieblich’s paper in detail would exceed the scope of this work. In short, he discusses the complex ethical and legal dilemmas surrounding such wars of recovery. According to Raz, he argued, law fulfils its role if it guides us towards moral behaviour.⁶³⁸ In wars of recovery, where a state attempts to reclaim lost territory, two principles are in conflict—territorial integrity and individual rights—, and this causes legal and moral ambiguity. Dworkin and Franck, he explained, argued that when rules are not coherently connected to underlying general principles, their legitimacy is compromised. The absence of clear moral and principle-based guidance results in the lack of any — S. 91 —authoritative answer on the issue of wars of recovery.⁶³⁹ Lieblich concluded that as long as international law oscillates between statism’s claim to territorial integrity and individualism’s emphasis on individual rights, the legal debate on wars of recovery is likely to continue. He observed that law’s legal–illegal binary approach is insufficient to address the topic. Instead, he proposed a more nuanced approach, viewing wars of recovery as non-prohibited rather than strictly legal or illegal. He bolstered this argument by referring to judge Simma, who in the ICJ Advisory Opinion on Kosovo introduced the concept of toleration regarding Kosovo’s independence.⁶⁴⁰ Such nonbinary concepts beyond legal–illegal ‘could open space for contestation of wars of recovery’.⁶⁴¹

Without further examining the lawfulness of wars of recovery,⁶⁴² for the purpose of this work, two points must be emphasized. First, the time factor plays a crucial role in frozen conflicts and thus in potential or actual wars of recovery. The passage of time complicates arguments for the lawfulness of wars of recovery because, despite their illegal status, the de facto states become more consolidated and stable and can benefit from the prohibition of the use of force. Although the passage of time does not affect the de jure territorial situation, it does strengthen the de facto states’ territorial situation. Time thus pulls in two different directions in frozen conflicts, exacerbating the disjunction between the de jure and the de facto territorial situations. This fact contributes to their deadlocked nature. Second, Lieblich’s suggestion of introducing more nuanced concepts than the binary of legal–illegal is of broad significance for this thesis. In many respects, frozen conflicts transcend widely accepted binary categories of law and challenge them fundamentally, with the consequence that more effort has to be put into legal argumentation for one position or the other. The transgression of binary categories by frozen conflicts will be examined in the next section.