Frozen Conflicts in International Law

Part I: Investigation of the Term

The initial focus of the research is to explore the term ‘frozen conflicts’ and to address several fundamental questions. The subsequent analysis requires understanding of the origin, use, and implications of the term. In particular, it is crucial to determine which situations are referred to as frozen conflicts. Moreover, this section examines conceptualizations and definitions that have been suggested by international lawyers and other scholars. Discussing tentative conceptualizations of frozen conflicts involves considering other legal concepts and examining their distinctions and overlaps with frozen conflicts. The aim is to present a comprehensive profile of the term ‘frozen conflicts’ that can provide a basis for further analysis.

I.Background of the Term

A.Origin

Defining the origins of specific terms and concepts can be both simple and complex. It can be simple when a term emerged following a specific and salient event, such as the ‘war on terror’, a phrase that can easily be attributed to US policy following the attacks of 9/11 in 2001. Conversely, it can be complex when a concept is not clearly attributable to a single event, when it remains fuzzy, and/or when it has not yet been fully conventionalized. Some think a phenomenon can certainly exist avant la lettre, whereas others assume that concepts always ‘set free new realities’.¹⁶ According to Skinner, changes in vocabulary correlate with changes in social perceptions and awareness.¹⁷

Identifying the first use and originator of the term ‘frozen conflicts’ is far from straightforward. Authors thus disagree on this question. Grant examined the etymology of ‘frozen conflicts’ and found ‘one of the first’ occurrences in — S. 8 —1911 in the journal The Academy and Literature.¹⁸ An article titled ‘Hero as a Baby’ pictures well-known individuals when they were little. One scene reads as follows: ‘Napoleon weeping bitterly because he was compelled to go early to bed and to leave his leaden soldiers in a frozen conflict on the table’.¹⁹ As the author does not further elaborate the scene, it is unclear whether this use of the term has a deeper meaning. The author trivially refers to toy soldiers that remain frozen, meaning in the exact same position, until the next day when little Napoleon will continue playing. The first use of the term to refer to a real conflict, rather than to leaden soldiers, according to Abline, occurred in 1998 in a OSCE report describing the situations in the Republic of Moldova (henceforth Moldova) and in Georgia.²⁰ In contrast, Grant points to the Yearbook of Polish Foreign Policy and suggests that the personal representative of the OSCE Chairman-in-Office called the conflict in Transnistria ‘frozen’ in 1994.²¹

However, leaving aside the fictional one of 1911 of Napoleon as a child, we found earlier examples of the term. In 1974, the North Yemenite conflict was described as ‘remaining frozen’.²² This conflict had begun in 1962 when the Republicans, led by the army, launched a coup d’état against the former dictator Muhammad al-Badr, leading to a civil war. By the late 1960s, the war had sunk into a stalemate with no clear end. The article published in 1974 explained that the conditions that had led to the conflict still existed and might lead to a new confrontation, especially because it was embedded in broader politics.²³ Furthermore, at the UNGA meeting on Antarctica in 1985, the Chilean representative emphasized the

success of the Antarctic Treaty system—which has frozen conflicts over sovereignty, fostered uninterrupted cooperation between rival countries, and created the first zone of disarmament and peace, the first ecological preserve and the first scientific laboratory territory.²⁴

The metaphor also appeared in the New York Times in 1969, although in a slightly different variation. An article titled ‘The Frozen War’²⁵ discussed the Cold War and whether the hostilities between the Soviet Union and the West had ended. The author considered the outbreak of a hot war between the two sides unlikely and suggested that the Cold War had become a frozen war: continuing ‘in the thermal metaphor, one might say that the cold war is frozen over’.²⁶ These references show the term’s appeal and its usage in various contexts. Yet, despite the lack of consensus on its introduction, there is broad consensus on when it gained momentum.

Apart from some trivial references, authors who engage with the term ‘frozen conflict’ seem to agree that it became popular in the 1990s.²⁷ The historical context was the aftermath of the collapse of the Soviet Union in 1991, the emergence of fifteen new states, and the disappearance of other entities.²⁸ The new states on the map were Armenia, Azerbaijan, Belarus, Estonia, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan. Their territories were identified according to the uti possidetis juris principle, which transformed the former Republics’ administrative boundaries into international ones. Secondary and tertiary entities within these Republics were not considered entitled to independence.²⁹ Nevertheless, some of these entities aimed, and still aim, for separation. Whereas some of the separation movements remained peaceful, for instance in the autonomous regions of Gagauzia in Moldova³⁰ and Ajaria in Georgia, others developed into wars, such as in Chechnya in Russia, which was especially severe.³¹ Moreover, these new entities, whether independent states or regions aiming for independence, suddenly had the choice to adjust their positions vis à vis the Russian successor state and the West. Conflicts emerged in the regions of South Ossetia and Abkhazia in Georgia, Transnistria — S. 10 —in Moldova, and Nagorno-Karabakh in Azerbaijan. As in Chechnya, the aspiration for independence led to war, but unlike Chechnya, these regions are still de facto separated from the parent state—except for Nagorno-Karabakh, which was reintegrated into Azerbaijan in 2023. The situation between separatists aiming for independence, backed by a patron state, and the parent state to which they formally belong has been termed ‘frozen conflict’. The term implies that the conflict remains unresolved and the situation is in limbo between war and peace.³² The broad consensus among scholars of various disciplines is that the four conflicts in South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh were the original or classic frozen conflicts and that the use of the term has expanded since then.³³ Interestingly, the term is the same in various languages: for example conflit gelé in French, eingefrorener Konflikt in German, заморожений конфлікт (zamorozhenyy konflikt) in Ukrainian, or замороженный конфликт (zamorozhennyy konflikt) in Russian. According to one Russian author, the Russian expression is a derivation of the English one, which reflects a critical attitude of the author to Russian policy.³⁴

B.Conventionality of the Metaphor

The term ‘frozen conflict’ thus emerged in the specific historical context of the dissolution of the Soviet Union. Nevertheless, it has been conventionalized and applied to other conflicts, even to some that had actually existed and been frozen before. For example, the conflict between India and Pakistan over Kashmir has persisted for decades. The first of five wars took place more than 70 years ago following the withdrawal of the British Empire in 1947, and the situation has ever since been characterized by the recurrence of violence between terrorist groups and the military, large-scale political protests, and police repression.³⁵ In public and academic discourse, however, the Kashmir — S. 11 —conflict has been called frozen only recently.³⁶ Another example of the term’s expansion is the conflict between the Republic of Cyprus and Northern Cyprus. Again, the conflict existed before the collapse of the Soviet Union, since the Republic of Cyprus’s independence from British rule in 1960 and the Turkish intervention in 1974, which led to a short war.³⁷ However, labelling it a frozen conflict has become frequent only in the last twenty years.³⁸ Moreover, the term has been used from the 2000s onwards in academic and public discourse to describe conflicts such as those in Kosovo,³⁹ Western Sahara,⁴⁰ Israel– Palestine (referring to the time before autumn 2023),⁴¹ Bosnia and Herzegovina,⁴² eastern Ukraine (referring to the time between 2014 and 2022),⁴³ Taiwan,⁴⁴ Catalonia,⁴⁵ North Korea–South Korea,⁴⁶ and probably more. Accordingly, the term’s use has extended from the particular post-Soviet context to more universal applicability.

The term’s spread and popularity enables further observations on the metaphor it embodies. A metaphor connects a ‘source domain’ with a ‘target domain’.⁴⁷ The two domains are interconnected by ‘mapping’, which is partial and non-absolute and allows features of the target domain to be highlighted.⁴⁸ Thus, the words ‘frozen conflict’ are the source domain, while the actual situation on the ground is the target domain. ‘Frozen’ maps onto unresolved, prolonged, enduring, low-intensity violence, without either war or peace. Metaphors are common in all idioms of the world and are used everywhere. They can contribute to making language more efficient,⁴⁹‘fill lexical gaps and motivate semantic change’,⁵⁰ and function as merely decorative elements.⁵¹ They are also used in law, where they can serve both epistemic and poietic functions.⁵² In any case, a metaphor needs time to become conventional because a new metaphor begins as nothing more than a comparison.⁵³ Only after having been employed frequently enough does a metaphor’s base term become polysemous and complement its original, literal meaning.⁵⁴

As noted above, the ‘frozen conflicts’ metaphor has been in use for some decades and has been applied frequently by various writers and speakers to describe a range of conflicts. Moreover, its meaning is highly intuitive, so the literal meaning of frozen has almost disappeared. People would not be surprised to learn that frozen conflicts are not actually frozen. One reason for this might be its semantic proximity to Cold War, a metaphor that has become highly conventional. The latter has lost some of its power because of its conventionality⁵⁵ and become a source domain when people speak of ‘a cold war’ instead of ‘the Cold War’.⁵⁶ More broadly, the range of thermal metaphors — S. 13 —employed within the context of frozen conflicts is striking. Phrases such as ‘threading on thin ice’,⁵⁷ ‘thawing a frozen conflict’,⁵⁸ ‘periods of thaw or freeze’,⁵⁹ ‘melting ice does not always freeze back in its previous shape’,⁶⁰ ‘the fighting cooled’,⁶¹ ‘punctuated by periods of hot war’,⁶² ‘Georgia’s frost’,⁶³ and ‘Georgia’s Secessionist De Facto States: From Frozen to Boiling’,⁶⁴ all presuppose the conventionalization of the frozen conflict metaphor. However, the phrase is often put in quotation marks as ‘frozen conflict’ or ‘frozen’ conflict.⁶⁵ The term is also regularly complemented by ‘so-called’: ‘the so-called “frozen” conflicts’.⁶⁶ This points to the term still being in a grey zone⁶⁷ and is evidence against its complete conventionalization. In contrast, Cold War does not usually appear in quotation marks, which implies that everyone is expected to understand it immediately. The frozen conflict metaphor evidently can be applied in various ways. Some consider it simply an ‘enticing’ metaphor,⁶⁸ a linguistic decoration or catchphrase, especially when appearing in newspaper headlines or book titles. However, others consider it an analytical concept. This work is interested in the latter approach. New phenomena continually emerge in the world, and both language and law constantly adapt by conceptualizing these changes, sometimes with metaphors.

C.Problematic Implications

It is part of the nature of metaphors that they are not neutral but normally imply some judgment, as they highlight just one characteristic of the target domain. One author explains the use of metaphors, particularly in law, with a metaphor: they are either used as weapons ‘where a given metaphor is mobilized in the legal/political stage’ or as a battlefield ‘where different groups fight over the control over the signification of a metaphor’.⁶⁹ With this in mind, the following four problematic aspects of the frozen conflict metaphor have been identified.

First, many criticize the adjective ‘frozen’ as misleading, because it suggests stability when in fact the situations referred to are very changeable and inherently risky.⁷⁰ They can turn into wars any time and can cause considerable casualties and destruction to a region.⁷¹ The war in South Ossetia and Abkhazia in August 2008 showed how a ‘frozen’ situation can ‘defrost’ (разморозка⁷², razmorozka) or ‘thaw’⁷³ very quickly. Whereas Transnistria is considered the quietest frozen conflict, Nagorno-Karabakh has repeatedly experienced outbreaks of violence throughout the last decades, most recently in 2016, 2020, and 2023, causing many deaths and the displacement of hundreds of thousands of people. The situation in eastern Ukraine, which had also been called a frozen conflict before 2022,⁷⁴ became a full-scale war with Russia’s invasion that year. Additionally, the alleged stability conveyed by the adjective ‘frozen’ implies a sense of passivity among both the parties involved and actors that are not involved. Labelling a conflict as frozen invites inaction by supporting the assumption that the conflict is of marginal relevance and that addressing it is not urgent.⁷⁵ To underline that frozen conflicts are not — S. 15 —actually frozen at all, one author consistently puts the word frozen in brackets: ‘(frozen) conflicts’.⁷⁶

Second, frozen conflicts are understood to be of long duration, existing for over three decades.⁷⁷ Yet from a historical perspective, their duration is not in fact particularly long. Bebler illustrated this criticism with the conflict over Gibraltar between Spain and the United Kingdom, which has endured for 300 years following the British occupation in 1704.⁷⁸ Although Gibraltar was donated to the United Kingdom by Spain following the Treaty of Utrecht in 1713, several Spanish governments have since attempted to reclaim the territory, arguing that British sovereignty there violates the territorial integrity of Spain.⁷⁹ The conflict, manifested in land blockades, disruptions of traffic, harassment of tourists and Gibraltarians, and intimidation by jets’ low overflights of the area,⁸⁰ has indeed been named frozen,⁸¹ which is appropriate due to its long duration, according to Bebler.

Third, the term is confusing because it remains unclear what exactly is described as ‘frozen’. It can refer to violent actions that are infrequent when compared to full-scale war, it can denote a stalemate in peace negotiations, and it has even referred to human rights being ‘frozen’,⁸² meaning that they are not respected in these regions. Acknowledging this inaccuracy, the present thesis approaches the term ‘frozen’ in relation to multiple factors at various levels that together convey frozenness.

Finally, some authors have criticized the term’s bias.⁸³ It is not neutral but implies a certain attitude towards Russian policy and interpretation of international law. Indeed, the term has not only emerged in the post-Soviet context but has also been linked to Russia’s practices in its ‘near abroad’ when — S. 16 —pursuing the project of русский мир (russkji mir: Russian world).⁸⁴ Despite, or indeed because of, these allegedly problematic implications, a holistic examination of the term is desirable.

II.Use of the Term

A.Empirical Analysis of UN Documents

An empirical analysis of the occurrence of the term ‘frozen conflict’ can generate quantifiable insights. As mentioned above, there is broad consensus on the term’s emergence in the 1990s and its original use to describe conflicts in the post-Soviet context. However, no attempt has yet been made to support this assumption with data. The present work aims to fill this research gap and has therefore empirically analysed the use of the term in the UN digital library. The UN digital library is the bibliographic database of the UN, established in 1979 and curated by the Dag Hammarskjold Library.⁸⁵ The UN database was chosen because of its international scope and relevance to both law and politics, the vast array of documents provided, and its inclusion of documents dating back to the foundation of the UN in 1945. The empirical analysis addressed two research questions: first, how often has the term ‘frozen conflict’ been used over time by state officials and international organization officials? Second, which states have most frequently been mentioned as being affected by a frozen conflict, and which states have most frequently used the term ‘frozen conflict’ to refer to other states?

The following method was applied: a full text search for ‘frozen conflict’ was performed on the UN digital library platform, and 234 records were found for the period between 1945 and 2023.⁸⁶ After excluding publications, NGOs’ letters and statements, footnote references, book quotations, quotations of previous speakers in meetings, and duplicates, exactly 200 documents remained. These documents include meeting records, letters, verbatim notes, reports, and written statements. No UNSC Resolution or UNGA Resolution has included the term yet.⁸⁷ In total, the term ‘frozen conflict’ occurs in these documents 241 times. For each of these instances, three pieces of information — S. 17 —were collected: the year of use, the speaker using the term, and the state mentioned as involved in a frozen conflict.

As far as this third item is concerned, one instance can refer to one or several states. For example, a record such as ‘Abkhazia and Transnistria are two of many frozen conflicts in the Black Sea region’ would be considered to mention both Georgia and Moldova as states involved in a frozen conflict. Moreover, implicit references were also recorded. For example, the statement that ‘we all know that there are dozens of Kosovos throughout the world, just waiting for secession to be legitimized, to be turned into an acceptable norm. Many existing conflicts could escalate, frozen conflicts could reignite, and new ones could be instigated’⁸⁸ was registered under ‘Kosovo’ as a state involved in a frozen conflict. However, overinterpretation of the statements was strictly avoided. Importantly, negations of a frozen conflict were also recorded. For example, ‘Ukraine is no longer a frozen conflict’ and ‘the term frozen conflict is misleading in the case of Nagorno-Karabakh’ would still result in Ukraine and Azerbaijan, respectively, being registered as the states referred to as affected by frozen conflicts.

B.Findings

The first research question, how often the term ‘frozen conflict’ has been used over time by state officials and international organization officials in the UN Documents, is answered in the chart below (Figure 1).

As the chart shows, the first use of ‘frozen conflict’ dates back to 1985. This instance of the term’s use has been mentioned above: Chile referred to the Antarctic Treaty System as having ‘frozen conflicts over sovereignty’.⁸⁹ Thereafter, the term was not used until 1998, when state officials started using it. International organization officials only started using it in 2000. This is slightly later than expected given that, as outlined above, many authors write generally of the emergence of the term in the 1990s. We see two explanations for this. First, although the ‘original’ frozen conflicts of South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh all faced wars after the dissolution of the Soviet Union in 1991, these conflicts only became frozen during the 1990s. As will be discussed later, such a passage of time is a key characteristic for conflicts to be perceived as frozen. Arguably, the longer a conflict is in a deadlock, the more appropriate the term ‘frozen’ is when referring to its unchanging state. Second, as shall also be discussed later, frozen conflicts are characterized by Russia’s approaches to international law and politics. It was in the 2000s that Russia’s perspective changed and became more competitive with the West, not in the 1990s, when it was still weak.⁹⁰ Accordingly, we think that the wider use of the term falls within this decade because only then were the patterns of Russia’s policies recognized.

Another finding concerns the consistency of use of the term since 1998 by state officials. In contrast, international organization officials did not use the term at all in 2001, 2020, or 2021; however, they have still used the term in most of the years since 2000. In fact, the term has never completely disappeared since 1998, although the frequency of its use has fluctuated. Two peaks are evident in the use by state officials, one in 2006 and one in 2017. Although no specific event explains the first peak in 2006, a UNSC meeting in 2017 with the agenda item ‘Conflicts in Europe’⁹¹ coincided with most of the references to frozen conflicts that year and thus contributed to the second peak. Finally, it is striking that state officials have employed the term much more often than international organization officials. Among these officials, the term was predominantly used by UN officials (30 out of 42 instances) and less by OSCE officials (6 out of 42 instances). Officials from the EU, ICRC, and the regional organization of Georgia, Ukraine, Azerbaijan and Moldova (GUAM) used the term less than five times.

The second research question is which states were referred to as affected by a frozen conflict, which states referred to frozen conflicts, and how frequently these references occurred from 1945 to 2023. The answer is provided in the chart below (Figure 2). The black bars show which states were referred to as affected by a frozen conflict. The grey bars show which states have used the term ‘frozen conflict’ when referring to other states. The numbers on the horizontal axis of the chart shows how often these references occurred between 1945 and 2023. — 11 —

Fig. 2: Use of the term ‘frozen conflict’ regarding states

Figure 2: Use of the term ‘frozen conflict’ regarding states

USA

Poland

France

Romania

Turkey

Cyprus

Israel-Palestine

South Sudan

Kosovo

Ukraine

Moldova

Azerbaijan

Georgia

States referred to as frozen conflict

States using the term ‘frozen conflict’

The graph indicates that Georgia, Azerbaijan, and Moldova are the states most frequently referred to as affected by frozen conflicts (all three states more than 60 times). In Georgia, the conflicts referred to involve South Ossetia and Abkhazia, in Azerbaijan, Nagorno-Karabakh is affected, and in Moldova, Transnistria. The conflicts in these four regions are considered the original or classical frozen conflicts, as has been noted above. However, until now no — S. 20 —data has been presented to confirm this assumption. Figure 2 now provides empirical proof. Additionally, Ukraine has been referred to as affected by frozen conflicts 38 times. These mentions refer to eastern Ukraine as well as Crimea. They became frequent only after 2014, when Russia annexed Crimea and supported the proclamation of independence of Donetsk and Luhansk.

Kosovo has also been referred to as being in a frozen conflict more than 20 times, the fifth most frequent state to be mentioned as such. Its declaration of independence from Serbia in 2008 remains unrecognized by Serbia and several other states, leading to persistent tensions that have repeatedly erupted into violence. The situations in South Sudan, Israel–Palestine, and Cyprus were occasionally referred to as frozen conflicts. However, at the time of writing, full-scale wars and major humanitarian crises are ongoing in both South Sudan and Palestine, such that the term ‘frozen conflict’ would currently seem inaccurate. States and territories that were mentioned as affected by a frozen conflict less than five times have been omitted from the figure. These include Antarctica, Afghanistan, Bosnia-Herzegovina, Djibouti, Ethiopia, Lebanon, North Korea–South Korea, Pakistan, Syria, and Western Sahara. Geographical regions that were mentioned additionally or alternatively to states as affected by frozen conflicts have also been omitted: the ‘Caucasus’ was mentioned 16 times, ‘Europe’ or ‘Eastern Europe’ (also ‘the OSCE region’) 16 times, ‘post-Soviet’ 15 times, the ‘Black Sea region’ 7 times, and ‘Balkan’ 4 times. The following regions were mentioned only once: Africa, Central Asia, Middle East, and the Sahel. Finally, 34 mentions of the term ‘frozen conflict’ could not be attributed to any specific state or region.

Findings are less clear as to which states mentioned frozen conflicts, as the term’s usage turns out to be widely dispersed. Ukraine comes out on top of the ranking, having used the term most frequently. Interestingly, 19 of 24 instances of Ukraine using the term happened before 2014 and do not refer to conflicts involving Ukraine but to the usual four regions: South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh. Moldova and Georgia used the term frequently, too, occupying places 2 and 3 in the ranking and referring mainly to their own affected regions. Surprisingly, Azerbaijan, which is affected by the fourth of the classical frozen conflicts, only used the term six times. This is less usage than by Turkey, Romania, or France, each of which used the term ten times or more despite not being affected by a frozen conflict. South Sudan, Israel, Palestine, and Cyprus, which have occasionally been referred to as involved in frozen conflicts, have never used the term themselves. States that used the term less than six times have been omitted from the figure. These were Albania, Andorra, Armenia, Austria, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Canada, Chile, Croatia, Cyprus, Estonia, Gabon, — S. 21 —Germany, Ghana, Hungary, Italy, Japan, Latvia, Lichtenstein, Lithuania, Luxembourg, Malaysia, Montenegro, The Netherlands, North Macedonia, Norway, Russia, Senegal, Serbia, Singapore, Slovakia, Tunisia, the United Arab Emirates, the United Kingdom, Uruguay, and the Holy See. Thus, a wide range of states have referred to frozen conflicts without being affected by one. Therefore, the conclusion can be drawn that the term ‘frozen conflict’ functions as a self-designator as well as an external designator.

In sum, the empirical findings clearly show that the term ‘frozen conflict’ has been used mainly by state officials since the late 1990s. Nevertheless, it gained momentum only in the 2000s and has been present in UN documents ever since. Furthermore, Georgia, Moldova, and Azerbaijan are the states mostly referred to as affected by frozen conflicts. It is therefore not only legitimate but also appropriate to focus on the four regions of South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh throughout this work. The results also suggest including their post-Soviet context in further analysis. However, the results are less clear about which states use the term. In fact, the states most often referred to as involved in frozen conflicts are not the same as the states that most frequently use the term, although there is some overlap with Georgia and Moldova being among the most frequent users.

III.Conceptual Approaches

After having studied the term’s background and use, this chapter next presents conceptual approaches to frozen conflicts. First, it will briefly focus on conceptual approaches and definitions offered by other disciplines than international law, including international relations, political science, and peace and conflict studies. Subsequently, legal approaches to frozen conflicts will be presented. This includes examination of the term’s use in legal discourse, suggested definitions and, finally, overlapping legal concepts.

A.Other Disciplines

Scholars of international relations, political science, and peace and conflict studies have suggested some definitions of frozen conflicts over the last decades, some of which will be outlined briefly as examples.⁹² In 1966, Holsti, a professor of political science, introduced a definition of ‘frozen disputes’, rather than ‘frozen conflicts’: — S. 22 —The category of ‘frozen’ dispute refers to those conflicts in which both sides have remained fully committed to their incompatible positions but where neither has yet dared to attempt resolution through accommodation, withdrawal, or military conquest.⁹³

As will be shown in Part IV, the parties’ commitment to incompatible positions is one of the core characteristics of frozen conflicts and is also reflected in their legal arguments. A more recent definition understands a frozen conflict as a state of conflict in which smaller acts of violence can occur, but largescale hostilities no longer do, and in which the parties have agreed on a ceasefire, but efforts to resolve the conflict and achieve lasting peace have failed.⁹⁴ Other examples state that frozen conflicts ‘in the post-Soviet space usually refers to four conflicts of an ethnic or ethnolinguistic nature that arose out of the collapse of the USSR in the early 1990s’,⁹⁵ and that ‘the violence stopped, but the underlying interests of the formerly warring parties have neither been abated nor addressed’.⁹⁶

In fact, these examples are working definitions that are better suited to the cases under discussion than to universal application. Scholars have also observed that the concept of frozen conflicts has remained undertheorized.⁹⁷ Accordingly, there seems to be an imbalance between its use as a term and its systematic analysis at a conceptual level. A suggestion for a more rigorous conceptualization has been offered by two international relations scholars, Jolicoeur and Campana, in 2009; they suggested that frozen conflicts exhibit the following characteristics:⁹⁸ the emergence of a secessionist movement following the dissolution of a communist state of federal type; the suspension of hostilities by the establishment of a ceasefire, usually reinforced by a peacekeeping operation; the victory of a secessionist party followed by the formation of a de facto regime, which in some cases has evolved into a state entity with ambiguous status; and the non-recognition of the victor by the — S. 23 —international community or, for South Ossetia and Abkhazia since September 2008, very limited recognition at the end of evolutionary processes both endogenous and exogenous to the conflict. However, Grant has criticized this definition for being geopolitically limited due to its reference to the ‘dismemberment of a communist state of federal type’.⁹⁹

The most comprehensive research on frozen conflicts has been presented by Klosek, Bahenský, Smetana, and Ludvík from the Peace Research Centre Prague: they rigorously conceptualized frozen conflicts and developed the ‘Frozen Conflict Dataset’, which identified 42 cases of frozen conflicts from 1946 to 2011.¹⁰⁰ Of these, 25 are ‘traditional’ interstate conflicts, 17 are conflicts between a state and a de facto state, and 23 were still ongoing in 2011.¹⁰¹ They presented their definition in a special issue of the Asia Europe Journal on frozen conflicts in 2019.¹⁰² The definition reads as follows:

a protracted, postwar conflict process, characterized by the absence of stable peace between the opposing sides. In frozen conflicts, core issues between the opposing sides remain unresolved, the dispute is in the forefront of mutual relations, and there is a looming threat of the renewal of violence. The conflict also remains highly salient in the domestic discourses of both policy makers and the general population.¹⁰³

They also state that a frozen conflict ends by one of three transformative mechanisms: violent thawing (e.g., Russia and Chechnya 1999), peaceful thawing (e.g., Sudan and South Sudan 2011), or conflict withering (e.g., Iraq and Kurdistan 2003).¹⁰⁴ Furthermore, they distinguish the concept from related ones in peace and conflict studies, such as enduring rivalries, strategic rivalries, and protracted conflicts, by emphasizing the specific postwar situation of frozen conflicts: the fact that they can only occur if there was a war at some point.¹⁰⁵ In 2021, the dataset was published in the Journal of Peace Research’s article ‘Frozen Conflicts in World’s Politics: A New Dataset’.¹⁰⁶ The approach is notable for being neither spatially nor temporally specific, because — S. 24 —the authors aimed to ‘overcome the prevalent understanding of frozen conflicts as a contemporary phenomenon in the post-Soviet space’.¹⁰⁷ Thus, they include conflicts that have not yet been analysed from this perspective.

Given its comprehensiveness and data-based approach, this conceptualization of frozen conflicts is promising as a future standard reference. Moreover, the development of an independent research project shows the growing interest in the topic and the necessity of its examination. This is also reflected in the establishment of a junior research group named ‘Frozen and Unfrozen Conflicts’ at the Leibnitz Institute for East and Southeast European Studies in Regensburg.¹⁰⁸ However, only time will tell whether the definition by Klosek et al. will prevail and become universally applied, potentially even beyond peace and conflict studies.

B.International Law

1.Use of the Term

Until now, only one thorough study on the etymology, use, and potential meaning of frozen conflicts in international law has been published, that by Grant in 2017.¹⁰⁹ To avoid repetition, his main findings on the use of the term in international law will be briefly outlined here and complemented with more recent findings. In international organizations’ practice, as of 2017, Grant found no appearance of the term in the Repertoire of the Practice of the Security Council nor in the Repertory of Practice of United Nations Organs.¹¹⁰ Although this assertion was still true of the latter publication in August 2024,¹¹¹ the collection of the UNSC now contains one match when searching for ‘frozen conflict’.¹¹² This reference appears in the reproduction of a speech by the High Representative for Bosnia and Herzegovina in 2021, who affirmed that ‘Bosnia and Herzegovina remained de facto a frozen conflict, where political leaders continued to pursue wartime goals and generate divisive narratives and nationalistic political agendas’.¹¹³ This document was also included in the empirical analysis on the use of the term in the previous chapter. Grant also observed — S. 25 —that UN officials seldomly use the term and that Francis Deng,114 UN Secretary General’s Representative on internally displaced persons, ‘appears to have been the first’115 in 2000. Yet, in the empirical analysis in this thesis 27 cases of UN officials using the term ‘frozen conflict’ have been identified by 2023. Among these were not only Deng’s successors as UN Secretary General’s Representatives on internally displaced persons, Kälin116 and Beyani,117 but also Secretary Generals Annan118 and Guterres119 in their reports to the UNSC. Grant further concluded that regional organizations had used the term more frequently than the UN, namely the Parliamentary Assembly of the Council of Europe, NATO, and the OSCE.120 This assessment is still accurate today: the term is found in the speeches and documents of the organizations in question.121 Interestingly, the OSCE uses the term even though it was the OSCE Chairman-in-Office who stated when briefing the UNSC in 2008 that frozen conflicts do not exist:

There really is no such thing as a frozen conflict. I think it is time to banish that term from the political lexicon and act quickly to settle any outstanding conflicts in our region. Let us forget all the talk about frozen conflicts and try instead to resolve them.122

Besides the term’s use in international organizations, Grant found that ‘frozen conflict’ had not been invoked in judgements, advisory opinions, or separate 114 UN Economic and Social Council, ‘Specific Groups and Individuals: Mass Exoduses and Displaced Persons, Report of the Representative of the Secretary-General, Mr. Francis Deng, Submitted Pursuant to Commission on Human Rights Resolution 200/53 (3 July 2000), UN Doc E/CN.4/2001/5/Add.2’ nn 5, 12, 16, 19.

115 Grant (n 4) 366.

116 UN Human Rights Council, ‘Report of the Representative of the Secretary-General on

the Human Rights of Internally Displaced Persons, Walter Kälin (9 February 2009) UN Doc A/HRC/10/13’ n 94 (e).

117 UN Human Rights Council, ‘Report of the Special Rapporteur on the Human Rights

of Internally Displaced Persons, Chaloka Beyani (2 April 2015) UN Doc A/HRC/29/34/ Add.3’ n 65.

118 UNSC, ‘Report of the Secretary-General on the Situation in Abkhazia, Georgia (28 Sep-

tember 2006) UN Doc S/2006/771’ (n 1) n 37.

119 UNSC, ‘Report of the Secretary-General on the Situation in Abyei (16 April 2019) UN

Doc S/2019/319’ (n 1) n 63.

120 Grant (n 4) 367.

121 See e.g. PACE, ‘The Role of the Council of Europe in Preventing Conflicts, Restoring

Credibility of International Institutions and Promoting Global Peace, Doc 15821 Report (5 September 2023)’ n 15; ‘Keynote Speech by NATO Deputy Secretary General Mircea Geoană at the Bucharest Forum: “Resilience, Pandenomics and the Great Acceleration”’ (NATO, 8 October 2020); Court of Conciliation and Arbitration within the OSCE, ‘Annual Report 2022’ (2023) 2, 17, 18.

122 UNSC, ‘Verbatim Records (26 September 2008) UN Doc S/PV.5982’ 4.

or dissenting opinions of the ICJ,¹²³ which is still the case today.¹²⁴ Neither has the term appeared in the jurisprudence of the ICTY, ICTR, or IRMCT Appeals Chamber.¹²⁵ Grant found an instance of the term being used in the South China Sea Arbitration before the Permanent Court of Arbitration, where the Philippines used the term in its argument in oral statements.¹²⁶ In ECtHR case law, the term ‘frozen conflict’ appeared in a case against Moldova and Russia, but it only consists of a report with the term in its title being cited and thus does not reflect the court’s practice.¹²⁷ The same is true for the formulation in the ECtHR’s case of Georgia against Russiathat ‘Russia was seen as the protagonist responsible for keeping the conflicts in the region frozen’,¹²⁸ which refers to the Report of the Independent International Fact-Finding Mission on the Conflict in Georgia (henceforth IIFFMCG).¹²⁹

Grant finally examined the use of the term by international law writers. He found that the term ‘frozen conflict’ has not been invoked by the International Law Commission (ILC), the International Law Association (ILA), or the Institut de Droit International.¹³⁰ Again, these findings still hold true at the time of writing.¹³¹ Moreover, the Max Planck Encyclopaedia does not contain an entry for ‘frozen conflict’.¹³² Nevertheless, as Grant noted,¹³³ the term does appear in other entries of the encyclopaedia: in the entry on ‘irredentism’, where it is explained how the nation-building process affects states with artificial or contested borders in particular such as in ‘the post-Soviet area, the frozen conflicts of Abkhazia, South Ossetia, and Nagorno-Karabakh’,¹³⁴ — S. 27 —and in the entries on South Ossetia¹³⁵ and Nagorno-Karabakh¹³⁶. Although not mentioned by Grant, the term also appears in the Max Planck Encyclopaedia’s entry on confidence-building measures (CBMs), which explains that

with the success of CBMs for Bosnia-Herzegovina, it is more likely that similar measures could be addressed in other post-conflict cases (or regions transiting out of conflict). ‘Frozen’ or low-intensity conflicts (such as Nagorno-Karabakh in Azerbaijan, Abkhazia and South Ossetia in Georgia, and Transnistria in Moldova) are matters of common concern for European security.¹³⁷

International law textbooks typically do not introduce frozen conflicts along with other types of conflict.¹³⁸ An exception might be the term’s occurrence in the publication of the International Committee of the Red Cross on contemporary practice in international humanitarian law (IHL), where Sassoli et al. discussed the temporal scope of the application of IHL and the corresponding difficulties: ‘frequently, most armed conflicts result in unstable ceasefires, continue at a lower intensity, or are frozen by an armed intervention by outside forces or by the international community’.¹³⁹ Finally, the term does not appear in the UN Treaty Collection.¹⁴⁰

2.Suggested Definitions

In contrast to the legal institutions and instruments discussed above, legal literature seems to use the term ‘frozen conflict’ more frequently. The most prominent and most widely quoted examples may be Bebler, Borgen, Mälksoo, Wittke, and the IIFFMCG.¹⁴¹ However, in general, the term is used in a descriptive vein and as a label that allows a variety of legal questions that characteristically arise in situations called frozen conflicts to be addressed.

These questions typically include secession, the use of force, regional interpretations of international law, and internal displacement. Few conceptual approaches or efforts have been made towards a legal definition. For instance, Bebler simply stated in his introduction to a work titled ‘Frozen Conflicts in Europe’ that the term’s use indicates that the violent stage of each of these conflicts ended in a stalemate and without a peace treaty.¹⁴² Another circumscription reads as follows: ‘It is common to call South Ossetia a “frozen conflict”, but this name implies certain features besides merely being protracted and unresolved: frozen conflicts typically concern claims to a separate territorial and sovereign status.’¹⁴³ And Paraskeva and Meleagrou noted in their book Cyprus at the European Court of Human Rights that

The term ‘frozen’ in connection with long-standing and unresolved confrontations indicates a lengthy passage of time with inevitable political social and legal changes which influence the negotiating positions of the parties and necessitate readjustments and compromise for the deadlock to be broken.¹⁴⁴

Some more examples of paraphrasing frozen conflicts in legal literature could be mentioned, as well as authors who reject the term and attribute it to other jargons or disciplines.¹⁴⁵ In contrast, Borgen’s oft-quoted study of the conflict in Transnistria on behalf of the New York Bar Association presents a more conceptual approach to frozen conflicts.¹⁴⁶ He studied the interplay of sovereignty, self-determination, and territorial integrity. Instead of suggesting a legal definition, he underlined the necessity of treating frozen conflicts as a legal phenomenon examined from a legal perspective, explaining that realpolitik has its limits and cannot fully explain or resolve these conflicts.¹⁴⁷

Yet, as mentioned above, the only suggestion for a precise legal definition of the term ‘frozen conflict’ so far has been offered by Grant.¹⁴⁸ In his paper, he identified a set of seven characteristics that pertains to all of the frozen conflicts in South Ossetia, Abkhazia, Transnistria, and Nargorno-Karabakh — S. 29 —‘and does not describe any situation that is not a frozen conflict’.¹⁴⁹ He followed the approach of what cognitive science calls the classic concept structure, whereby a case is either included in or excluded from the category of frozen conflicts. The seven characteristics are as follows:

(1) armed hostilities have taken place, parties to which include a state and separatists in the state’s territory; (2) a change in the effective control of territory has resulted from the armed hostilities; (3) the state and the separatists are divided by lines of separation that have effective stability; (4) adopted instruments have given the lines of separation (qualified) juridical stability; (5) the separatists make a self-determination claim on which they base a putative state; (6) no state recognizes the putative state; (7) a settlement process involving outside parties has been sporadic and inconclusive.¹⁵⁰

Despite the identification of these seven characteristics, Grant concluded that the concept of frozen conflict ‘remains at best at the edges of legal discourse’ and views it as useful not for legal analysis but for the ‘exposure of a strategy’.¹⁵¹ He bolstered this conclusion by three arguments: first, the juridical characteristics of frozen conflicts are highly diffuse and have little or no obvious connection to one another; second, the term is not neutral because it suggests that there is no hope for a resolution of the conflict, whereas in fact lawyers have the duty to find solutions and act as problem-solvers without prejudging the situation; and third, the use of the term implies legal fragmentation with a regional international law imposed by Russia that is at variance with general international law.¹⁵² Apparently, Grant has a very precise idea of what qualifies as a legal concept: it entails a single readily identified core phenomenon, is neutral, and is truly universal in the sense that it does not suggest any regional rules of international law. We certainly agree that frozen conflicts are a complex phenomenon that indeed exhibits a ‘mélange of juridical concepts’¹⁵³ and thus challenges binary legal thinking, in which objects, ideas, and phenomena are either included in or excluded from a category. However, we differ from Grant by thinking that this reveals as much about international law as it does about the phenomenon of frozen conflicts. Rather than excluding frozen conflicts from the legal realm due to their ambiguity, the more interesting project is to more closely examine this apparent mismatch — S. 30 —between international law and frozen conflicts. Thinking of frozen conflicts as a legal topic or concept in a wider sense than Grant supposes allows reflection on the legal dilemmas underlying frozen conflicts. It offers an opportunity to discuss how international law deals with ambiguity in general. Thus, the aim of this thesis is not to offer a legal definition of frozen conflicts. Instead, its aim is to study the relationship between international law and frozen conflicts holistically and recognize the ambiguity of the latter. Despite the occasional use of the term ‘frozen conflict’ by international law authors, this project has not yet been undertaken.

3.Overlapping Legal Concepts

When exploring conceptual approaches to frozen conflicts, the question arises as to which uncontested legal concepts are typically invoked together with that of frozen conflicts. The legal concepts applied to frozen conflicts are primarily international or non-international armed conflict, de facto state, and belligerent occupation. These concepts will repeatedly be revisited throughout this work. For now, it suffices to emphasize briefly that they are connected to frozen conflicts, perhaps even overlapping with the concept to a certain extent, but we consider these three concepts not to be interchangeable with frozen conflict. Consequently, the research that has been conducted on them does not render research on the concept of frozen conflicts redundant. Even if these concepts apply to the same situation, they have different qualities and functions. It is worthwhile to briefly distinguish frozen conflicts from these three concepts.

The first overlapping legal concept is international or non-international armed conflict. International law is relatively straightforward when classifying conflicts. Leaving aside the reasons and motives for the conflict, jus ad bellum, jus in bello recognizes only two distinct types of conflict: international and non-international armed conflicts.¹⁵⁴ A similar distinction is made by international criminal law, which distinguishes between war crimes during international and non-international armed conflicts and defines these in the same manner as IHL.¹⁵⁵ A potential third category called ‘protracted armed conflict’, mentioned in article 8 (2) (f) of the Rome Statute,¹⁵⁶ was discussed by some — S. 31 —writers but finally rejected.¹⁵⁷ Considering the complexity of reality and the manifold situations of violence in the world, this dichotomy appears to be an oversimplification — in contrast to the much-lamented fragmentation of international law. Many contemporary armed conflicts challenge these categories due to their combination of non-international armed conflict within a state, international interventions, and spillover effects on other territories.¹⁵⁸ Of course, frozen conflicts can be classified within the categories of international and non-international armed conflict to identify what law is applicable. Typically, frozen conflicts involve aspects of international as well as non-international armed conflict.¹⁵⁹ However, frozen conflicts fit these two categories only as long as there is an armed conflict at all. The ‘hot’ phases of war that all frozen conflicts have gone through¹⁶⁰ meet this condition. But, as will be outlined later, a frozen conflict is mostly characterized by low-intensity violence and thus stands in contrast to an armed conflict, whether international or non-international. Restricting the analysis to the prevalent categories of international and non-international armed conflict would overlook precisely what makes frozen conflicts unique: the state of frozenness and the absence of obvious armed conflict.

The second overlapping legal concept is that of the de facto state or de facto regime. According to the Max Planck Encyclopaedia of International Law, de facto regimes are entities that have existed for long periods, claim to be states or governments, and control defined territories, but lack recognition, at least by the majority of states, such as North Korea, Taiwan, Northern Cyprus, South Ossetia, and Abkhazia.¹⁶¹ This definition includes the claim to be a state or a government, but other definitions distinguish between de facto states and de facto regimes. Accordingly, ‘the entity that constitutes a de facto state seeks full constitutional independence and widespread recognition as a sovereign state, for example the Republic of Somaliland and the Republic of Kosovo’.¹⁶² In contrast, an entity that constitutes a de facto regime ‘aspires to — S. 32 —be recognized by the international community as being the official government of an already existing state’.¹⁶³ The difference between a de facto regime and a de facto state, therefore, is in their ambition: whereas the de facto state aims to be a state through secession or independence from the parent state, the de facto regime seeks recognition as the legitimate government without necessarily violating the parent state’s territorial integrity.¹⁶⁴ Writers addressing frozen conflicts normally include a description of a de facto state.¹⁶⁵ As will be outlined later, the concept of a de facto state, the term that will be used henceforth, is one of the core characteristics of frozen conflicts. Accordingly, if no de facto state remains, following absorption by the parent state,¹⁶⁶ we consider a territory to be no longer involved in a frozen conflict. This scenario occurred when Nagorno-Karabakh was absorbed by Azerbaijan in 2023. Nevertheless, the concepts of de facto state and frozen conflict are not interchangeable. Some de facto states have never been involved in a frozen conflict, for instance Somaliland.¹⁶⁷ The concept of the de facto state is important, even in cases of obligatory non-recognition following a UNSC decision, to draw the consequences of the fact that a territory is fully governed by the entity.¹⁶⁸ Such entities are partial subjects of international law and are thus holders of some rights as well as some obligations.¹⁶⁹ Interestingly, the concept of the de facto state blurs the binary legal logic of states versus nonstates. Similar to frozen conflicts, it softens rigid categorizations.

The third overlapping legal concept is belligerent occupation. As will be explained in greater detail in Part III, regions affected by frozen conflicts are characterized by a high military presence. According to article 42 Hague Convention IV,¹⁷⁰ a territory is considered occupied when it is placed under the authority of a hostile army without the consent of the territorial state to which — S. 33 —it belongs. Although no international tribunal has affirmed belligerent occupation in the cases of South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh, the situations have sometimes been qualified as such. The Rule of Law in Armed Conflicts (RULAC) online portal run by the Geneva Academy classifies the parent states, Georgia and Moldova, as occupied by Russia and, up to 2023, classified Azerbaijan as occupied by Armenia.¹⁷¹ However, literature on belligerent occupation barely discusses the frozen conflicts of South Ossetia, Abkhazia, Transnistria, or Nagorno-Karabakh.¹⁷² It can be inferred that frozen conflicts are not textbook examples of belligerent occupation. Two aspects might be atypical: first, a belligerent’s control over another actor can be regarded as occupation by proxy when a subordinated de facto administration controls the territory.¹⁷³ Second, the occupation is combined with claims for self-determination and the separation of the territory.¹⁷⁴ Therefore, the question of belligerent occupation plays a role in frozen conflicts but it is not congruent with the latter concept.

IV.Conclusion

Investigations of the term ‘frozen conflict’ produce various findings. First, it remains unclear when and by whom the term was first introduced. Although some trivial references have been found before the 1990s, the consensus is that its origin is in that decade. The historical context was the dissolution of the Soviet Union in 1991 and the outbreak of wars in regions that sought independence. These wars quickly entered a less violent phase and thus were described as frozen conflicts. Many authors speak of the emergence of the term in the 1990s without specification. The empirical analysis of UN Documents allows for some specifications: the analysis shows that the term emerged at the very end of the 1990s but suggests that its use gained momentum only in the 2000s, when state officials and international organization officials started using it frequently. Therefore, the emergence of the phenomenon at the beginning of — S. 34 —the 1990s has to be distinguished from the emergence or conventionalization of the term in the late 1990s and 2000s. This is because only after the passage of time were these conflicts perceived as frozen and the metaphor became conventionalized. Furthermore, it was only in the 2000s that Russia changed its policy towards surrounding states. The empirical findings also show that the four conflicts in South Ossetia, Abkhazia, Transnistria, and Nagorno-Karabakh were the first to be labelled frozen and are still the ones most frequently referred to as such. Nevertheless, over time, the term’s application has expanded to other conflicts beyond the post-Soviet context.

Second, despite the frequent use of the term ‘frozen conflict’, few conceptual approaches have been attempted. To fill this research gap, the Peace Research Centre Prague offered a rigorous conceptualization in 2019 and identified 42 cases of ‘frozen conflicts’ from 1946 to 2011. Whether this definition becomes a standard reference in peace and conflict studies and beyond remains to be seen. International law has shown only marginal interest in conceptual approaches to frozen conflicts. One reason may be that the term rarely appears in legal instruments. Only international legal writers, state officials, and international organization officials have employed it occasionally. So far, only Grant has examined whether the term ‘frozen conflict’ has any legal content and presented a legal definition. He concluded that the concept is not relevant to legal analysis.

Finally, it has been acknowledged that looking at the situations called frozen conflicts from a legal perspective typically brings other legal concepts into play. These are the undisputed legal concepts of international or non-international armed conflict, de facto state, and belligerent occupation. These concepts may overlap with frozen conflicts and lack a clear distinction from them. However, they are not interchangeable with the concept of frozen conflict. Moreover, it is the very ambiguity of the concept of frozen conflict, including its overlapping with other concepts, that characterizes the relationship of international law with frozen conflicts. Closer examination is required not only of the phenomenon but also of the discipline’s categories.