To Be or Not to Be : The Ontological Predicament of State Creation in International Law

International law is classically based on a system of states whose members it attempts to identify by virtue of their effectiveness, their recognition by other states, and their creation in accordance with the rules of international law. In this article, I illustrate the indeterminacy of these three dimensions and argue that assessments of individual state creations are instead necessarily based on blunt, but silent, ontological commitments to any potential state’s full presence or absence. While the ‘great debate’ between declaratory and constitutive doctrines of recognition has emphasized, but not finally determined, the ontology of the state, attempts to find compromises between material effectiveness and constitutive recognition as well as the turn towards the proper legal regulation of state creation have only bracketed and invisibilized its decisive role. The deconstruction of state identifications reveals an essentially empty state ontology that confronts scholars and practitioners with the predicament of having to ultimately presuppose any particular state’s existence or its absence. This not only allows for reflecting on the stakes of individual assessments but also shows how all state identifications inevitably reproduce the hegemonic image of an exclusive and neatly delineated state system that brings its unruly fringes under control time and again. If we give the state a transcendental status, it disappears from the world; if we see it merely as a set of empirical attributes, it disappears in the world. – E. Ringmar, ‘On the Ontological Status of the State’1 * European School of Political and Social Sciences, Université Catholique de Lille, Lille, France; University of Helsinki, Helsinki, Finland. E-mail: janis.grzybowski@univ-catholille.fr. The author would like to thank Martti Koskenniemi, Anne Peters, Jens Bartelson, Wouter Werner, Ukri Soirila, Farrah Hawana and two anonymous reviewers for their helpful comments on earlier drafts of this article. He would also like to thank the Max Planck Institute for Comparative Public Law and International Law for a scholarship to work at its premises in Heidelberg, Germany, as well as the Swiss National Science Fund for earlier support lent to this research. 1 E. Ringmar, ‘On the Ontological Status of the State’, 2 European Journal of International Relations (1996)

If we give the state a transcendental status, it disappears from the world; if we see it merely as a set of empirical attributes, it disappears in the world.
-E. Ringmar, 'On the Ontological Status of the State' 1

Introduction: International Law and State Ontology
Despite its ostensibly theoretical nature, the ontological question of what states actually 'are' is in fact located at the very heart not only of the international legal order but also of current controversies about practical legal questions.A case in point is Palestine's bid to put itself under the jurisdiction of the International Criminal Court (ICC) and even become a proper state party to the Rome Statute. 2 While the ICC rejected Palestine's request for opening investigations in April 2012 because it had doubts about its status as a state, 3 it eventually welcomed it as a new member in 2015, after the United Nations (UN) General Assembly had granted it the status of a 'non-member observer state'. 4Behind the controversy over Palestine's accession to the ICC lurks the question of Palestine's state status, 5 which in turn revolves around how states are determined in international law.While supporters of Palestine's accession had argued for a long time that it was a state due to its legal entitlement as a self-determination unit and widespread recognition by other states,6 sceptics held that the fragmented control exercised by the Palestinian National Authority under Israeli auspices lacked, among other things, the required effectiveness and independence for Palestine to qualify as a state. 7Similar debates concern the rights and obligationsand, therefore, the status -of other contested 'states', from Kosovo to South Ossetia and Abkhazia and from Somaliland to Taiwan.Far from being detached from practice, the questions of what the state is and how individual states are identified inform some of the most entrenched territorial disputes in the world and promise to retain their significance in conflicts from the post-Soviet space to the Middle East.Any identification of concrete states implies some ontology of the state as an at least implicit understanding of its being.Yet while there is a wide agreement that the state is composed of the elements of territory, population and government, 8 usually complemented by the capacity to enter into relations with other states, 9   individual states is a much more controversial business.As I argue in this article, the standard criteria of effectiveness, recognition and legality are co-dependent and indeterminate.From an international law perspective, the state is an entity neither located entirely within law nor entirely outside of it, despite all attempts to reduce it to one or the other dimension.Indeed, 'there are few issues that better epitomize the combination of law, fact and power more enigmatically than the question of statehood', 10 and, yet, '[i]t is in statehood that one seeks to find a seamless amalgam of legal doctrine and social reality'. 11By state ontology, then, I refer neither to its formal legal status nor to the material features of its social reality but, rather, to what these dimensions are meant to capture and what holds them together -that is, the very being of the individual state in and beyond both law and extra-legal reality. 12uriously, however, state ontology has played a diminishing role in discussions of state creation in international law since the days of the 'great debate' between the constitutive and the declarative position. 13As this classical controversy over grounding the state either on internal order or external recognition proved irresolvable, today it is often regarded as outdated and misleading. 14Instead, international law has turned to the properly legal regulation of state creations, enabled by the '"objective" criteria of statehood' that would guard against both the politics of recognition and the politics of claiming legal consequences on the basis of supposedly evident facts. 15Yet the focus on legal rules to determine what should be acknowledged as a state creates problems in its own right; after all, entitlements to statehood are not tantamount to actual state creations, while the supposed illegality of state creations does not guarantee their actual failure either.Hence, the persistent resort to classical state criteria and recognition, as in the arguments about the formal status of Palestine and others.
Against this background, I propose to reconsider this classical ontological problem through the lens of an analytical perspective that avoids the pitfalls of both the traditional attempt to approach it head on and the modern attempt to effectively circumvent it.Although it directly engages with legal arguments, the article takes an 10 French, 'Introduction', in D. French (ed.), Statehood and Self-Determination in International Law (2013)  external vantage point on international law, rather than a perspective of international law.The goal is not to replace legal analysis but, rather, to complement it by a reflective perspective on its silent presuppositions and argumentative constraints. 16nstead of taking state ontology literally, as an essence to be fully disclosed or a mere illusion to be strictly avoided, I suggest that it operates as a deep discursive structure of international law beneath the surface of state effectiveness, recognition and legality.My aim is not to discover what the state really is but, instead, what state ontology does by compelling legal arguments to provide determinate answers as to whether some entity is or is not a state.
More specifically, I argue that while the ontology of the state is essentially empty -a gesture of grounding where there is no ground -it is discursively decisive because it allows for the representation of the contradictory 'multiplicity and fractionality' of state dimensions in the form of a 'virtual singularity' of state presence or absence. 17By way of a deconstructive analysis, I illustrate how the three conventional state dimensions are used to construct state identities in legal arguments, although none of them provides for a harmonious 'single story' of clearly successful state creation. 18ogically bereft of any objective ground, concrete state identities can thus only be presupposed.Indeed, they must be presupposed because international law works on the assumption that states really exist and that they can be differentiated from each other as well as from non-state entities.This does not mean that determinations of states are 'purely political', since any assessment must still be articulated and rendered plausible in the vocabulary of international law.However, it does mean, first, that legal arguments and political commitments are indeed enmeshed, which is an old insight from critical legal studies 19 and, second, that they operate against an enabling and constraining horizon of state ontology as their fundamental condition of possibility, a horizon that, finally, international law reproduces in turn precisely by acting upon it in practice.By imposing a logic of state presence or absence, international law eclipses the imagination of a genuine non-state world order. 20The standard modern map of state territories is an expression of this ontological horizon of the state system -exclusive and neatly ordered.Law, supra note 12, at 50.For a classical deconstructive analysis of international law, see M. Koskenniemi,  From Apology to Utopia: The Structure of International Legal Argument (2005). 19See, e.g., Koskenniemi, supra note 18.   20   This also entails a caveat for my own argument; since I look explicitly into how international law determines individual state creations and reproduces the state system by way of imposing a silent state ontology, I cannot at the same time consider the potential for imagining a radically different non-state world order.
Western Sahara or secessionist Abkhazia, indicate doubts about the full presence or absence of certain states.
I first discuss in the second section of this article how the declaratory position once raised the ontological question as a means for promoting equality by binding (non)recognizing states to determinate standards of simple effectiveness -and how this turn to facts failed to determine states objectively since the waxing and waning reality of material political orders does not easily translate into the strict legal status of state presence or absence.In the third section, I turn to various attempts to bracket the ontological question by finding a compromise between the declaratory and the constitutive positions and illustrate how they tend to collapse ultimately into either one or the other.The fourth section discusses how the move from the identification of states to the regulation of their creations has invisibilized, but not displaced, the ontological question, which still haunts assessments of state creation.Throughout all these sections, I illustrate the problems with identifying states in international law by deploying examples of contested or contestable state creation, from Somaliland to Western Sahara, which show that none of the three state dimensions are objectively determinate and that assessments are instead presupposed before -and silenced behind -legal arguments.I conclude by emphasizing the role of the ontological predicament in legal assessments for the preservation, but also the flexibility, of the state system and highlight the opened space for participating consciously and reflectively in the rearrangements of 'states' in international law.

Identifying the State between Recognition and Effectiveness
As conventionally retold, with the ascendency of positivism in 19th-century international law -and the imperialist expansion of a Western-modelled global system of states with its (semi)-colonial periphery -the identification of sovereign states increasingly became a problem in its own right. 21Following the decline of natural law and dynastic 'legitimism', the question of state creation came to be defined in doctrinal terms by the fundamental opposition between the constitutive doctrine that 'through recognition only and exclusively a State becomes an International Person and a subject of International Law', 22 and the declaratory position that if a 'State, … exists in fact, [it] must exist in law' too. 23Since then, approaches to the state as either effective political order or as legally recognized status have comprised two fundamental dimensions of the state concept in international law.
It is widely understood today that neither theory is able to provide for an objective identification of states on its own, but neither the reasons for, nor the implications of, this failure are entirely clear.As I suggest, the argument has classically run in a ritual circle around the question of state ontology.Recognition is an arbitrary political act; hence, effectiveness should determine states.Yet effectiveness is empirically contestable and indeterminate, so recognition is required to interpret effectiveness, although whether recognition in turn either reflects or else really creates states also remains indeterminate.The underlying presumption that structures this debate is that states do exist and that they can be individually identified.The decisive role of this presumption needs to be drawn out.

A Anchoring International Law in Facts: From Constitutive to Declarative Recognition
The factual notion of the state has been codified in the 1933 Montevideo Convention on the Rights and Duties of States. 24While there are differences among scholars in emphasizing particular criteria as decisive, 25 the content and purpose of the factual notion of the state is rather straightforward.The state is regarded as an effective entity that exists before and irrespective of whether it is properly recognized or otherwise legally sanctioned by international law.Even today, 'effective control of territory and people remains the hallmark of what constitutes a state'. 26he declarative theory is directed against the classical constitutive doctrine and its imperialist heritage, which bluntly maintained a hierarchy between different state-like entities that is radically at odds with the commitment of modern international law to justice and equality among states. 27If international law was merely an instrument for maintaining an international hierarchy by means of regulating legal status, then legal recognition would be, as James Brierly puts it, 'an attorney's mantle artfully displayed on the shoulders of arbitrary power'. 28Lassa Oppenheim's elaborations illustrate the nonchalance with which recognition could be separated from the question of proper state identification.The already-cited stipulation that '[t]hrough recognition only and exclusively a State becomes an International Person and a subject of International Law' follows a revealing clarification that '[t]here is no doubt that statehood itself is independent of recognition' and that 'International Law does not say that a State is not in existence as long as it is not recognized, but it takes no notice of it before its recognition'. 29Ti-chiang Chen wrote passionately against this function of recognition and drew the conclusion that the constitutive doctrine 'may serve the purpose of Machiavellian statesmen' well since it 'provides them with a justification for ignoring the existence of other entities and denying them rights under international law'. 30f international law was to live up to its universalist aspirations, so the declarative position claimed, it would have to acknowledge and incorporate all states on an equal basis: 'If an entity in fact exists, the refusal to treat it in accordance with international law would incur the same risks and perils as would be incurred had the treatment been refused to a recognized entity.' 31 The principle of effectiveness thus serves as a legal basis for international law as a system regulating relations between actually existing states so that the recognition of effective states is 'not creative, but declaratory'.32 Ex factis jus oritur.Still today, it is argued that 'the status of an entity as a State is, in principle, independent of recognition'.33 Moreover, for a declarativist, '[i]t is further clear that a sovereign state cannot be created through recognition by other states' since the recognition of legal personality presupposes an effective entity: 'No amount of recognitions can supply the lack of the fulfilment of the requirements laid down by international law.Recognition, in such a case, is simply ineffective in law.' 34 Hence, international law has been seen as a 'realistic legal system' that 'provides that only those claims and situations which are effective can produce legal consequences'.35 Thus, 'if a State emerges from secession, it will be able to claim international status only after it is apparent that it … controls a specific territory and the human community living there'.36 Yet, while the criterion of actual effectiveness seems to promise an objective ground on which to evaluate the success or failure of state creations and thus avoid the political dilemma of constitutive recognition, the identification of effective states is not as readily apparent as the declaratory position assumes.Effective control is a 'matter of degree' 37 -'there is no bright line between effective and ineffective' 38 and 'no specific requirements as to the nature and extent of this control, except that it includes some degree of maintenance of law and order and the establishment of basic institutions'.39 It is widely agreed, and even Chen admits, that the conditions for statehood 'can only be stated in general terms' and that 'it is obviously impracticable, for instance, to prescribe any fixed standard of [territorial] size and population'.40 Effective government control is particularly hard to measure, as illustrated in cases of so-called de facto states discussed below.But, if so, the view that a state, 'once having satisfied certain objective tests, ipso facto becomes a person in international law' 41 has to be rejected as 'epistemologically naïve'.42 Clearly, there must be some 'materiality' to states, but this does not justify a 'materialism' according to which the significance of material features is objectively intelligible.43 The question then is how one can categorize waxing and waning orders of material power as states or non-states at all.

B Effective but Unrecognized: The Status of Entities Not Formally Recognized as States
The challenge of so-called de facto states or regimes, from the Turkish Republic of Northern Cyprus (TRNC) to Somaliland, epitomizes the lasting significance of claims to statehood based on effectiveness as well as the problematic implications of upholding the constitutive doctrine of recognition. 44Although for different reasons not widely recognized as states, these de facto entities are said to 'internally' measure up to effective statehood. 45Furthermore, the fact that they often entertain various diplomatic relations with recognized states and conclude treaties shows their capacity to enter into external relations. 46he case of Somaliland is illustrative because, in contrast to other cases, it is often seen as not having violated any pre-emptory norms of international law in its (attempted) secession. 47Somaliland rebels and clan leaders declared the territory independent in 1991 and, since the mid-1990s, the polity was on a path of stability in comparison to the rest of Somalia. 48It has therefore been frequently discussed by international law scholars as a case of unrecognized, but effective (and not illegal), state creation.Some have argued that 'Somaliland meets the objective criteria of statehood' and 'is a state that simply lacks recognition', 49 but others maintain that even though 'the notion of the de facto regime has been pressed to its ultimate … Somaliland is not yet a State'. 50Two reasons are usually cited in favour of the latter and dominant conclusion.For one, effectiveness itself is apparently less evident as sometimes presumed.James Crawford, for instance, who maintains that 'the status of an entity as a State is, in principle, independent of recognition' and who acknowledges that '[i]t is said that from 1993 onward, a Somaliland government has functioned effectively in the territory', still counts Somaliland under 'unsuccessful attempts at secession', apparently also because 'the stability of Somaliland is fragile' after all. 51Assessments of national courts and various UN agencies concerned with the question of Somaliland's stability have also come to divergent conclusions. 52his disagreement illustrates the difficulty in determining effectiveness in practice.The second reason why it is usually not considered a state is that 'no third State has recognized the independence of Somaliland' -but this assumes that recognition determines whether an entity is or is not a state, after all. 53owever, if questions of fact require interpretation and the competent authorities in international law to decide upon any particular entity's claim to statehood are the existing states -although courts may occasionally play a role too -then the constitutive doctrine is essentially reinstalled.Indeed, as Crawford warns, 'if there are no such [statehood] criteria, or if they are so imprecise as to be practically useless, then the constitutive position will have returned, as it were, by the back door'. 54Yet this also implies, contra the view of Crawford and others, that we cannot conclude either that 'Somaliland is still not a State, not even a non-recognized one'. 55Just as there are legitimate doubts about the presence of its statehood, there are also legitimate doubts about its absence.Assessments in favour of, or against, Somaliland's statehood, such as those cited here, cannot be objectively grounded on either (non)-recognitions or measures of effectiveness, but rely on a blunt ontological commitment to Somaliland's (non)-existence as a state, if, to be sure, they are informed by arguments as to the determining role of effectiveness and recognitions.Since the dominant practice is to not regard Somaliland as a state, the entity is, in contrast to Somalia, not a member of the UN and arguably lacks the legal protection of other states. 56uch of this controversy and ambiguity applies to most supposed de facto states or regimes, leaving aside for the moment questions of illegality, as posed by cases such as the region of Nagorno-Karabakh or the TRNC. 57Biafra, for instance, was a controversial case of effective state creation, which remained unrecognized except by Tanzania, Gabon, Zambia and Haiti. 58While the dominant assessment is that 'Biafra was not a State', 59 others have argued that '[t]here is no doubt that Biafra met the requirements of population, government, permanence, and a reasonable measure of effectiveness' before being defeated and reincorporated by the government of Nigeria. 60Arguably, Biafra had remained in a permanent state of war and could never claim to have established lasting control, 61 but whether or not control lasts is essentially an ex post judgment.It is not entirely clear how the 'ultimate success' of a factual secession should be objectively established on the basis that the 'former regime … can no longer successfully restore its authority'. 62Cases in point are the 'frozen conflicts' over Abkhazia or Nagorno-Karabakh, which Georgia and Azerbaijan, respectively, have sought to bring under their control by military force since the early 1990s and which they still regard as parts of their territories.Despite the repeated failure of these military campaigns, both de facto entities remain largely unrecognized.
However, ignoring the state features of relatively effective entities has legal and political consequences.Even Taiwan remains vulnerable due to non-recognition, although the end of the Chinese civil war and the achievement of sufficient state effectiveness are hardly contested. 63The conventional conclusion is that 'Taiwan is not a State because it still has not unequivocally asserted its separation from China and is not recognized as a State distinct from China'. 64Yet this puts on a fragile basis the argument that we should assume the existence of 'a cross-Strait boundary for the purposes of the use of force' 65 and, thus, a prohibition for the mainland People's Republic of China (PRC) to bring Taiwan under control militarily. 66Unsurprisingly, the PRC regards any potential use of force against Taiwan as an internal affair. 67The status of the state entails legal consequences, including for the 'internal' use of force, and the ontology of the state compels us to think in terms of mutually exclusive state presences. 68In this light, the notion of de facto regimes implies a degradation since it is not at all clear that recognitions signify statehood, as I will further explain in the next section.

Bracketing State Ontology? Compromises on State Identification
by virtue of certain rules and practices', then this raises the question how recognition can be an act of legal interpretation rather than political discretion. 69Compromise between the acknowledgement that states are social entities not simply made by legal recognition and the insistence that it must be legal assessments that distinguish states from other entities come in either a theoretical or a pragmatic form.Ultimately, however, both approaches fail to bracket the fundamental ontological dilemma between the declarative and constitutive positions.

A A Theoretical Way Out? The State as Legally Recognized Fact
The widespread uneasiness of understanding states to be products of fact only has been paradigmatically expressed in Hans Kelsen's seminal critique of the declarative position and his insistence that 'there are no absolute, directly evident facts, facts "in themselves," but only facts established by the competent authority in a procedure prescribed by the legal order'. 70The state, then, is a 'legal fact'. 71While, in contrast to the classical constitutive doctrine, Kelsen also insisted that '[t]he legal act of recognition is the establishment of a fact' and, thus, 'cognition rather than re-cognition', his compromise leaves open how and by whom individual states can be determined without resorting either to supposedly objective facts, even if evaluated from a 'legal' perspective, or to decisive recognition by other states. 72elsen himself returned to a constitutive position in reformed shape.Since states 'alone are the authorities empowered by general international law to decide the question' of whether or not any entity meets the statehood requirements, he concluded in what resembles Oppenheim's aforementioned classical formula: 'Before recognition, the unrecognized community does not legally exist vis-à-vis the recognizing state.' 73 Although the insistence that states are legal subjects and not merely social facts has been largely endorsed, Kelsen's neo-constitutive position and, in particular, the relative notion of a state existing merely in relationships with states that have recognized it, has been rejected in the literature because of the supposed 'absurdity of conceiving a State as existent and non-existent at the same time'. 74This critique nicely illustrates the ontological desire to decide the question of being for good.
Hersch Lauterpacht, another important proponent of a theoretical compromise, ran into similar problems with his formula that recognition was 'declaratory of facts and constitutive of rights'. 75If there would be a 'duty to recognize' if a state really existed and a prohibition of 'premature recognition' if it did not, this would presuppose 69 Crawford, supra note 7, at 5. 70 Kelsen, supra note 25, 606. 71Ibid.(emphasis added). 72Ibid., at 608. 73 Kelsen, supra note 25, at 608.See also Dugard and Raič, 'The Role of Recognition in the Law and Practice of Secession', in Kohen, Secession, supra note 55, 94, at 96-97. 74Chen, supra note 23, at 41; see also Crawford, supra note 7, at 21; Dugard and Raič, supra note 73, at 97.For positions in favour of the relative character of recognition, see G. Schwarzenberger, A Manual of International Law (1967), at 73-74; Oppenheim, supra note 22, at 124, para.72. 75Lauterpacht, supra note 25, at 75. that states could be discerned as objective facts or otherwise by reference to an evident truth not determined by state recognition. 76The distinction that '[r]ecognition declares the existence of a physical, not of a legal, phenomenon' only superficially conceals the basic dilemma that states can only be either identified by recognition or be independent of it -there is no logical middle ground on the underlying premise that states either exist fully or not at all. 77Since Lauterpacht considers the legal assessment of the effectiveness of any state-like entity to be objective, his explanation for disagreement is not that there might be legitimately diverging interpretations but, rather, that some states act in bad faith by deliberately misrepresenting the facts. 78But to say that effective statehood as 'a physical fact … is of no relevance to the commencement of particular international rights and duties until by recognition -and by nothing elseit has been lifted into the sphere of law' is simply a restatement of the constitutive position. 79ellingly, Lauterpacht recognizes the problem, yet proposes not a legal, but a political, solution: the 'collectivization of the process of recognition'. 80By achieving 'a high degree of political integration of the international community in form of an international organization of States', recognition could be decided upon in each case by an 'appropriate majority'. 81Yet if there was some mechanism by which states could be made to agree in their assessments about an entity's state character, there would be no formally contested cases of state creation to begin with. 82More importantly, that at least unanimous recognition effectively makes states is still a voluntaristic assumption that is undermined not only by supposed de facto states but also by fragile recognized states, as discussed below.Hence, the 'great debate' on whether states are ultimately identified by virtue of a certain objective effectiveness or by formal recognition remains theoretically unresolved.

B The Pragmatic Compromise: Conflating Effectiveness and Constitutive Recognition
The great debate has been regarded as misleading and outdated not only because neither doctrine actually provides for an autonomous role of international law in determining statehood 83 but also because, supposedly, 'the differences between declaratory and constitutive schools are less in practice than has been depicted'. 84This sets the stage for the now common perspective on the cumulative effect of state effectiveness 76 Ibid., at 7-12, 73. 77Ibid., at 75. 78 Ibid., at 76. 79 Ibid., at 75. 80 Ibid., at 67. 81 Ibid., at 68. 82 Arguably, the admission to the United Nations (UN) presents a process of collective decision making, as foreseen in Art.4(2) of the UN Charter.However, Art. 4(1) requires new members to already be states, as pointed out by Frowein, supra note 44, at 5. 83 Crawford, supra note 7, at 5. 84 Ibid., at 28. and recognition, which is content with the view that '[p]robably the truth lies somewhere between these two theories'. 85Crawford, for instance, although he rejects the strict constitutive position, argues that if 'an entity is recognized as a State [this] is evidence of its status' and 'where recognition is general, it may be practically conclusive'. 86This compromise is not new, of course, and was already succinctly formulated by Quincy Wright in 1955: In principle, recognition is declaratory of facts which under international law constitute the status in question, but, because these facts are usually so uncertain that international law cannot be applied automatically, and because states may promote their policies by recognizing facts not yet established or by refusing to recognize facts which are at the moment established, recognition may, in practice, be in considerable measure constitutive of that status. 87milarly, in his 'synthesis', Michael Schoiswohl argues that while in undisputed cases recognition is declaratory, in 'borderline cases' it is 'evidentiary' and assumes a 'semi-constitutive' function. 88Yet the fundamental problem with having it both ways is still how to determine any individual instance of complete state creation at any moment in time, as international law ultimately demands.To begin with, if there is no third category added -that is, cases that do not even become states when recognizedrecognition is simply constitutive since it elevates every entity to state status.Second, if recognitions 'consolidate … effectiveness … [a]nd this consolidation in turn means reaching the legal threshold of statehood', the determination of when the threshold is crossed is still a question of either fact or recognition. 89If the effect of recognition depends on whether or not the entity in question is just effective enough to become a state upon recognition, this lower threshold has to be determined outside of international law by an empirical measurement -just as the threshold of effectiveness from the simple declarative perspective discussed above.Third, in order not to return to a classical constitutive position, it is often added that recognition has to be granted in accordance with international law. 90Much along the lines of Lauterpacht's compromise, '[a]t least where the recognizing government is not acting in a merely opportunistic way, recognition is important evidence of legal status'. 91Yet on which basis can it be established whether a state acts in good or in bad faith, if state effectiveness cannot be regarded as a readily apparent objective benchmark? 92hus, however plausible it might be that in many cases of state creation recognitions partly reflect and partly shape the 'facts on the ground', the mix of the two logics does not provide for any better approach to determine individual states at any point in time, as required by the ontological premise that any particular state is ultimately 85 Starke, supra note 12, at 133. 86 Crawford, supra note 7, at 27; see also Peters, supra note 37, at 5. 87 Wright, 'The Chinese Recognition Problem', 49 AJIL (1955) 320, at 324. 88 Schoiswohl, supra note 49, at 39-47. 89Peters, supra note 37, at 5. 90 Schoiswohl, supra note 49, 42. 91Crawford, supra note 7, at 93 (emphasis added). 92Koskenniemi, supra note 18, at 279. either present or absent.Thus, irrespective of popular formulas for compromise, the state is still torn between the two foundational dimensions of material effectiveness and recognized status.

C Does Recognition Signify Statehood? State Fragility and Contested Recognitions
In light of fragile compromises, it might seem that grounding state identities on recognition for practical purposes at least promises some legal certainty.Yet even leaving aside the cases excluded by such a neo-constitutive view, such as Somaliland or Taiwan, relying on recognition is also problematic for other reasons.To begin with, even where recognition has an arguably real effect on the development of a 'state' in terms of its material effectiveness, this effect depends on other factors on the ground.For instance, the recognitions of Slovenia, Croatia and Bosnia-Herzegovina in the wake of the dissolution of Yugoslavia in early 1992 played out very differently in the respective processes of 'state creation'. 93While Slovenia was cut loose after marginal skirmishes, Croatia became involved in an internal civil war with a sizeable Serbian minority that sought to either preserve the Yugoslav federation or else secede from Croatia.Only the violent collapse of the break-away Krajina region in 1995 silenced the competing claim to statehood.Bosnia-Herzegovina, finally, ended its civil war with Croat and Serbian groups in a federal arrangement that remains fragile until today and renders the effectiveness of Bosnia-Herzegovina highly doubtful. 94Likewise, the case of South Sudan, which slid into a civil war shortly after gaining recognition for its independence in 2011, casts some doubt on the assumed constitutive effect of unanimous recognition. 95ore generally, a thematic purview largely held outside the scope of international law is the material ineffectiveness of entities that have been unanimously recognized as states -for instance, Somalia, which is the prime case of 'state failure'.Although the debate on so-called 'state failure' or 'fragility' has been popularized in the discipline of International Relations, 96 some international law scholars have also argued that the turn from effectiveness to legality has endangered international law because it treats entities as states irrespective of whether or not they are effective. 97In this view, even if 'effectiveness is [only] a necessary, but no sufficient criterion of statehood', it 'remains indispensable.It must not be substituted by indices of legality or legitimacy, because such an approach would transform the international legal system into a purely virtual one which could not perform its ordering function'. 9893 98 Peters, supra note 37, at 13.As discussed above, 'because effectiveness is a relative concept, it is difficult to call a government arrangement completely non-effective'. 99Revoking the recognition of supposed 'failed states' has thus been largely rejected by international law scholars. 100State practice too illustrates support for the continued existence and territorial integrity of recognized states, as best evidenced by the notorious example of Somalia. 101Correspondingly, Crawford complains about the 'conceptual confusion at … [the] core' of the state failure approach, because it mistakes 'crises of government' and cases in which the 'regime has failed' for the 'extinction of states'. 102This critique reiterates that the status of a state cannot be reduced to the degree of its effectiveness, but it leaves open whether there could be a point at which the degree of territorial control is simply too low to uphold any credible state status at all.After all, international law assumes some effectiveness as a basis for the rights and obligations with which it endows states. 103The unwavering support for recognized states is therefore also based on a blunt presupposition of their statehood.
A second problem already touched upon arises when an entity is recognized by some, but not by other, states, leading to an ambiguous status that defies the promise of recognition to provide legal certainty.As aforementioned, this is one of the lasting theoretical problems of the constitutive doctrine that led to Lauterpacht's formula of a duty to recognize and his call for collective recognition.For instance, Russia and some European states as well as the USA have accused each other of purely political considerations in their respective (non-)recognitions of Kosovo, on the one hand, and of Abkhazia and South Ossetia, on the other. 104Kosovo has been recognized by more than 100 states but remains unrecognized by many others, including Serbia, Russia, China and five European Union (EU) member states. 105The International Court of Justice (ICJ) reasoned in 2010 that the declaration of independence did not violate international law but left open whether Kosovo had actually become a state and, if so, how -that is, by effectiveness, legal entitlement, or international recognitions. 106Kosovo remains contested in terms of all its state dimensions, thus illustrating the shaky ground of any combination as well.The same holds true for Abkhazia and, if less effective, South Ossetia, which were recognized by Russia, Venezuela, Nicaragua and some island states, but remain unrecognized by all others.The revealing question for the recognition approach to state creation is thus 'how many and whose recognitions are necessary?' 107While it appears that due to its recognition by many other states Kosovo is considered more of a state than Abkhazia and South Ossetia, the sizable number of recognitions of Western Sahara and Palestine did not have a similar effect on their assessment and treatment as states, as discussed further below. 108n conclusion, even if the resort to recognition follows the pragmatic attempt to provide certainty on membership in the international state system, it is limited in this regulatory function, as evidenced from recognized fragile states to unrecognized de facto states.Here, the main issue is not that (non-)recognizing states determine new states politically but, rather, that we do not know whether they actually determine anything else than a formal and perhaps entirely virtual status not tied to any 'effective' local practices of statehood.For instance, while some cases, such as Slovenia and Croatia, show how recognitions participate in relatively effective state creations, other cases, from Bosnia to South Sudan, illustrate that they not always do so, bringing the 'extra-legality' of the state back to the fore.Behind the question whether the state is ultimately determined by its legal status or its empirical materiality remains the deeper ontological assumption that any given entity either is a state or it is not.This entrenched ontology is indeterminate at the core, founded neither on recognized status nor on effective order.Yet because the great debate failed to resolve the dilemma head on by reducing it to one or the other decisive dimensions or by simply combining both in a shaky compromise, it was increasingly neglected in newer trends towards the proper legal regulation of state creations.

Invisibilizing State Ontology: The Legal Regulation of State Creation
The classical positions and even some compromises appear ultimately old-fashioned because they do not take into account the evolution of the proper legal regulation of state creation in modern international law. 109In Crawford's authoritative statement, '[n]either theory of recognition satisfactorily explains modern practice'. 110Beyond the entrenched debate over the priority of supposedly evident facts and authoritative recognitions, modern discourse holds that 'the creation of States might be regulated by rules predicated on other fundamental principles' than effectiveness, and 'even if effectiveness is the dominant principle, it must nonetheless be a legal principle' to be legally assessed. 111wo logics or principles of legal regulation have been particularly relevant -that is, the right to self-determination of peoples and the illegality and non-recognition of state creations that involve the violation of pre-emptory norms such as self-determination and the general prohibition of the use of force.Ex injuria jus non oritur.But whether the creation of states as entities ontologically located at the same time within 108 Sterio, 'On the Right to External Self-Determination: "Selfistans", Secession, and the Great Powers' Rule', 19(1) Minnesota Journal of International Law (2010) 137. 109Shaw, supra note 9, at 177; Crawford, supra note 7, at 5, 97-99, 105. 110Crawford, supra note 7, at 5. 111 Ibid.
and outside of international law can be captured by this shift from identification to regulation is doubtful at best.In fact, I argue that it can neither determine incontestably which entities should be considered states nor whether entitlement does lead to statehood and prohibition prevents it.The regulatory approach invisibilizes the problem of direct identification; it concedes the extra-legal reality of state creation in order to bracket it and focus entirely on the legal side to the problem and, thus, international law's internal discourse on (il)legality.Yet the harmony achieved is an illusion.Not only are questions of the (il)legality of state creations themselves inherently difficult to settle, but it also becomes unclear whether any entity accepted or rejected as a state within the legal discourse is actually a state in that extra-legal reality that has been excluded at the beginning.Ultimately, the ontological question of whether a state has come into existence or not cannot be displaced.

A Between Bangladesh and Northern Cyprus: The Legal Indeterminacy of Rights to Self-Determination and of the (Il)legal Use of Force
That 'self-determination is, at the most basic level, a principle concerned with the right to be a State' 112 has changed the logic of state creation by putting legal entitlement up front actual state emergence. 113The decolonization of former mandate and trust territories as well as other non-self-governing peoples following World War II led to a vast expansion of the number of states formally recognized under international law. 114In light of the widespread consensus in favour of the legality of state creation by decolonization and the (eventual) willingness of most existing states to recognize most dependent entities as new states, 115 a more controversial question has been how the right to self-determination might be invoked in favour of state creations outside the context of decolonization. 116 first problem is the identification of units of self-determination as peoples separate from the populations represented by the recognized states to which they formally belong. 117In fact, determining peoples as authentic entities entitled to self-determination outside the decolonization context turns out to reproduce the original difficulties with determining states as legal subjects. 118Since the population of each state constitutes a people in its own right, any emerging claim to external selfdetermination is bound to clash with the existing state's claim to represent the people on its territory in its entirety. 119The so-called 'safeguard clause' in the Declaration of Friendly Relations reaffirmed, despite all avenues outlined for self-determination, the territorial integrity and right to self-determination of the peoples embodied by existing states, thus providing an effective legal argument against unilateral secession. 120econd, self-determination can also be exercised internally, and peoplehood outside the context of (de-)colonization is not understood to provide for an explicit right to independent statehood exercised against the will of the formal parent state. 121The argument advanced in the case of Kosovo's bid for independence -namely, that a people or even a minority can obtain the right to external self-determination against the parent state's will as a means for 'remedial secession' -has gained some acceptance in the international law literature but still remains widely contested. 122Also, it raises the question of how the threshold of oppression for any remedial secession can be objecti vely delineated and applied in individual cases, a question that further complicates the determination of a right to self-determination.
Conversely, the invalidation of state creations that involved the violation of jus cogens rules is sometimes also hard to judge, especially in cases outside the decolonization process. 123Violations of two pre-emptory norms are held to be particularly important -that is, the right to self-determination and the prohibition of the use of force. 124A rather clear case of the violation of the principle of self-determination was the (attempted) state creation by the 'illegal racist minority regime' of (Southern) Rhodesia, 125 which was not accepted as an independent state upon its declaration of independence in 1965 and which crumbled in 1979. 126Yet, in more contested cases, such as Bangladesh, the TRNC and Kosovo, the problem for determining a violation is closely linked to the aforementioned difficulty of establishing a right in the first place.Similarly, the legal invalidity of an entity's status as a state as a result of the illegal use of force, which played a role in Manchuria after the Japanese invasion and in northern Cyprus after the Turkish invasion, depends on determining the use of force as illegal intervention rather than as justified assistance to a self-determination unit and emerging new state. 127he difficulty in assessing the entitlement to and, conversely, (il)legality of any state creation is illustrated by the divergent trajectories and evaluations of the (attempted) secessions of Bangladesh and the TRNC.Both the Pakistani population in the east of the country and the Turkish population in northern Cyprus -assisted in their secessions by India and Turkey respectively -had not been considered selfdetermination units beforehand. 128Yet the supposed (il)legality of their state creations played out differently.Although, at first, both interventions were considered to be illegal, in the case of the secessionists in Eastern Pakistan, India's military assistance was justified by the argument that, under the specific circumstances of the situation, Bangladesh had obtained a quasi-right to self-determination. 129Yet the many factors invoked in favour of such an extraordinary right, including repression against the ethnically distinct population and its express will to separate, also applied in the case of northern Cyprus. 130At the very least, 'the distinctions are not so plain as to speak for themselves'. 131Yet, despite these similarities, the use of force by India was ultimately considered legitimate assistance to Bangladeshi selfdetermination, whereas the use of force by Turkey in Cyprus was mainly rejected as an illegal intervention and occupation. 132he attempt to maintain a difference between the two cases in terms of legality has also been undermined by explanations based on effectiveness and recognition.It was pointed out, for instance, that '[d]espite the presence of Indian troops on its territory, its doubtful stability, the refusal of Pakistan's recognition until 1974 and the probable illegality of Indian intervention, Bangladesh was rather rapidly recognized as a State'. 133Whether recognitions were therefore decisive is unclear too, however.While some argue that Bangladesh 'was recognized by many States at a time when there was no effective government', 134 others maintain that even before Pakistan's acquiescence, 'Bangladesh existed de facto as a viable State', claiming that, despite controversial discussions over the legality of India's intervention, what motivated its recognition 'was simply the principle of effectiveness'. 135By contrast, Turkey's intervention in Cyprus and the creation of the 'putative state' of the TRNC are not only widely considered illegal but have also not been recognized by any other state than Turkey, measures of actual effectiveness notwithstanding. 136his brief comparison illustrates the long-standing insight of critical legal studies that any legal interpretation is vulnerable to counter-arguments, 137 and it suggests that the dominant views on the state character of Bangladesh and the TRNC are neither based on a compelling argument for legal entitlement or its absence nor on recognitions and effectiveness but, rather, on dominant presuppositions of the presence or absence of statehood consolidated over time among third states, international organizations and scholars.Perhaps assessments of Kosovo's and Abkhazia's statehood too will consolidate in one way or another over time, but this too depends on the ontological politics of assuming state presences and absences.

B From Rhodesia to Western Sahara: On the Potential Virtuality of Legal Status
Apart from the question of the legal indeterminacy of entitlements to, and invalidations of, state creations, the turn to legality and, thus, the shift from state identification to the regulation of state creations might lead to what Peters aptly calls the 'virtuality' of statehood in international law -that is, the detachment of the legal discourse from the social reality it attempts to grasp and transform. 138Some cases illustrate the limited transformative role of even dominant legal assessments because entitlements do not simply make states and prohibitions do not simply dissolve them either.
The case of Rhodesia is a prime example usually invoked to show the important effect of additional and genuinely legal state criteria; in this case, the pre-emptory norm of self-determination and the prohibition to prevent self-determination by force. 139Of course, ultimately, the white minority regime was wrestled down by international non-recognition, economic sanctions and indigenous guerrilla movements. 140But those measures took 14 years to end the more or less effective material existence of this unrecognized entity, and, in retrospect, its demise might look all too automatic.When assessing Rhodesia's claim to statehood in 1974, nine years into its existence, Christian Okeke came to the conclusion that 'Rhodesia ranks among the entities which are endowed with statehood under international law', even if 'unrecognized', and that 'international law should apply to its activities'. 141In addition, with the partial exception of South Africa, Rhodesia could not count on lasting allies, as 136 Crawford, supra note 7, at 144, 140-147. 137Cf.Koskenniemi, supra note 18. 138 Peters, supra note 37. Virtuality means simply illusive here.By contrast, as an essentially irreducible presupposition of presence, state ontology is always 'virtual' in some sense, if also partly substantiated in legal and extra-legal practices. 139Crawford, supra note 7, at 128-131. 140 Minter and Schmidt, 'When Sanctions Worked: The Case of Rhodesia Reexamined', 87 African Affairs (1988) 207, at 233-234. 141Okeke, supra note 64, at 105.Of course, as Okeke emphasizes, this 'does not necessarily preclude any … collective action against such an entity.' can, for instance, Abkhazia and other pro-Russian 'states', which will therefore be much more difficult to wish away.But whether or not Rhodesia was a state between 1965 and 1979 is thus a question that cannot be objectively settled by resort to either effectiveness or legality.The factual power of the normative, which the case is supposed to show, is limited, as further illustrated by the persistence of, for instance, 'the illegal' TRNC.Conversely, there is still a number of clearly identified non-self-governing territories with a confirmed right to external self-determination that are not yet widely regarded by other states, courts or international law experts as having achieved statehood. 142n illustrative example in this respect is Western Sahara, the right to self-determination of which was explicitly reiterated by the ICJ, which also rejected the counterclaim to the territory by Morocco (and Mauritania). 143Yet the projected referendum on independence never took place, and Morocco assumed effective control of most of the territory -and only a few years ago concluded an international treaty with the EU, in which it traded fishing rights in the territorial waters of Western Sahara as if exercising legitimate sovereignty over them. 144The exile government of the self-declared Sahrawi Arab Democratic Republic is virtually powerless in most of Western Sahara.Thus, even legal entitlement and widespread recognition together do not lead to the widespread acknowledgement, or actual making, of a state. 145his should not lead to the conclusion that in the end '[t]he reality of effective authority tends to prevail, notwithstanding the maxim ex injuria jus non oritur'. 146ndeed, '[i]t may be that international law's main contribution in such cases … is to keep the issue on the agenda until the circumstances change and a settlement becomes possible'. 147Palestine's claim to statehood, for instance, has regained some momentum in the international arena, although Palestine too is 'only' widely recognized and legally entitled, but without effective control in most areas of its territory.Its accession to the ICC is a case in point.
In conclusion, it is true that 'the question is precisely whether the term "State" should be regarded as for all purposes equivalent to certain situations of power' and 'that international law risks being ineffective precisely when it does not challenge effective but unlawful situations'. 148The advancement of a genuinely legal approach to state creations has provided international law with an autonomy from the ultimately extra-legal claims of empirical facts and (non)-recognitions.However, this strictly legal approach already juxtaposes the identification of states with the regulation of their creation and affairs, which assumes that we know states, or at least entities materially or per recognition qualifying as states, other than by their lawful emergence.In other words, the original question whether or not particular entities are states is first and foremost not a question of whether they should or should not be states.Ignoring the extra-legal comes with the risk of simply taking legal representations for reality and undermining international law's ability to regulate the relations between the forces that be, including the actual capacities of (would-be) states to wage war, police territory, organize public services and assume the rights and fulfil the obligations of states under international law -capacities that apparently partly predate and surpass the constitutive power of legal designations and recognitions.Indeed, it is still widely held that, in principle, international law is silent on secessions that are neither legally endorsed nor prohibited; 149 and yet it has to identify new states emerging from secessions.Criteria of (il)legality, then, have changed the game of presupposing states into or out of existence, but the ontological question of presence or absence still silently structures legal arguments and decisions.

Conclusion: State Creation in the Discursive Web of State Ontology
The question of state creation strains international law's stance between its normative aspiration 'to make "legality" prevail over sheer force' and its 'realistic' role that 'takes account of existing power relationships'. 150Even more fundamentally, it puts into question what exactly the reality of the state is if neither recognitions nor measures of effectiveness can ultimately determine it.As illustrated in this article, the three reference points of statehood in international law -that is, effectiveness, recognition and (il)legality -are indeterminate of individual state identities.More or less effective territorial control by a self-declared government can put persistent non-recognition into doubt, but it cannot be simply acknowledged as statehood either because effectiveness waxes and wanes or assessments of the entity's stability may legitimately differ.Yet recognitions, while promising determinate decisions by competent authorities, in turn fail to reflect or simply constitute sufficiently effective entities and cannot dispel doubts about the mere 'virtuality' of recognized states.Finally, considerations of legal rights to, and prohibitions of, particular state creations constitute international law's autonomy over recognitions and empirical claims, but this shift confuses identification with regulation and thus conceals the ontological question that remains decisive in identifying states beyond the question of (il)legality.I have focused on the logic of legal arguments to show that behind their indeterminacy it is the ontological structure of state presence and absence that is really determinate in the sense that it necessitates clear answers and blunt commitments.The commitments are ontological in the sense of signifying 'being' irreducible to any particular dimension of the state.They are further silent because the fundamental assumption that states are either present or absent remains unarticulated; indeed, international legal arguments assume that effectiveness, recognition or legality determine state identities, although they do not.As this ontological dichotomy constructs an image of certain identity required by the exclusive structure of the state system, it provides a platform for both the contestation of the status quo membership by marginalized entities at the fringes and the discursive control over such contestations, which are articulated and can thus also be rejected in the language of international law.At the same time, international law cannot objectively identify states and, yet, has to identify states under objectivist pretensions.It thereby participates both in the emergence and demise of particular states and in the re-production of the international state-based order as a whole.This illustrates both the possibilities and the limits of re-imaging and re-constructing states in international law.No matter how technically fain-grained, any legal analysis of state creation is ultimately enabled by, and inevitably reproduces, this ambiguous ontological horizon.
What, then, 'is' the state beyond the deceptive image of a full presence?It appears that the sense of its reality emerges at the intersections of different dimensions and approaches to it.Although the three main principles of state identification in international law are inconclusive, their use in fostering potential cases of state creation shows how they provide for a discursive web of more or less plausible claims in which arguments in favour of particular state identities can be nested.We may think of the web as working only by the combined strength of the strings and knots of which it is composed, while anything the web carries might slip again through the holes, depending on the density of the web at any specific point.Effective, but unrecognized, Somaliland, partly recognized but, largely rejected, Abkhazia and legally endorsed and widely recognized, but clearly ineffective, Western Sahara all sit on few strings and knots, if indeed on different ones.They struggle to hold on to the web and climb their way up to where entities reside that are routinely and unquestionably considered states, such as the PRC or Israel, notwithstanding their contestable origins, nonrecognition by several UN members and the occasional turmoil and instability in, for instance, Xinxiang and Jerusalem.Ultimately, the difference lies in the routine reiteration of blunt presuppositions of the state character of these latter entities in everyday legal, diplomatic and other practices.
Legal assessments of state creation thus participate in what they claim to assess, without ever gaining the entirely determining power with which a naive constitutive position might credit recognition.When fostering a case for or against any particular entity's statehood, practitioners have no choice but to resort at least implicitly to theories of state identification and to accept the basic question posed by the ontological dichotomy of presence and absence.In this sense, legal assessments about statehood are like all legal assessments: a practice of articulating contingent and always vulnerable positions in a professional, but open-ended, language. 151In such a reflective perspective, the question is not whether Palestine 'is' truly a state or not but how, where and when it is assumed to be one and which further effects any presuppositions of its statehood entail in concrete legal or political practices. 152This is not a call for some non-legal viewpoint that would provide a ground for 'right' decisions.Rather, the point is to become conscious of the need to make, and the implications of making, a concrete commitment and thus enlarging the scope of concerns, although, in the end, decisions cannot be determined either by law or by some notion of extralegal reality outside of it.Against the horizon of a continuously reproduced state system, contested individual cases call for prudent arguments and decisions on concrete claims to state creation that will never live up to some ideal standard, but that can be made in reflection of the overall predicament and potential or likely consequences: 'In an ontological politics we might hope ... to interfere, to make some realities realer, others less so.The good of making a difference will live alongside -and sometimes displace -that of enacting truth.' 153
1, at 1. On the centrality and ambiguity of the state concept in modern political thought, see J. Bartelson, The Critique of the State (2001).11French,supranote10,at1.12For the production of the legal and the extra-legal in international law, see F. Johns, Non-Legalities in International Law: UnrulyLaw (2013).By tracing the metaphysical dichotomy of presence and absence, my approach is loosely inspired by J. Derrida, De la Grammatologie (1967) but it also draws on the understanding of 'ontological politics' articulated by Mol, 'Ontological Politics: A Word and Some Questions', 47(1) Sociological Review (1999) 74; J. Law, After Method: Mess in Social Science Research (2004).However, while this article examines the role of state ontology within legal discourse, including the non-legal realities this discourse invokes and in turn co-produces, it does not engage with state ontologies outside the legal discourse.
149Abi-Saab, 'Conclusion', inKohen, Secession, supra note 55, 470, at 473.For an account of divergent processes of secession in practice, see also A. Pavković and P. Radan, Creating New States: Theory and Practice of Secession (2007).This does not mean that international law is neutral and state practice would not indicate a strong inclination to preserve existing states and territories.See Oeter, 'Recognition and Non-Recognition with Regard to Secession', in C. Walter, A. von Ungern-Sternberg and K. Abushov (eds), Self-Determination and Secession in International Law (2014) 45, at 52, 54.But it does mean that the ultimate answer to the question whether or not a new entity has emerged cannot be based purely on legal norms. 150Cassese, supra note 26, at 12-13.See also Shaw, supra note 9, at 42-44.