The conventional identification of states: legal recognition in IR

Despite IR's traditional (Neo-)Weberian commitments, the discipline largely depends on legal recognition to identify states. As Strang (Reference Strang1991, 150) points out, although many scholars invoke Weber's (Reference Weber2004, 33) definition of the state and his emphasis on the monopoly over the means of force, in practice ‘[r]esearchers do not ascertain which organizations continuously patrol a given territory’ but rather ‘use lists of members of the United Nations’ to identify states. Even Tilly, an important inspiration for studies of unrecognized state making (Kingston and Spears Reference Kingston and Spears2004; Niemann Reference Niemann2007; Hagmann and Hoehne Reference Hagmann and Hoehne2009), ultimately falls back on legal recognition as the marker of individual states (Tilly Reference Tilly1992, 2). Thus, traditionally, state formation is expected to unfold within recognized or otherwise already presupposed states (Tilly Reference Tilly1992, 191–225; Clapham Reference Clapham1998; Sørensen Reference Sørensen2001).

The literature on quasi- (Jackson Reference Jackson1990), failed (Rotberg Reference Rotberg2004), or collapsed (Zartman Reference Zartman1995) states builds on the same dichotomy of legal status and empirical order to investigate the fragility or lack of ‘internal’ statehood within ‘externally’ delineated states in which such shortcomings can be empirically observed. These ‘quasi-states’ (Jackson Reference Jackson1990) are said to exist only on the international legal plane, while lacking the substance of domestic order. Yet, because the spaces in which individual state collapse and state failure can be empirically explored are still delineated as units by virtue of their legally recognized status, the assessment of state failure does not suggest that these recognized entities are no longer states. Indeed, for Crawford (Reference Crawford2006, 722), it is a confusion to think states have ceased to exist where governments have failed, including in notorious cases such as Somalia. That there are different ‘types’ (Sørensen Reference Sørensen2001) or ‘degrees of statehood’ (Clapham Reference Clapham1998) thus does not undermine, but rather reinforces, the reliance on legal recognition for the identification of individual states as the concrete ‘starting point of international relations’ (Bull Reference Bull1977, 8), whatever their international ranking, degree of political autonomy, or domestic capacity to govern. The assumption of already given states is thus both the condition of possibility of IR and its inherent limitation (Walker Reference Walker1993; Bartelson Reference Bartelson2001).

Arguably, a burgeoning literature taking a more explicit approach to recognition has recently sprung up within IR (Fabry Reference Fabry2010; Lindemann Reference Lindemann2013; Coggins Reference Coggins2014; Lindemann and Ringmar Reference Lindemann and Ringmar2014; Daase et al. Reference Daase, Geis, Fehl and Kolliarakis2015). However, much of this literature is concerned with ‘political recognition’ between already presupposed states, rather than with the identification of these entities as states in the first place (Bartelson Reference Bartelson2013, 113). Where ‘legal recognition’ is the subject (Fabry Reference Fabry2010; Bartelson Reference Bartelson2013; Coggins Reference Coggins2014), the discussion is often informed by a version of the classical debate in international law about whether recognition is ‘constitutive’ or ‘declarative’ of statehood (Grant Reference Grant1999; Crawford Reference Crawford2006), with a distinct interest in normative and political considerations informing practices of recognition, rather than state making as such (Fabry Reference Fabry2010; Coggins Reference Coggins2014). The focus on recognition alone, moreover, either simply presupposes states as already there, or else credits recognition with the power to generate them. Indeed, the claim that entities require recognition to be regarded as proper states (Fabry Reference Fabry2013, 165; Coggins Reference Coggins2014, 8) is, if not a mere tautology, an equation of statehood with recognition by definitional fiat, rather than an explanation of state creation, as it shuns the prior question of what states actually are and how they can be identified as such.

If states are not reducible to legal recognition, one cannot be content with an analysis of how candidate entities were or were not recognized by other states, or, in classical English-School parlance, admitted to an expanding international society (Bull and Watson Reference Bull and Watson1984). Instead, the more fundamental question is how we can know the candidate entities in question to be ‘states’ or ‘almost-states’ in the first place. The answer requires what Bartelson (Reference Bartelson2013, 108) calls ‘epistemic recognition.’ Yet, as he points out (Reference Bartelson2013, 108–109):

Thus, although IR is somewhat wary about its central assumption of a state system already delineated by legal recognition, for the most part, it does not divert from this key assumption. Yet, if states are neither simply given nor reducible to recognized status, then they have to be identified prior to and independently of recognition, at least in a rudimentary form. As I shall argue below, this is what the de facto state literature tries, but ultimately fails, to do.

Beyond the fringe? Conceptualizing unrecognized state making

The notion of de facto states – which I use as a short-hand also for ‘contested’ (Geldenhuys Reference Geldenhuys2009) or ‘unrecognized’ (Caspersen Reference Caspersen2012) states – directly challenges the conventional presupposition of states in IR and rejects legal recognition as the decisive criterion for determining states. Instead, it approaches statehood from the domestic side, drawing largely on historical sociology and a (Neo-)Weberian understanding of state formation (e.g. Tilly Reference Tilly1992), as well as the notion of effectiveness in international law. What it means to bring ‘states’ at the margins of international society to the fore, however, depends on how wide the conceptual net of supposed unrecognized state making is cast.

Many political scientists invoke ‘separatist states’ (Lynch Reference Lynch2004), ‘phantom states’ (Byman and King Reference Byman and King2011), ‘mini-states’ (Davies Reference Davis2009), or ‘states-within-states’ (Kingston and Spears Reference Kingston and Spears2004) to capture a host of different ‘state-like features’ (Kingston Reference Kingston, Kingston and Spears2004, 6) of potent rebel groups, warlord fiefdoms, or other political communities. This is often done for heuristic purposes to reflect their accomplishments in developing particular levels of territorial control, bureaucratic organization, or public services that are ‘state-like’ (Kingston Reference Kingston and Spears2004, 2). Yet, for all the summoned power of the analogy to state formation, these authors often shrink back from the full-blown claim that these entities are or have turned into ‘actual’ states. For instance, Spears's ‘key elements of a Weberian definition’ are merely a ‘range of state-like structures’ under a definition that ‘has intentionally been left imprecise’ (Spears Reference Spears, Kingston and Spears2004, 16–17). Thus, ‘states-within-states are often incomplete insofar as they do not have all the substantive attributes of statehood’ (Spears Reference Spears, Kingston and Spears2004, 30) and merely show a ‘potential’ to ultimately become fully developed states (Kingston Reference Kingston, Kingston and Spears2004, 6). In practice, the permissive nature of the concept makes it difficult to distinguish between ‘states-within-states’ and non-state actors. As pointed out by Reno (Reference Reno2009, 361), some supposed ‘states-within-states’ would have been ‘liberated zone[s]' in ‘1970s lingo.’ Others simply speak of ‘rebel rulers’ (Mampilly Reference Mampilly2011). As a result, this wide notion of unrecognized state making has been criticized as overstretched (Geldenhuys Reference Geldenhuys2009, 27; Caspersen Reference Caspersen2011, 3).

By contrast, the research program on ‘de facto’ (Pegg Reference Pegg1998), ‘unrecognized’ (Caspersen and Stansfield Reference Caspersen, Stansfield, Caspersen and Stansfield2011; Caspersen Reference Caspersen2012), and ‘contested states’ (Geldenhuys Reference Geldenhuys2009) employs a more restrictive notion of de facto states as ‘entities which feature long-term, effective, and popularly-supported organized political leaderships that provide governmental services to a given population in a defined territorial area,’ but which lack ‘widespread juridical recognition’ (Pegg Reference Pegg1998, 4). Rather than a vague allusion to statehood, the claim here is essentially literal: these entities are states and therefore ‘deserve to be called “states”’ (Geldenhuys Reference Geldenhuys2009, 26). Faced with ‘collective non-recognition’ (Geldenhuys Reference Geldenhuys2009, 22), they ‘find their very right of statehood being challenged by their parent states and the broader international community’ (Geldenhuys Reference Geldenhuys2009, 3). This view is often backed by recourse to the widely cited Montevideo Convention on the Rights and Duties of States (1933), which – echoing the Weberian notion of the state – defines states by a permanent population, defined territory, government, and capacity to enter into relations with other states (Pegg Reference Pegg1998, 27; Lynch Reference Lynch2004, 15; Geldenhuys Reference Geldenhuys2009, 8; Caspersen Reference Caspersen2012, 17). While I shall illustrate that these criteria are less easily operationalized than some proponents of de facto states would suggest, the classical definition of effective statehood without recognition gives the literature its conceptual basis. Provided that these criteria are fulfilled, ‘[t]he defining characteristic of unrecognized states, the factor that determines their position in the international system and predominates in internal debates, is their lack of recognition’ (Caspersen Reference Caspersen2012, 16).

Contesting legal recognition

The literal claim to statehood differs from mere allusions to state-like features in that the statement of fact expresses a normative demand to acknowledge that the entities in question really are states. Misrecognizing them as non-state entities makes a difference both analytically and practically. This position is reflected in diverse issue areas into which the de facto research program has ventured, including the global exclusion of populations living in unrecognized territories (Pegg Reference Pegg1998; Geldenhuys Reference Geldenhuys2009; Caspersen Reference Caspersen2012), the politics of recognition and secret diplomacy (Ker Lindsay Reference Ker Lindsay2012; Pegg and Berg Reference Pegg and Berg2016), and the peculiar domestic dynamics of nationalism, democracy, and security in unrecognized states (Pegg Reference Pegg1998; Kolstø and Blakkisrod Reference Kolstø and Blakkisrud2008; Caspersen Reference Caspersen2011; Berg and Mölder Reference Berg and Mölder2012).

Two implications of unrecognized status are particularly noteworthy. First, the very claim of de facto state making suggests an unacknowledged creation of effective states, including in areas often deemed unsuitable for state formation. This challenges common views on state failure and fragile quasi-states from Sub-Saharan Africa to the rim of the post-Soviet space (King Reference King2001; Kingston and Spears Reference Kingston and Spears2004; Niemann Reference Niemann2007; Hagmann and Hoehne Reference Hagmann and Hoehne2009). As assessing state formation and state fragility depends on a prior identification of the units within which these processes are to be observed, the de facto state claim also challenges these units. Indeed, in this view, the failure of affected recognized states to maintain control in their entire territory might be precisely due to the rise of a de facto state in part of that territory.

Second, unrecognized states are legally excluded from the international system and therefore particularly vulnerable to renewed armed conflict. They have not only often ‘effectively defeated the armies of recognized governments in open warfare’ (King Reference King2001, 526), and thus ‘frequently emerge from wars,’ but they are also commonly ‘sustained by the threat of further fighting’ (Byman and King Reference Byman and King2011). Indeed, secessionist wars feature among the most devastating armed conflicts since World War II and many remain volatile long after the fighting ends (Coggins Reference Coggins2014, 3). Crucially, the status of separatist entities in conflict situations is itself an object of contestation, with serious consequences. The two types of conflict – that is, intra- and inter-state war – become virtually indistinguishable and can in fact only be distinguished at the cost of making a political choice. Particularly telling illustrations of the implications of denied status are the eventual destructions of the Chechen Republic in Russia, the Republic Krajina in Croatia, and the Tamil Tigers of Tamil Eelam in Sri Lanka (Caspersen Reference Caspersen2012, 11). In all these cases, civil wars between the separatists and the official government had ended in peace agreements or permanent ceasefires (Pegg Reference Pegg1998, 66–84; Stokke Reference Stokke2006; Caspersen Reference Caspersen2012, 11). Yet, hostilities resumed later in full force, ending with the defeat of the ‘rebels’ and the complete dismantling or violent re-integration of their polities as a matter of their respective parent state's ‘internal’ affairs. In this sense, ‘state death’ (Fazal Reference Fazal2004) might still occur, but since it does not concern recognized states, it remains undetected.

In response, some international law scholars propose to treat all conflicts involving de facto states or ‘states in statu nascendi’ as ‘quasi-international armed conflicts’ (Wills Reference Wills2012) to extend the legal protection against aggression. In this view, the effective statehood of the entities involved, and not their legally (un)recognized status, should determine whether a conflict is intra- or inter-national (Wills Reference Wills2012). Others disagree with this as a general rule, but still propose a functional assumption of interstate conflict in specific cases. To take the example of the controversial relationship between mainland China and Taiwan (Mengin Reference Mengin2008), in which the People's Republic maintains that any potential armed conflict with Taiwan ‘falls within the domestic affairs of China’ (Chan Reference Chan2009, 492), Crawford argues that although Taiwan cannot formally be regarded as a state, we should assume ‘a cross-Strait boundary for the purposes of the use of force’ (Crawford Reference Crawford2006, 221). This would put any attack by China under the restrictions of the general prohibition of the use of force in interstate relations.

Both implications of supposed de facto state making rest on the assumption that we are in fact dealing with states, albeit in an unrecognized form. Indeed, while the major studies on de facto states (Pegg Reference Pegg1998; Geldenhuys Reference Geldenhuys2009; Caspersen Reference Caspersen2012) paper over the precise definition of the outcast entities, they all converge on the claim that the existence of states is a matter of fact. Although generally treated as a mere starting point for diverse empirical investigations in the de facto state literature, this central claim presents itself a potentially audacious challenge to the convenient assumption of a readily demarcated state system composed of recognized states. It is to its supposed substantiation that I turn in the next section.

Formal criteria applied: the flawed attempt to identify states as objective facts

The de facto state concept draws on the established dichotomy of effective order and legal status in IR by maintaining that the ‘de facto state’ is ‘the flip side of the quasi state coin’ (Pegg Reference Pegg1998, 5). While quasi-states are supposedly ‘not states in the strict sense, but only by courtesy’ (Bull and Watson 1984, 430), de facto states are considered actual states because of their supposed effectiveness (Pegg Reference Pegg1998; Geldenhuys Reference Geldenhuys2009). However, any attempt to grasp the substance of the state by virtue of its ‘effectiveness’ calls for the operationalization of criteria for empirical analysis. As debates within the literature show, determining the specific criteria and employing them for empirical analysis only reveals the empirical indeterminacy of state status.

As mentioned in the previous section, scholars drawing on accounts of state formation in historical sociology regard state-like entities beyond recognition as ‘Tilly-like states in the making’ (Kingston Reference Kingston, Kingston and Spears2004, 2). Invoking ‘a Weberian definition of statehood,’ Spears (Reference Spears, Kingston and Spears2004, 17–18) argues, one could make the case that ‘the processes that Tilly describes with respect to Europe were very much at work in creating and strengthening […] states-within-states.’ However, coercion and extraction in populated territories do not necessarily amount to the creation of states (Kingston Reference Kingston, Kingston and Spears2004, 6; Spears Reference Spears, Kingston and Spears2004, 17) and state formation itself is a continuous process that does not stop at a particular point. In fact, Tilly's (Reference Tilly1992) narrative of state formation over a millennium is not primarily concerned with – and is arguably even opposed to – defining a particular benchmark for fully developed statehood that could be used for assessing individual cases or measuring an entity's achieved ‘statehood.’

Hence, those who advance a literal claim of de facto state making generally seek to define a threshold that would ‘exclude more ephemeral entities’ (Caspersen Reference Caspersen2012, 7), that is, ‘insurgent states, black spots, and states-within-states’ (Caspersen Reference Caspersen2012, 8; see also Geldenhuys Reference Geldenhuys2009, 26–27). An empirical operationalization which sets de facto states apart from less effective entities seems crucial for making the point of actual state making beyond recognition. For this purpose, Pegg (Reference Pegg1998, 27; emphasis added) argues that a ‘de facto state's capabilities’ are sufficient to indicate statehood if ‘it has an organized leadership; it has reached such a level that it is able to provide a degree of governmental services; it is able to control a given territorial area for extended periods of time.’ Caspersen (Reference Caspersen2012, 11) too acknowledges that, in reality, statehood is a question of degree, so that, for instance, ‘de facto independence,’ which she regards as a criterion of statehood, does not necessarily entail full territorial control, but should ‘cover at least two-thirds of the territory to which it lays claim (…) including its main city and key regions.’ Furthermore, since a formal declaration of independence can be strategically disadvantageous, she argues, it is sufficient that the entity has ‘demonstrated clear aspirations for independence’ (Caspersen Reference Caspersen2012, 11). Additionally, Pegg (Reference Pegg1998, 32) holds that states should not be ‘here today and gone tomorrow,’ although he is unsure whether ‘one month’ or rather ‘one year’ would be an appropriate threshold. Others set it to ‘three years’ (Geldenhuys Reference Geldenhuys2009, 4) or ‘at least two years’ (Caspersen and Stansfield Reference Caspersen, Stansfield, Caspersen and Stansfield2011, 3; Caspersen Reference Caspersen2012, 11). Pegg (Reference Pegg1998, 32; emphasis added) ultimately concedes that setting a specific temporal threshold remains ‘somewhat arbitrary.’ This is no small concession. Rather, as I illustrate below, the entire language of sufficient ‘degrees,’ ‘extents,’ and ‘levels’ suggests a scale on which any specific threshold cannot but appear as arbitrary.

Unsurprisingly, based on their respective specific criteria, Pegg (Reference Pegg1998), Geldenhuys (Reference Geldenhuys2009), and Caspersen (Reference Caspersen2012) variously include some entities as supposed de facto states while excluding others. For instance, Caspersen's (Reference Caspersen2012) criterion of control over ‘two-thirds of the territory’ keeps Somaliland in, but leaves Western Sahara and Palestine out, while Geldenhuys (Reference Geldenhuys2009) includes all three cases in his list of ‘contested states.’ Similarly, while Geldenhuys (Reference Geldenhuys2009) counts Taiwan in, for Caspersen (Reference Caspersen2012, 10–12) it is a ‘borderline case’ of an unrecognized state because it is recognized by 23 states and maintains informal relations with many others. The criterion of having ‘declared formal independence or demonstrated clear aspirations for independence’ (Caspersen Reference Caspersen2012, 11) also poses problems. For instance, the Kurdish Autonomous Region in Iraq was included although it had not formally declared independence, while South Ossetia and the Turkish Republic of Northern Cyprus, which did declare independence, were included despite their continued dependence on their respective patrons, Russia and Turkey (Caspersen Reference Caspersen2012, 8–12). As Mengin (Reference Mengin2008) suggests in the case of Taiwan, foreign relations with other countries can also be actively concealed through various legal fictions so as to evade the question of independence and avoid provocations. Regarding the hallmark criterion of effective territorial control, even a poster case of de facto statehood such as Somaliland has been contested. While often presented as ‘maintain[ing] effective control in its territory’ (Geldenhuys Reference Geldenhuys2009, 12; see also Caspersen and Stansfield 2011; Caspersen Reference Caspersen2012), a number of scholars have cast serious doubt on the degree of government control and the stability of clan compromises in Somaliland (Zierau Reference Zierau2003; Kolstø Reference Kolstø2006, 727–279; Renders and Terlinden Reference Renders and Terlinden2010). Considering these ambiguities, Crawford (Reference Crawford2006, 417) concludes that although ‘the notion of the de facto regime has been pressed to its ultimate […] Somaliland is still not a state.’ As such, it appears that whether Somaliland, Taiwan, Northern Cyprus, and other contenders ‘really’ are states or not remains empirically indeterminate.

Expanding or erasing the gray zone? De facto ‘states’ in conceptual limbo

The contested effectiveness of many de facto states and the difficulty of establishing an objective threshold for a sufficient degree of material statehood also poses a problem for distinguishing de facto states from non-state entities. This is important because, conceptually, the de facto state claim depends on a double move. While erasing the difference between de facto states and recognized states ‘above’ them as artificial and merely a product of recognition, rather than effectiveness, it simultaneously establishes a difference vis-à-vis other entities ‘below,’ which lack sufficient effectiveness for statehood. Yet, if de facto states cannot be clearly shown to be states by virtue of their effectiveness, the difference between them and entities without a sufficient degree of political organization and effective territorial control, such as the various rebel rulers and ‘states-within-states,’ also remains fuzzy. As I discuss below, this threatens the de facto state concept by either expanding or erasing it.

Although they ‘are generally weak, poor, and very corrupt,’ unrecognized states are supposedly still not to be confused with ‘anarchical badlands,’ since this situation is ‘not all that different from the countries of which they are formally part’ (Caspersen Reference Caspersen2012, 22). Indeed, many recognized states also lack full control, are minuscule (Geldenhuys Reference Geldenhuys2009, 10; Caspersen Reference Caspersen2012, 5), or have their independence regularly compromised (Geldenhuys Reference Geldenhuys2009, 15–18; Caspersen Reference Caspersen2012, 5). This is, in Krasner's (Reference Krasner1998) terminology, part of the ‘organized hypocrisy’ of international relations. Yet, acknowledging that deficits in actual control are not limited to de facto states only shows that all ‘states’ are effective only to some degree, and sometimes hardly at all. It does not help us draw a line between de facto states and other, less effective entities. In fact, the threshold of effectiveness is so vague that supposed unrecognized states are themselves regarded to ‘move in and out’ of other, non-state categories, including the aforementioned ‘black spots’ and ‘insurgent states’ (Caspersen Reference Caspersen2012, 12). The difficulty of distinguishing alleged de facto states from non-state groups therefore puts their genuine state character further into question (Byman and King Reference Byman and King2011).

In short, de facto states are not determined by a particular level of effectiveness – but neither are recognized states. As a result, recognition itself is sometimes brought back in to find a way out of the conceptual quagmire. Caspersen (Reference Caspersen2012, 150), for instance, links some deficits of de facto states to their lack of recognition and even provides an argument for why recognition, rather than effectiveness per se, makes the difference:

This attempt to square the circle in the face of the empirical indeterminacy of state status leaves de facto states in a conceptual limbo and returns us to the classical choice between simply presupposing states or identifying them via legal recognition. If recognition provides a ‘certain rigidity’ to statehood, while de facto states remain characterized by ‘fluidity’ (Caspersen Reference Caspersen2012, 150), this distinction is either essential for state status, in which case recognition reemerges as the actual marker of proper states, or else the lack of recognition is an accidental condition of de facto states that may hamper their development and endanger their survival but has no bearing on their essential state character.

In the latter case, however, it is not by crossing a particular threshold of effectiveness that de facto states are identified as states. Instead, by studying their ‘domestic’ politics and ‘foreign’ relations, Somaliland, Abkhazia, Taiwan, and others are invoked as states for analytical purposes, despite the difficulties in properly identifying them as such. This allows for these ‘anomalies in a world of sovereign states’ (Caspersen Reference Caspersen2012, 3) to be empirically investigated as unrecognized states, but at the price of simply presupposing their state character, rather than substantiating it. This effectively curbs the challenge posed by the de facto state concept. Rather than turning the tables on legal recognition as the marker of states, the de facto state claim exposes a more radical uncertainty about the ontological status of states between recognition and effectiveness since neither can make a solid case for state identification.

Circling around state identification: the constitutive and declaratory doctrines

Like IR, public international law traditionally operates on the presumption of an international system that is neatly divided into states. In contrast to IR, however, international law has been seriously grappling with challenges of state creation and recognition for more than a century (Kohen Reference Kohen2004; Crawford Reference Crawford2006; French Reference French2013). The two principal approaches took shape during the long 19^th century in response to the decline of earlier doctrines based on dynastic legitimism and imperial sovereignty (Alexandrowicz Reference Alexandrowicz1958; Anghie Reference Anghie1999; Grant Reference Grant1999). The constitutive theory, which holds that ‘through recognition only and exclusively a State becomes an International Person and a subject of International Law’ (Oppenheim Reference Oppenheim1905, 110 cited in Crawford Reference Crawford2006, 15), emerged as an expression of legal positivism and standard-of-civilization discourse to tell fully sovereign states from other entities in the nascent statist international system. The counter-position, maintaining that states are a matter of fact and recognition is only declaratory, has its antecedents in the Latin American bids for independence from the Spanish Empire, where it was eventually articulated in the principle of uti possidetis juris (Fabry Reference Fabry2010, 50). Thereafter, the emphasis on ‘effectiveness’ found its paradigmatic expression in the aforementioned Montevideo Convention of 1933 (Grant Reference Grant1999), which enumerates the material criteria of statehood and stipulates that states exist independently of recognition. Their long history of opposition notwithstanding, international law does not clearly privilege either recognition or effectiveness as the marker of successful state creation. Instead, today it is widely held that ‘[n]either theory of recognition satisfactorily explains modern practice’ (Crawford Reference Crawford2006, 5; see also Vidmar Reference Vidmar2012, 701–704).

On the one hand, the constitutive theory has been rejected because it appears ‘unacceptable’ that ‘an entity otherwise qualifying as a State’ would be treated ‘as if it was not a State’ by states not formally recognizing it (Crawford Reference Crawford2006, 27). In terms of the classical critique launched by the declarative position, if a ‘State, (…) exists in fact, [it] must exist in law’ (Chen Reference Chen1951, 38) too. For if the identification of states was left exclusively to the discretion of other states, international law would be merely ‘an attorney's mantle artfully displayed on the shoulders of arbitrary power’ (Brierly Reference Brierly1944, 13 cited in Grant Reference Grant1999, 3). Premature or otherwise unjustified recognition is therefore also considered invalid (Crawford Reference Crawford2006, 21), although to determine prematurity obviously requires taking criteria of statehood other than recognition itself into account. Many international law scholars also doubt whether recognition, without consideration of effectiveness, would actually constitute real states (Kunz Reference Kunz1950, 718; Kreijen Reference Kreijen2004; Peters Reference Peters2010). Recognition, in other words, already presupposes their prior identification as entities ready to be recognized as proper states.

On the other hand, the declaratory doctrine mirrors the problems encountered by the de facto state claim when put to the empirical test. It rests on the view that in order to capture actual states beyond recognition, international law requires ‘“objective” criteria for statehood’ (Crawford Reference Crawford2006, 45). Yet even Ti-Chiang Chen (Reference Chen1951, 59), the great advocate of the declarative position, admits that the conditions for statehood ‘can only be stated in general terms,’ preventing their uncontestable application to specific circumstances. In particular, ‘[t]here is no bright line between effective and ineffective’ (Peters 2010, 3) 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’ (Crawford Reference Crawford2006, 59). As the attempts to empirically determine a sufficient degree of effectiveness of de facto states illustrate, the purely declarative position is ‘epistemologically naïve’ (Koskenniemi Reference Koskenniemi2001, 385). If situations of supposed effectiveness require an interpretation, and in the decentralized international system this interpretation is conducted by other states, then (non-)recognition as the expression of such interpretations has returned ‘by the back door’ (Crawford Reference Crawford2006, 28).

Thus, recognition by other states is an expression of their political preferences, rests on contestable interpretations of both law and facts, and does not necessarily create effective states on the ground. However, effective statehood is in turn not an obvious fact and interpretations of whether it is actually present or not legitimately differ (Koskenniemi Reference Koskenniemi2005, 272–281). Hence, although both doctrines seek to determine statehood, they are eventually forced to silently assume states that neither of them can actually grasp (Grzybowski Reference Grzybowski2017).

Bridging or escaping the gap: in search of solutions by legal regulation

Modern international law has sought to overcome this predicament in two different ways. First, many scholars have argued that ‘the differences between declaratory and constitutive schools are less in practice than has been depicted’ (Crawford Reference Crawford2006, 28). This has led to various formulas for compromise which emphasize the joint impact of the two factors, so that recognition requires a degree of effectiveness on the ground, while the effective statehood of any given entity is in turn enhanced by recognition (Wright Reference Wright1955, 324; Crawford Reference Crawford2006, 27; Peters Reference Peters2010, 5). The intuition that in practice different factors play a role in state making is essentially correct, as I shall argue in the next section, but within the constraints of international law these compromises are still geared toward determining states in terms of definite status, which reproduces the opposition between declaratory and constitutive approaches. Either the degree of effectiveness necessary for an entity to become a state upon recognition can be objectively assessed, then recognition is supplementary and withholding it appears to disregard actual statehood, or the entity's effectiveness is open to contestation but becomes sufficient upon recognition, in which case recognition is thought to determine statehood. It thus still appears that ‘at a fundamental level a choice has to be made’ (Crawford Reference Crawford2006, 27), one that is somewhat arbitrarily privileging one side over the other.

Second, legal scholars insist that statehood is a legal status and should therefore neither simply derive from facts nor from what are, ultimately, political decisions. The state is ‘not a fact in the sense in which a chair is a fact,’ but rather ‘a legal status attaching to a certain state of affairs by virtue of certain rules and practices’ (Crawford Reference Crawford2006, 5). Genuine legal rules in state creation include, for instance, the right to self-determination, the protection of territorial integrity of existing states, and the general prohibition of the use of force. That Palestine and Western Sahara have a right to statehood is neither a question of effectiveness nor of recognition, and the same holds true for the denial of this right to Southern Rhodesia between 1965 and 1979. Territorial disputes, self-determination conflicts, and questions of state dissolution and succession have thus inspired a vast body of legal knowledge and regulation that goes beyond a simple opposition between effectiveness and recognition (Kohen Reference Kohen2004; Crawford Reference Crawford2006; French Reference French2013).

However, the shift toward proper legal regulation does not deliver any objective ground for identifying states either. For one, the question of whether or not specific rights or prohibitions related to state creation are applicable at any point gives rise to controversial legal debates in its own right. The dispute over the status of Kosovo and the advisory opinion of the International Court of Justice is an illustrative case in point (Milanović and Wood Reference Milanović and Wood2015). For another, even where legal considerations align, regulation eclipses but does not replace state identification. For example, many de facto states have persisted for decades without legal entitlement, while the acknowledged right to self-determination and widespread recognition of Palestine and Western Sahara have not ended debates about their actual statehood that still revolve around questions of effectiveness and recognition (Crawford Reference Crawford2006, 434–448).

As international law is premised on both regulating a social reality of statehood outside of its own normative confines and deciding on the state status of individual entities by resorting to the interpretations and decisions of other states, it remains caught between indeterminate arguments about effectiveness and recognition. There is no escape from this circle, which revolves around the basic assumption that states really exist and must be individually identified, although any attempt to pinpoint the essence of states reveals its indeterminacy.

The foundational paradox of state identification

That IR and international law arrive at the same paradox is not accidental. While the two disciplinary perspectives have set different priorities and developed distinct approaches to come to terms with state creation, their shared predicament results from a prior distinction between individual states and the state system. It is this fundamental distinction which in the last instance generates all problems of state identification in international theory and practice, since it paradoxically grounds the state system on pre-existing states while regarding political entities as proper states only once they are part of the system. This essentially presents a ‘chicken and egg’ problem of whether individual states preceded the state system or vice versa (Coggins Reference Coggins2014, 21). Irresolvable at its root, the intertwined ontology of the state system and its component states renders the identification of individual states simultaneously necessary and impossible. As I explain below, however, we should not understand the paradox as a problem that requires a substantive solution, but rather as a foundational predicament that drives the (re-)productive dynamic of the state system itself. Before turning to the productive function of the paradox, I briefly explore how it has been rendered invisible by the division between legal and sociological approaches to the international and the state.

It might be tempting to trace the predicament of state identification between legal principles and political facts back to the disciplinary divide between international law and IR in the early 20^th century, as paradigmatically articulated by Lauterpacht and Morgenthau (Koskenniemi Reference Koskenniemi2001). However, both authors already presume an international system made up of states, as did Weber (1978) and Kelsen (Reference Kelsen1928, Reference Kelsen2005) before them in their respective sociological and legal approaches to the state. Both had developed rigorously disciplinary perspectives to overcome what they regarded as ‘metaphysics’ in state theory, that is, the unfounded presupposition of the state's existence. For Weber, it was in legal constructions of statehood that state metaphysics could be detected, whereas sociology as an ‘empirical science of behavior’ (Weber Reference Weber1978, 4) identified states with the ‘human community that (successfully) claims the monopoly of legitimate physical violence within a particular territory’ (Weber Reference Weber2004, 33). Conversely, for Kelsen (Reference Kelsen1928), such a monopoly could not exist as a matter of fact, but only as a legal assumption or ‘basic norm’ (Kelsen Reference Kelsen2005, 201). Ironically, when faced with the challenge of deciding which legal order applied in case of conflicting basic norms, Kelsen (Reference Kelsen1928, 98–99; Reference Kelsen2005) himself invoked empirical effectiveness as the determinate criterion. This effectively closes the circle, as the factual state depends on the legal state, and vice versa. Thus, rather than overcoming the ‘metaphysical’ presupposition of the state, the division between legal and sociological views not only further perpetuated but also concealed it.

In fact, the paradox of state identification emerges from the prior distinction between the domestic and the international as two sides of the same coin in the political order of modernity (Walker Reference Walker1993; Bartelson Reference Bartelson2001). Within this order, political entities can only become part of the state system if they are states, while in order to be states, they must be part of the state system. States cannot be reduced to either their international or their domestic side, they ontologically combine both. State identification confronts a paradox because any attempt to originally identify states – by either international status or domestic effectiveness – must presuppose these states as evidently already there. ‘Epistemic recognition’ is thus impossible, and yet necessary as the ‘starting point’ of the international.

Instead of regarding this paradox as a problem, we should see it as a fundamental driver of social order. In system-theoretical terms, paradoxes generate order by allowing for differentiations that reduce complexity by imposing particular forms of observation (Luhmann Reference Luhmann1993, 770). In this sense, the social form of the state reflects the imagined distinction between the domestic and the international that cannot be in turn reduced to anything else. Conversely, it has itself shaped social reality in such a way that states and the state system have become the primary lens through which political claims and conflicts are articulated, assessed, and settled. Indeed, the perceived need of international law and politics to identify individual states as well as the resilience of problems of state identification testifies to the continued dominance of the social form of the state. From this angle, state identification has never been the problem, but was always part of the solution, not because it identifies states out there, but because it enables their constitution, and that of the international, in diverse practices of ‘state’ making, recognition, and legal contestation.

In the everyday of international affairs, the foundational paradox of state identification has arguably become invisible behind routine practices. In most ‘international’ settings, practices appear as ‘competent’ precisely because they reflect a ‘background knowledge’ (Adler and Pouliot Reference Adler and Pouliot2011) that is premised on given state identities and the distinction between domestic and international law and politics. By contrast, contested states mark the indeterminate or undecidable line of distinction between the inside and the outside, and thus bring the foundational paradox back to the fore. What is significant about attempts to identify states via recognition or effectiveness is therefore not whether they appear to be right or wrong, but that they share a commitment to the form of the state and thus, in their competing attempts at state identification, inevitably contribute to the delimitation of the statist international order.

Unfolding the paradox: performative state making and the reconstruction of statist ontology

The paradox of state identification circumscribes the scope and limits of substantialist approaches to the state in international law and IR, but it does not mark the end of critical inquiries into how the paradox unfolds. Two important lines of research in this respect are the study of performative state making and the reconstruction of the ontology of the statist international. I consider each in turn before discussing how they shed light on their respective blind spots, and how they can jointly contribute to further explorations of state making and the paradox of state identification in international theory.

First, if states have no essence grounded in either material facts or normative principles and yet adhere to an ontological form of the state, then they can be said to be performatively enacted, even without ever becoming fully ‘present.’ This has been variously argued in IR both with regards to state making (Visoka Reference Visoka2018) and in other contexts (e.g. Weber Reference Weber1995, Reference Weber1998; Jackson and Nexon Reference Jackson and Nexon1999). Such performative state practices include, for instance, the constitution of agency in diplomatic relations and at international conferences (Wille Reference Wille2017; Visoka Reference Visoka2018), where the state represented by a delegation can become tacitly accepted and hence co-produced by other states and international organizations. Legal constructions and decisions can have a similar effect (Grzybowski and Koskenniemi Reference Grzybowski, Koskenniemi, Stirk and Schuett2015). This can be observed, for instance, in the accession of Palestine to the Rome Statute after having been granted the status of a ‘non-member observer state’ by the General Assembly of the United Nations (International Criminal Court 2014), or in the confirmation of the right of Western Sahara to its territorial waters, despite Moroccan occupation, that was reiterated by the European Court of Justice with respect to the EU–Moroccan fishery agreement (Court of Justice of the European Union 2018). Precisely because statehood is a comprehensive social order irreducible to any single attribute, the various legal, political, social, economic, and diplomatic practices interact with each other across different fields, from diplomatic representation through taxation and border controls to the display of symbols such as flags and uniforms.

Any turn to the enactment of states should take a multitude of such practices into account, and investigate how they inform each other and become identified with a particular state, say Kosovo or Somaliland, by both participants and outside observers. Identification thus goes hand in hand with various practices of state making, is multi-sited, and concerns state projects ranging from routinely uncontested Sweden or the United States to variously contested Palestine, Somaliland, or Kosovo. On this methodological basis, the particular effect of recognition, or the lack thereof, can be studied beyond the misleading focus on whether particular aspirants really are states or not, while findings about various state practices of de facto states can be considered without either assuming or trying to determine their state status. It is important to note, however, that while in this perspective the ontology of the state becomes visible in its performative effects, it is at the same time reproduced as the very frame through which we understand political claims and conflicts. Thus, even an inquiry into how the state form is enacted in practice reproduces this form intellectually.

The second major line of research I consider reconstructs the theoretical texture and historical diffusion of the political ontology of the state and the state system. It draws on two different strands of research that have recently become more closely tied together. First, critical international relations scholars have noted in mainly theoretical elaborations that it is extraordinarily difficulty to think of alternative political orders to the modern international because the distinctions between inside and outside, modern and pre-modern, state and state system, are too deeply entrenched in modern political thought (e.g. Walker Reference Walker1993, Reference Walker2010; Bartelson Reference Bartelson2001, Reference Bartelson2014; Prozorov Reference Prozorov2009). Second, the work of legal and conceptual historians on political formations predating modern nation states and international relations contributes to grasping how the political ontology of the state emerged, took hold, and diffused (e.g. Bartelson Reference Bartelson1995; Koskenniemi Reference Koskenniemi2001; Benton Reference Benton2010; Armitage Reference Armitage2013; Benton and Ford Reference Benton and Ford2016). Together, these two strands of work lend themselves to a historical reconstruction of the political ontology that underpins the paradox of state identification.

Ideally, such a reconstruction would include inquiring into how the vocabulary of the state and the international came to shape concrete political and legal claims, conflicts, and solutions over time. The history of the American declaration of independence, and its global diffusion as a template, is a prominent case in point (Armitage Reference Armitage2013, 215–232). Another is the emergence of new states in Latin America in the first decades of the 19^th century. As the Spanish empire imploded, elites in the colonies pushed for change, first under and then against the Crown. Despite inspirations from the American revolution and European enlightenment, they had no ready-made design for instituting such new states (Adelman Reference Adelman2006). Rather, the division between inside and outside emerged progressively through constitution-drafting, border struggles (Adelman Reference Adelman2006), and the development of the legal principle of uti possidetis juris in international law, which privileged ‘effectiveness’ over dynastic title (Fabry Reference Fabry2010, 50–70). To be sure, the episode indicates neither the beginning nor the end of a long process of constituting the statist international, but it illustrates how the forms of the state and the state system gradually became the self-evident conceptual lens through which we understand political disputes and solutions.

Although history may illuminate the emergence of the statist international and our predicament with state identification, it does not explain its current resilience. In fact, as the analysis above shows, the endurance of statist ontology in international theory is tied to the very practices of state identification, which in turn mostly react to claims of state creation. Indeed, advancing alternatives to statist ontology can only go so far, if the concrete problems posed and the proposed solutions reproduce the form of the state. As disputes over the status of Palestine, Kosovo, Western Sahara, Taiwan, or Abkhazia indicate, both observers of and participants in these conflicts continue to invoke the form of the state, whether to articulate secessionist claims or reject them in favor of the parent states. Hence, just as studies of the performative enactment of states cannot move outside of the ontology of the state, the study of political ontology cannot ignore the enactment of states in practice.

While the two broad lines of inquiry pursue different objectives, they are intimately connected and can be analytically interwoven to capture routines of, and potentially ruptures in, the (re-)production of states and the international. As such, the state form might have proved resilient in international theory not so much because of the impossibility of thinking beyond the state and the international per se, but rather because of its prevalence and reproduction in practice. Conversely, states might be re-enacted in practice not so much because social reality naturally consists of separate communities and territories, but because political actors and legal regulations invoke the form of the state for the satisfaction of political claims and the settlement of political conflicts. The difficulty of articulating political alternatives, both in concrete practices and in ontological reconstructions, is thus likely due to the multi-layered and multi-sited enactment of states and the state system, where practices tend to become entangled in the naturalized ontological form of the state.