Global Ordering and the Interaction of Communities of Practice: A Framework for Analysis

Several new frameworks for the study of the differentiation of and relations between global orders have recently been introduced to International Relations. This article demonstrates how the communities of practice (COP) framework provides complementary as well as novel answers to processes of global ordering. COP theory has become a thriving research framework and has led to substantial innovative work on the internal logics of international institutions, including the North Atlantic Treaty Organization and the European Union


Introduction
Several frameworks for conceptualizing the differentiation of global order into suborders have recently proliferated in the discipline.Frameworks, such as regime complexity, network theory, or field theory, share the ambition to provide new understandings of the dynamics inherent in suborders and how these shape global orders.In this contribution, we demonstrate how communities of practice (COP) theory provides an alternative account that offers genuine advantages over the other approaches.
COP theory is a thriving research framework in the International Relations (IR) discipline and cognate social science fields.Inspired by Etienne Wenger's (1998) groundbreaking work on COP, the framework was advanced in the field of IR in particular by Emanuel Adler (2005Adler ( , 2008 ) ).It has led to substantial innovative research on international institutions, including the North Atlantic Treaty Organization (NATO), the European Union (EU), or the World Bank, and a diverse spectrum of transnational communities, ranging from diplomats to pirates. 1 Indeed, COP theory has become one of the pivotal contributions in the thriving debate on practices in IR. 2  A community of practice presents a unique unit of analysis for IR.In this sense, it complements the ontological vocabulary of the discipline and competes with analytical aggregates, such as regimes, systems, fields, networks, or assemblages.Contrary to these concepts, where significant efforts have been made to spell out a general theory for IR, COP theory has so far not progressed into a genuine and more general theory that allows us to theorize and investigate global ordering at different scales.
In this contribution, we argue that in order to take key steps toward a more general IR theory based on COP, we need to explore how the international system and global order can be interpreted through such lenses.If COP are the constituent units of the international system, how do they interact with each other?Such questions not only prompt theorizing the interaction of COP, but they also demand examining the ways in which such interaction drives global ordering and the international system, and how such dynamics spur conflict and cooperation.
Scholars have not sufficiently theorized this interaction so far.The debate is characterized by a proliferation of studies that use COP theory to analyze single communities and their driving elements and logics at micro-scales.In order to generalize ideas from the COP approach, we argue, three distinct moves are required: First, we need to go beyond the study of isolated communities.Second, we need to recognize that the framework is open in scale and hence allows for larger scopes and the study of macro-level entities.Third, we need to theorize how community interaction takes place and to what effect.
To develop these arguments, we introduce recent theorizing on the interaction between differentiated social spaces ( Liu 2021 ).We also explore key categories of interaction introduced in the original works of Wenger (1998Wenger ( , 2000 ) ), which have so far received little attention.On this basis, we introduce a taxonomy of the interactions between COP.We distinguish between two different scales: (i) the interaction of communities through spaces , that is, how communities are nested in each other in distinct locales and how they meet in particular shared arenas and (ii) interaction through agents that perform particular boundary roles, here identified as "guardians," "brokers," and "space travelers."We first develop these concepts and mechanisms in the abstract and then turn to the case of global ocean governance to illuminate how the framework translates into a research strategy.Since COP theory is in principle open in scale, our conceptualizations will be likewise relevant to those interested in large-scale global phenomena as well as those interested in particular institutions or micro-situations.
ize and study the interaction of COP.The following section then details our conceptual outline of how COP interact.We sketch out a range of mechanisms of how interaction takes place and then illustrate these through the empirical case of ocean governance.We study the interaction between three ocean governance COP that are centered around the concepts of 'maritime security', 'blue economy', and 'ocean health'.In the concluding section, we call for systematic research on the emergence and interaction of COP that draws on these categorizations and further explores the ordering effects produced by them.

The State of COP Research in IR
Theorizing International Order(ing) Understanding the factors that drive types of international order has been a persistent focus in the field of IR.While international order was classically understood as one coherent space-"the international system"-mainly characterized and driven by the interaction of great powers, at least since the 1980s, a pluralistic understanding of order has evolved that understands the international as composed of various overlapping and competing orders.The 1970s and 1980s regime debate has done much to foreground this pluralistic understanding ( Behnke 1995 ), as has research on regional orders, such as security communities ( Adler and Barnett 1998 ).
Regime theory has been recurrently criticized for being particularly vague about the precise mechanisms that produce orders-or, phrased differently, how ordering unfolds and degrees of coherence between international units are achieved ( Steffek, Müller, and Behr 2021 ).Since the 2000s, different social theories have been brought into the debate to offer such mechanisms.This includes revamped versions of regime theories under the notion of regime complexity, as well as network and field theories.
Regime theories posit that orders emerge through actors negotiating and agreeing on norms and ideas that give coherence to different social spaces ( Kratochwil and Ruggie 1986 ).As the contemporary debate on regime complexity highlights, this often leads to the situation that contemporary problems of the international are objects within more than one regime, and hence a situation of complexity arises ( Raustiala and Victor 2004 ).The majority of regime theorists contend that the increasing regime complexity, resulting from the proliferation of international organizations and pluralization of non-state actors, has led to fragmentation, contestation, and disorder rather than overarching, fully integrated orders.While the concept of fragmentation is, in principle, value-free, theorists argue that regime overlaps with diverging norms are sources of competition ( Alter and Meunier 2009 , 19-20) and conflict ( Margulis 2013 )even a "pathology […] that threatens governance effectiveness" ( Abbott et al. 2015 , 7).As a result, fragmentation is considered a stumbling block toward building overarching global orders (e.g., Benvenisti and Downs 2007 ;Gomez-Mera 2016 ).This negative outlook on the ordering effects of regime overlaps has been countered by more recent work that demonstrates that norm conflicts do "not necessarily undermine the global legal order," for they are most often cooperatively managed ( Kreuder-Sonnen and Zürn 2020 , 359;Krisch et al. 2020 ).
Irrespective of how scholars evaluate the outcome of fragmentation, regime theorists predominantly hang on to a rationalist understanding of order formation, in which states are the key actors that produce international orders by agreeing on how a particular domain should be governed.What continues to be problematic in such an understanding is not only that the opportunity to disaggregate the state, which is anything but a coherent unit, is missed, but also that transnational actors, including international organizations, transnational movements, experts, or nongovernmental organizations, are often inappropriately considered.Moreover, the literature continues to lack a deeper understanding of how precisely the norms and ideas underpinning regimes are actually produced, and how they may constitute a source of coherence among rather than fragmentation across regimes.
Network theories, introduced in different variants since the early 2000s and prominently associated, for instance, with Anne-Marie Slaughter's A New World Order (2005), offer a different reading of international ordering.For Slaughter and others, orders are made through the interaction of a rich set of governmental officials ( Avant and Westerwinter 2016 ).The level of analysis, hence, moves to the interaction of subgovernmental units, with more recent research gradually integrating other types of officials and nongovernmental representatives.
Network theories open up the debate on ordering considerably since they leave the question of scale (that is, what size an international order may have) principally open: They approach the question of agency inductively by looking at who contributes to ordering.Identifying the ordering agents is the primary outcome of network analysis, which elucidates those nodal points where interactions come together.The long-standing literature on "transnational advocacy networks" (TANs) has been especially helpful in identifying those groups of actors that advocate for a specific cause.These groups, composed of nongovernmental or unofficial actors that share specific value commitments, arguably "mobilize information strategically to help create new issues and categories, and to persuade, pressurize, and gain leverage over much more powerful organizations and governments" ( Keck and Sikkink 1999 , 89).Actively creating links across a diverse number of stakeholders, TANs promote key norms of the international system and have the power to reconfigure or even transform the structure of the international system.Network theory also opens up possibilities for quantitative analysis and considering large-scale datasets of interactions, and thereby allows for the identification of clusters and patterns through replicable methodology.The core limit of quantitative network analyses is, however, that they tend to reduce order to quantifiable forms of interaction.Hence, they often have little to say about the quality of interactions and overlook potentially other forms of coherence, such as the prevalence of joint understandings or the norms highlighted in the regime debate.
In contrast, qualitatively oriented network analysis considers shared norms and values as critical to a network's coherence and success.Yet, this strand of research comes with its own limits, making value commitments a prerequisite for rather than the result of participation in a network.This significantly raises the bar for engaging in joint projects because value diversity arguably hinders the formation of networks.A further shortcoming relates to the rational actor assumptions underlying network analysis: Agents are here understood as strategic, utility-maximizing actors who seek to promote their campaigns through issue framing, pressure politics, and persuasion.The knowledge to be diffused is here reduced to mere information that can be traded as a commodity in exchange for other services.What is essentially left out of view in this framework is the prereflexive and practical know-how that is enacted in and through prac-tice.This know-how is, however, key to a group's coherence.As a condition for intentionality, this background knowledge orients its members' practices toward making sense of the world and constructing their goals, principles, and "rules of engagement" in the first place ( Adler 2005 , 20).
A third and more recent contender for the explanation of ordering is field theory.Originally associated with the work of Pierre Bourdieu, IR is increasingly moving forward the wider debate on international fields ( Nexon and Neumann 2018 ;Musgrave and Nexon 2018 ).Field theories join network theories in arguing that interaction is the key mechanism in driving ordering.Yet, they go beyond network theories and add important insights on the quality of interactions to explain both the coherence and differentiation of fields.Through their interaction, actors gain a practical sense of what is at stake in a field, and they strategically compete over these stakes.Such stakes might be authority, recognition, or material resources that actors participating in the field aim to gain.
Field theorists also increasingly highlight the importance of boundary discourses, that is, debates among participants on what belongs to a field and what does not ( Lesch and Loh 2022 ).Consequently, analysis starts out from observing the interaction in and across fields (often drawing on similar quantitative tools, such as network theory), and then seeks to understand the stakes in a field and its boundaries.While very promising, the way in which field analysis pays attention to the qualities of interaction risks reducing such qualities to competition and struggles among actors.In doing so, it neglects other forms of coherence, such as cooperation that proceeds along shared normative understandings.
The COP framework shares many assumptions of network theory and field theory,3 yet it brings a thicker understanding of different forms of interaction and shared understandings that enriches our understanding of ordering.As explained in the next section, COP are distinct forms of orders; like networks and fields, they are open in scale and center on the study of interaction.Yet, the approach adds important additional dimensions concerning the quality of interaction and sources of coherence, such as the importance of joint projects, prereflexive know-how, and shared normative understandings that build trust, mutual accountability, and commitment.It highlights how the interaction across COP can significantly contribute to the ordering of global spaces and thus provide impetus for cooperation rather than competition and conflict in the global realm.
In the next step, we introduce the COP framework and discuss the mechanisms it identifies.In order to advance the approach and bring it on a par with regime complexity, field theory, and network debates, we argue for the need to study the interaction of COP to scale up the approach and allow for observations on world ordering more broadly.

Communities of Practice
COP were introduced to IR as a core unit of analysis within the "practice turn."One of the core claims of the practice turn is that order is the outcome of practices and their enactment.This implies thinking of order as an ongoing achievement, which is why the majority of international practice theorists prefer the verb "ordering" ( Bueger 2014 , 393;Bueger and Edmunds 2021 ).While different understandings of the concept of "practice" prevail, there is some agreement to conceive of practice as an organized nexus of doing and sayings, with some divergence over what precisely organizes the nexus.The organization of the nexus is linked to a background of shared practical understandings, that is, standards of how a practice should be performed and when an activity becomes recognizable to constitute a practice.These standards have been variously understood as norms ( Wiener 2014( Wiener , 2018 ) ), tacit and formal rules ( Frost and Lechner 2016 ;Hofius 2016 ), and emotional stances or as future-looking goals and objectives ( Schatzki 2002 ).
In the COP framework, communities are the key form of order and container of practice to provide continuity over time.Wenger highlights three criteria that constitute COP: (i) the mutual engagement of actors through which they agree on the meaning of their activities; (ii) joint enterprises, that is, shared understandings of what matters, of what is worthwhile and requires attention, what requires justification or is tacitly accepted, and what actions require refinement and adjustment; and (iii) a shared repertoire, which includes all sorts of practical resources, ranging from concepts and metaphors to material tools and artifacts (see Bueger and Gadinger 2018 , 53-54).These three criteria allow for the identification of a community of practice at different scales.The framework primarily provides sensitizing concepts in that it gears attention to particular mechanisms through which coherence and communities are produced.Claiming that a particular configuration can be meaningfully understood as a community of practice therefore implies that the three criteria are fulfilled and evidence for mutual engagement, joint enterprises, and shared repertoires can be provided.

Communities of Practice in International Relations
In IR, the COP approach has been used primarily to rethink earlier concepts of transnational communities, such as security communities and epistemic communities.This analytical move is plausible since it allows authors to claim accumulative knowledge and draw on prior and established empirical notions of community.The North Atlantic community and its organizational form, NATO, for instance, have been frequently evoked as an example of a community of practice.Substantial earlier research has argued that there is a recognizable (nonwar) community, and what requires explanation is how it functions ( Adler and Barnett 1998 ).
Table 1 provides an overview of the studies in IR that use the COP framework.As the table documents, the majority of studies take a well-studied political configuration as their empirical case; they focus on either regional organizations, such as ASEAN, the EU, or NATO, or on recognized professions, such as diplomacy, international lawyers, or humanitarian aid workers.Conspicuously, communities are most often equated to the formal organization itself, with their boundaries neatly coinciding with those of the organization.Efforts to identify emerging or not-yet-known transnational COP that evade organizational classifications are sparse by comparison.Still, a handful of studies aim at conceptualizing piracy, terrorism, or crisis early warning systems in such a way ( Bueger 2013a ;Kenney 2020 ;Zwolski 2016 ).
Moreover, as Table 1 indicates, the vast majority of research investigates single COP.Only a minority of IR scholars have studied, for example, how two previously distinct professions or organizations come to form a community of practice ( Bueger 2013b ;Graeger 2016 ).Only more recently, a new wave of studies has become interested in the environment in which COP operate and, for instance, explores what happens at the boundaries of COP.Recognizing the need to examine the interaction between two or more COP, scholars have started to shift their analytical focus toward the boundaries of specific COP ( Adler and Greve 2009 ;Bremberg 2016 ;Hofius 2016 ), conceiving of boundaries as a testing ground for the relative depth, reach, and stability of shared practices.Hofius (2016Hofius ( , 2022 ) ) examines the "boundary work" of COP members, while Graeger (2016) and Sondarjee (2021) venture into the study of the practices and spaces in-between two communities, where individuals and groups traverse not only institutional and professional boundaries, but also more informal ones.These studies present valuable steps toward highlighting the informal nature of communities and the possible change in practices resulting from boundary encounters, knowledge exchanges, and the brokering of meaning across two or more communities.This is an important move towards generalizing the framework.as the nation-state.This encompassing perspective enables us to set the COP framework on a par with competing approaches, such as regime, field, and network theories, that aim at understanding global ordering and differentiation.

Interactive Forms: A Framework for the Study of COP
The objective of the following section is to identify a discrete set of ways in which COP interact.We draw on Wenger's original outline, concepts from symbolic interactionism that have inspired Wenger's formulation, as well as the recent theorization of the interaction of differentiated social spaces as advanced by Liu (2021) .On this basis, we present a framework of interactive forms with methodological intent.Similar to Liu (2021 , 125), we seek to theorize the "formal properties" of the interactions between COP rather than their substance, and therefore exclude hypothesizing about the possible motivations that bring about different forms of interaction.Our heuristic is structured into two distinct categories.The first category is spatial: Communities interact by meeting in spaces and at particular sites.They form relations that can be described as "nested," "overlapping," or "peripheral."The second category is agential: COP interact through functions performed by agents, which we conceptualize as "guarding," "brokering," and "space traveling."Taken together, these interactive forms provide a comprehensive framework for the study of the interactions of COP in global ordering.

Shared Spaces and Interaction of Communities
COP first interact by sharing spaces.In such spaces, they form relations with differing densities.One form of density is that of "nested" relations.This is the highest degree of interaction in terms of intensity of interactions, depth, and interdependency.A second form is "overlap" and provides looser ties: There is sustained and relatively deep interaction, though this does not imply that one community is embedded in another, but they function independently.A third form of density is "peripheral" interaction, which is marked by the least strong forms of interaction, and where there is little to no interdependence.
In the case of nested interaction, one or several COP are fully embedded and integrated within another one that might be broader in social and spatial scope.Accordingly, one community provides the space within which another one is situated.As Fligstein and McAdam ( 2005 , 59) indicate, in such a setup, we are looking at a Russian doll, "whereby actors that make up smaller collectivities are located within larger strategic action fields that contain larger collectivities."Some scholars, reflecting Adler's (2019) argument that COP are embedded in social orders, contend that situations of nestedness necessitate the introduction of a broader category to speak about the environment of COP.There is a rich body of work that draws upon Bourdieu's notion of "fields of practice," where various fields are embedded within one another, and particularly the field of power (see Eyal 2013 ;Steinmetz 2016 ).In his most recent formulation, Etienne Wenger calls for a concept of "landscapes of practice" with an emphasis "… on the multiplicity of practices involved, the importance of boundaries among them, and with problematizing identification and knowledgeability across these boundaries" ( Wenger-Trayner and Wenger-Trayner 2014 , 27).Since COP are open in scale, the introduction of an additional concept is, however, not required.Vital in embeddedness is that one community of practice "shares all its actors and positions within the larger space" ( Liu 2021 , 8).
Within nested interaction, there is a variety of relations depending on the structures of the overarching community.One form of structure is hierarchical, suggesting that only limited contestation can occur between nested communities due to the influence of the more encompassing community.Fligstein and McAdam (2005) , however, argue that dependence might go both ways.At times, the smaller community may be dependent on the larger, and at others, the larger community is dependent on the smaller ( Fligstein and McAdam (2005) , 60).Nested interaction also implies that communities can complement one another productively, avoiding any forms of contestation.
The middle proximate form of interaction is that of overlap: two COP maintain a direct and sustained interpenetration, creating a common area in which some actors from both COP may be located.This is not wholly nested, and practices and actors generally remain distinct, but interaction across this overlapping common area is marked by an overlap between communities, their practices, and some actors ( Wenger 1998 , 115-6).The complexity of this form of interaction is centered on the overlapping boundary space itself and the way in which actors operate within it.As Eyal ( 2013 , 175) argues, these "fuzzy zones of separation and connection" are less regulated than the core of the communities themselves.Spaces of overlap are not static.While they might have a degree of longevity, they have the potential to grow or decline as actors move in and out.In this form of interaction, there is a substantial degree of interdependence, and communities mutually influence the space of overlap ( Liu 2021 , 131).If sufficiently sustained, there is the potential for spaces of overlap to become independent COP in their own right.
Due to their nature, the overlapping zones can result in contestation or consensus.Competition, for example, may lead to the reinforcement of existing practices as interaction with outsiders gives actors within a community of practice something to relate through the contrast presented ( Liu 2021 ).In such a condition, the overlapping zone itself becomes an area of strong contestation, which is unable to permeate the communities' cores as it strengthens the boundary between the overlapping space and the core of the COP.If they are highly similar , however , in terms of their actors and practices, it may result in similar modes of interaction and consensus, or even allow for minor forms of renegotiation as the boundaries increasingly blur.The outcome of this form of interaction is heavily influenced by the actors within the overlapping zones, their positions, practices, and the nature of the overlap ( Barrett et al. 2012 ;Pyrko et al. 2019 ), an issue to which we turn below.
More socially distant are peripheral interactions between COP.While in spaces of overlap, interaction is relatively thick, in peripheral spaces it is thin.Interaction takes place on the outer edge of COP, rather than in a space between them.There is less sustained interaction and less interdependence, as interaction does not need to flow both ways.As Wenger (1998 , 117) argues, access to the community's boundary is casual but legitimate, and marked by observation or limited forms of interaction with this outer edge.These forms of interaction do not take place in an overlapping space, and boundaries remain relatively strong ( Abbott 2005 , 255;Liu 2021 , 129).They instead may be linked in some manner by boundary objects, common issues and interests, or actors who move across different communities ( Abbott 2005 , 255, 265).
Due to its relatively thin and shallow nature, a peripheral interactive form may be assumed to result in an outcome of less renegotiation, as interaction occurs on the boundary and is distant from the community's core.Wenger (1998 , 188), however, argues that as the rules are looser at the periphery, there is a greater likelihood for contestation to occur, and that the inside/outside duality of the boundary allows for this renegotiation to permeate deeper away from the periphery itself.

Interactive Agents: Guardians, Brokers, and Space Travelers
A second key dimension for the interaction of communities is agents.While surprisingly little empirical attention has been paid to the agents who "do" the interaction, 4 the COP framework is not short of considerations of the different roles actors may assume.Different roles have been conceptualized and here we focus on three ideal types: the broker, the guardian, and the space traveler.
In his original outline, Wenger (1998 , 105, 109-10) generically speaks of "brokers" to refer to those agents who enable the interaction of COP by participating in multiple communities and transferring elements of a practice from one to another.Inside organizations, he argues, brokers are usually individuals who are "in charge for special projects across functional units" and are by default tasked to span these functional boundaries on an everyday basis ( Wenger 1998 , 109).To function as competent brokers, however, they need sufficient legitimacy and authority in each of the communities they engage in.They also require special skills of translation, mediation, and perspective-taking to facilitate interaction across boundaries.
In 2000, Wenger nuanced this idea and proposed to think in terms of "boundary spanners," "roamers," "outposts," and "pairs" ( Wenger 2000 , 235-36)."Roamers" are portrayed as constantly moving in space, and do not necessarily belong to any one community of practice in particular."Outposts," by contrast, are rooted in one specific community of practice and regularly return to their home community from the frontlines ( Wenger 2000 , 235).Wenger's conception of "pairs" is to denote brokering "through a personal relationship between two people from different communities" ( Wenger 2000 , 236).These conceptualizations provide degrees of brokering, but little indication of the effects that they have on the interaction.
The literature on boundary work sheds light on Wenger's blind spots to highlight the diverse modes of boundary work between social spaces. 5In their survey on boundary work, Langley et al. (2019) identify three types: "Competitive boundary work," which denotes practices of demarcation for the purpose of sustaining distinctions; "collaborative boundary work," which implies practices of linking to facilitate collaboration and cooperation among these units; 4 To date, few studies exist that specifically zoom in on the key drivers and systematically analyze their different forms of interaction, including actors' backgrounds and repertoires, as well as the various roles they play and effects they have on the interaction dynamics between COP.Work in the IR literature on interorganizational cooperation has recently sought to capture the multiple roles that individual agents may play when international organizations interact.Schneiker and Joachim (2021) , for instance, refine Organ's (1971) "linking pin" concept in management and organization studies to demonstrate that "linking pins" can fulfill the dual functions of gatekeeping and "boundary spanning." 5In the cognate fields of sociology and STS, especially the sociology of professions, early work focused on the primarily exclusionary effects of boundary work, in which the demarcation of jurisdictions is considered key to maintaining a given boundary between professions ( Gieryn 1983 ), later work also acknowledges, even insists on the possibility of more inclusive effects ( Yagi and  and "configurational boundary work" through which purposeful individuals engage in practices of differentiating and/or integrating as to effect a specific broader design.Boundary work can hence be exclusive and inclusive, and can have a direct bearing on the spaces that communities share. In this light, Wenger's conceptualization of brokers is only one type of several.His brokers' practices are highly compatible with "collaborative boundary work," in which boundaries are actively "negotiated" ( Langley et al. 2019 , 715-7) and differences or divisions between groups "downplayed to get work done" ( Langley et al. 2019, 714).Conceptually, however, Wenger's "brokering" concept is unable to capture those constellations in which communities either refuse to collaborate and instead guard their distinctive practices or are ignorant of boundaries all along, preferring to move in and across multiple communities without active brokerage.Liu (2021) provides the most advanced current framework for how differently positioned actors engage in and enable the interaction of COP boundary work.As the most "basic categories" of a variety of actors, he distinguishes "guardians" from "brokers" and "space travelers" ( Liu 2021 , 132ff.)."Guardians" are gatekeepers who "guard" the expertise and repertoires of their community against external or new COP who threaten to intrude on their domain of practice.Accordingly, guardians seek to fend these off and erect boundaries that limit access to effect "social closure and monopoly over identity, resources, and status" ( Liu 2021 , 132).The likely result is the maintenance of "autonomy and distinction" ( Liu 2021 , 133), in which incumbents, who dominate a social space, continue to do so.It follows that no active brokering takes place unless incumbents see a strategic advantage to span boundaries across domains or challengers, such as external actors and novices, contest this form of domination.Guardians therefore aim at ensuring that individual COP remain distinct and separate.
With his concept of "brokers," Liu refers to those actors who Wenger equally conceptualized as competent enough to span and connect COP.Due to their long-time experience and work in influential positions, they are capable "to build bridges rather than barriers" ( Liu 2021 , 133).Through the work of brokers, communities become "porous" through the frequent interaction as well as the exchange of knowledge and resources ( Liu 2021 ).Whether two or more communities ultimately merge, however, is dependent on both the brokers and the structural prerequisites of the spaces of interaction.On the one hand, it depends on whether brokers see mutual benefits in merging or separation.On the other hand, it hinges on the structural similarity of the communities and whether their relations are already nested, overlapping, or peripheral.The frequent cooperation and exchange of resources might lead to "a division of labor" that is best accomplished in separate spaces rather than through mergers ( Liu 2021 , 134).Hence, COP might form close symbiotic relations, with participants moving across COP in a "revolving doors" fashion ( Liu 2021 , 128, 129).However, ultimately, they remain separate.
Space travelers, Liu's third type, are the most difficult to classify in terms of membership.They travel from post to post, either occasionally or regularly, and often do not occupy a fixed position in any one community.Consequently, they are not boundary workers, since they do not have a boundary to negotiate or defend.Instead, they embrace the styles and repertoires of different communities and therefore "develop plural and complex identities" ( Liu 2021 , 133).Their experience of multi-membership can also have adverse effects on their identity, entailing those "oc-Downloaded from https://academic.oup.com/isagsq/article/4/1/ksad079/7602888 by guest on 22 February 2024 cupational hazards" that Wenger associates with brokering: "[u]prootedness, homelessness, marginalization, and organizational invisibility" ( Wenger 2000 , 236).And yet, Liu suggests that space travelers "can be dominant actors in multiple social spaces […] without losing their status" ( Liu 2021 , 133).For COP interactions, this means that space travelers are essential to foster closely linked COP that, if not completely merged, reflect kindred relations, with actors adopting similar rules, ways of doing, and seeing.
In sum, the degree and quality of COP interactions are a result of both the spaces within which they are embedded and the actors through which they operate.Together, they define which ordering dynamics ensue from the interaction of communities, whether they are consensual and cooperative, competitive, or even conflictive.In what follows, we aim at illustrating the interaction forms by drawing on empirical instances from global ocean governance and oceanic orders.We do so largely with methodological intentions, that is, to demonstrate which empirical instances come into focus and what research questions the framework spurs.Living up to our expectation that COP theory is open in scale, our focus is on a high level of aggregation, and our cases are COP that are active on a global and transnational level.Ocean governance offers interesting empirical sites: The ocean is not only a space where civilizations have interacted on a global level since ancient times, but also a configuration of COP that has intensified in the past decade with the rise of several ocean crises and the political attention to them ( Bueger and Mallin 2023 ).Yet, it lacks attention from contemporary IR theorizing.

Illustration: The Interaction of COP in Ocean Governance
Global ocean governance is receiving growing attention in IR. 6 Yet, it is mainly international legal scholars who have investigated ocean governance as a form of global order.Here, the UN Convention on the Law of the Sea (UNCLOS), entering into force in 1994, is often seen as the "global constitution" of the ocean.Yet, several other international treaties and international and regional organization also attend to the space.Working mainly in the framework of regime theory, legal scholars describe ocean governance as highly complex as well as fragmented. 7They suggest that the growing complexity and fragmentation increasingly imply disorder.So far, little work has been conducted by drawing on alternative theories, and the oceans have not yet been studied from the viewpoint of newer theories of ordering.As such, ocean governance calls for new forms of theorizing that allow to cut through complexity and provide better understandings of the sources of fragmentation and contestation, as well as cooperation and order.
In the following, we explore ocean governance from a COP perspective, drawing on the outlined categories.We start out with a brief discussion of relevant COP that have been identified.We then show how ocean governance has been changing and can be understood as an increasingly socially differentiated social space in which COP interact with one another.In a final step, we provide examples of actors and discuss their importance in driving the interaction of COP.
A wide set of international actors uses the sea and governs it in one way or the other.On the one side, this includes users, such as the transport and extractive industries, fisheries, the leisure industry, and telecommunications.On the other side, a variety of state agencies address the oceans, ranging from law enforcement to security agencies (navies, coast guards, marine police, border agencies, fishing inspection agencies, and marine safety agencies) to various ministries involved in regulation and marine conservation (environment, trade, fisheries, economy, etc.).Some of these tasks are handled by regional organizations, such as the EU's European Maritime Safety Agency or the Indian Ocean Tuna Commission.
What are the relevant COP to understand global ocean governance?As argued in section "Communities of Practice," identifying COP implies detecting (i) mutual engagement, (ii) joint enterprises, and (iii) shared repertoires.While communities of leisure users, fishers, steamship captains, engineers, or submariners could certainly be identified in this way, such communities might not be the most relevant for understanding the governance of the oceans.Another indisputably useful way forward would be the identification of communities by drawing on established and well-documented international professions (diplomatic, legal, military, and safety).This would drive analysis toward the theory of professions and preconceived understandings of communities in which the ocean does not necessarily stand in the focus of attention (e.g., following the assumption that such communities operate roughly the same way-independently of whether they concern the land or sea).
In the following, we draw on a study by Bueger and Mallin (2023) .Bueger and Mallin start out with the observation that concepts are often the key vehicles of mutual engagement.They, hence, detect key contemporary ocean concepts and infer COP from them, showing that actors, which converge around these concepts, have developed significant joint enterprises and shared repertoires.Bueger and Mallin observe four ocean governance COP anchored in the concepts of maritime security, blue economy, ocean health, and blue justice.They argue that the latter, the blue justice community, is only nascent, and hence we will not include it in the following discussion.
The maritime security community of practice is composed of actors who understand oceans as a space of insecurity that requires protection from threats, including by military means.While the joint enterprise centers on securing marine activities, in particular shipping, the repertoire consists of shared securitizations and tools, such as law enforcement operations and maritime surveillance known as Maritime Domain Awareness.The blue economy community of practice interprets the oceans primarily as an economic realm that needs to be governed to ensure economic growth and new employment opportunities while paying attention to sustainability of business.The core joint enterprise is, hence, to find ways of sustainably exploiting ocean resources.Dedicated tools are part of the repertoire, including marine spatial planning, economic valuations, and blue economy strategies.The ocean health community of practice understands the oceans as an ecosystem and as an endangered space of biodiversity that is threatened by extraction, overuse, pollution, and climate change.Environmental protection and the rehabilitation of ecosystems, such as coral reefs, are the main enterprises of the community, with the repertoire consisting of oceanog-raphy , marine biology , systems analysis, or restoration techniques.
Each of these communities contains internal struggles over power and meaning, prioritizations, and the value of particular tools in the repertoire.Yet there is a sufficient level of coherence that allows for addressing ocean governance at this scale.Different professions are more relevant in one community than the other, and the importance of state-sponsored actors might also differ.In maritime security, private and public security professionals are important, while in the blue economy, it is economists and planners, and in ocean health, conservationists, biologists, and environmental agencies.State agencies are particularly important in maritime security, but they equally matter in the two other communities.Nongovernmental actors are active in all three communities, whether they are scientists, philanthropists, investors, technology companies, shipping giants, or global watchdogs.Grasping collectives beyond the classical dichotomies, such as state/nonstate, private/public, civil/military, or categories such as nationalities, is precisely the analytical power of the COP framework.
Bueger and Mallin do not provide much of an indication of how these communities interact, mainly pointing to the need for studying contestations and synergies.The framework laid out provides us with the opportunity to address this question through a focus on spaces and agents.In the following, we provide indicative examples that demonstrate what empirical foci and questions the framework of interactive forms spurs.We start with a discussion of spaces and turn to actors next.

Material Spaces of Interaction: The Ocean
To some degree, it could be argued that the "ocean" provides the space in which the three communities interact.Indeed, the argument can be made that within a larger ocean governance community of practice all three communities overlap.While this is true, it oversimplifies matters.The above-described communities have radically different understandings of the ontology of oceans, why they are in crisis, and how to govern them ( Bueger and Mallin 2023 ).Also, the assumption of spatial unity does not hold long.To start with, it is unclear whether ocean should be used in plural and whether there is a difference between "oceans" and "seas."With the ambiguous nature of ice, shifting tides, flooding, and sea level rises, it is also difficult to state where precisely the oceans start and the land ends.
Governance mechanisms have also created complexity.The proliferation of legal treaties and technical regimes has fragmented the oceans into different spatial constructions ( Peters et al. 2022 ).Even if one draws exclusively on the most foundational regime, UNCLOS, the ocean is already carved up into different zones: territorial waters, archipelagic seas, contiguous zones, exclusive economic zones, and international straits ( Ryan 2019 ).Each of these legal zones implies different rules, responsibilities, and rights prescribed under UNCLOS, and they are spaces that prescribe roles for states, leaving it to them to organize relations internally.Yet, the oceans are also carved up in a substantial number of functional and pragmatic spaces, which include search and rescue zones, transport corridors, marine protected areas (MPAs), no-fishing zones, areas of interest, and high-risk areas ( Bueger 2022 ).Given that these spaces are functional and problem-driven, they are more interesting in terms of the interaction of the three COP they host.
Planning Spaces: Marine Protected Areas A good case are Marine Protected Areas (MPAs).These are zones created for conservation and economic purposes and are an increasingly widespread national spatial tool in governing the oceans ( Gruby et al. 2016 ).They are spaces that cut across territorial waters (where states have full sovereignty) and Exclusive Economic Zones (where, under UNCLOS, states have not only the rights to resource exploitation but also obligations to protect the environment).As marine geographers have shown, MPAs are rich in political interaction, not the least given that they often have multiple purposes and objectives and are linked to other regional zones ( Gruby et al. 2020 ).
While they differ in type and form, MPAs provide an example of the overlap and peripheral interaction of the three ocean COP.Planning MPAs is a core response to the problem of the multiplicity of ocean uses with divergent interests (e.g., fishing, swimming, oil and gas extraction, or conservation).They are a tool in the repertoire of both the ocean health and blue economy communities, and indeed, the planning processes bring both of these communities together where they overlap.Yet, in these spatial planning and management processes, maritime security and law enforcement issues are often hardly considered or only vaguely addressed as matters of "compliance" ( Pieraccini, Coppa, and de Lucia 2017 ).In other words, while one would expect that the law enforcement of the maritime security community would be vital to protect such zones, in practices of planning and implementation, this community is marginal.Paradoxically, then, MPAs are a peripheral space for the maritime security community of practice.

Governance Spaces
Spaces where ocean COP interact are, however, not necessarily "in" the oceans.The oceans are also objects in other spaces.This includes spaces where the oceans are governed through practices, such as debate, consent formation, or rule-making.Such governance spaces include informal ocean summits, such as the UN Ocean Conference, the Our Ocean Conference, and the World Ocean Summit; coordination bodies on a global or regional level, such as UN-Oceans or regional Coast Guard Forums; and also formal international organizations that govern the sea, such as the International Maritime Organization or the UN Office on Drugs and Crime.The Our Oceans Conference, for instance, is heavily dominated by the ocean health community with the blue economy community nested or overlapping.The UN Office on Drugs and Crime runs a Global Maritime Crime Programme, which is primarily concerned with maritime law enforcement.There are some overlaps with the blue economy community, and ocean health sits at the periphery.

Virtual Spaces
Another case of spaces is virtual in kind.Communities interact through the mediation of digital technologies and onscreen realities.The majority of states and regional organizations run digital surveillance platforms, which provide representations of marine activities.Such digital platforms, often known as maritime domain or situational awareness systems, are developing a shared understanding of what happens at sea.At the heart of such systems is the projection of vessel movements on the basis of radar and satellite data.This is used not only to remotely monitor maritime transport routes and fishing activity, but also to identify anomalous behavior through automatic detection.
The space of interaction in such cases is a shared digital platform that different communities use for their understanding of (normal and suspicious) activity at sea.The regional network of such platforms is, in many ways, at the heart of the maritime security community of practice ( Bueger 2020 ).As a key law enforcement tool, such platforms are, however, also used by other communities.To provide one example, the European Maritime Safety Agency develops a joint surveillance picture for the EU institutions and member states ( Dupré and Guy 2012 ).This picture is not only used in the EU's naval operations, but it is also used to monitor oil spills, to conduct environmental assessments, and to improve the flow of shipping in the continent's waters.
The cases provide us with some empirical examples of where COP interact and what one might focus on in analyzing the interaction.Material spaces, such as the sea; planning spaces like MPAs; governing spaces, such as informal summits and international organizations, as well as virtual spaces, are all sites where we will be able to identify and study interactions and their density.

Interactive Ocean Agents
What should count as an actor within COP theory is an open question.While some would argue that the focus must be on individuals, others would include collective and corporate actors, while yet others would prefer the term actants and extend the notion to nonhumans.From an empirical perspective, we think that the inclusion of collective actors next to individuals is important, while the nonhuman extension leads too far into alternative theoretical terrains, that is, actor-network theory and assemblage theory ( Bueger and Gadinger 2018 ).In the following, we provide three illustrative cases of ocean governance actors that enable the interaction of the ocean COP.As with our discussion of spaces, these illustrations indicate on which kind of actor the researcher might want to focus.Each of the following cases illuminates the spectrum of empirical instances and provides a different form of agency (individual or collective, the latter in the form of corporations or international organizations) as well as an illustration for the three identified roles (space traveler , broker , and guardian).

Space Travelers
The adoption of the UN Sustainable Development Goal for the Oceans (SDG 14) led to a new office: In 2017, the United Nations Secretary-General appointed Ambassador Peter Thomson of Fiji as his Special Envoy for the Ocean.Since taking office, Thomson has become a major space traveler whose membership or loyalty cannot be associated with any one community in particular.8While a core part of his job is to attend ocean-related events, conferences, and summits, his objective is to forge shared frames of meaning and ways of understanding and seeing the ocean.
Specifically through speeches he gives at these events, he not only translates meaning between communities but also introduces tools across the communities.Thus, Thomson is an actor who brings communities into interaction, but without a distinct set of practices that emanates from a single community.An example is his call to submit "Ocean Voluntary Commitments" for the 2022 UN Oceans Conference ( Thomson 2022 ).Inclusionary in nature, all communities can present their practices as commitments, where he brings them together into interaction with one another.A commitment could be to develop "diverse collaboration frameworks" to tackle ocean plastics (ocean health), produce a "blue economy global report" (blue economy), or "combat IUU fishing" (maritime security).

Corporate Brokers
The Economist is not only a renowned media giant, but it has also expanded its portfolio as an event management company.One of its most important events is the organization of the annual World Ocean Summit.Launched in 2014, the summit has become a major event in the global ocean governance calendar ( The Economist 2023 ).Held annually, it brings together not only heads of states, delegations from different international organizations, philanthropists, celebrities, and activist groups, but also companies.It thereby acts as a broker who has sufficient epistemic authority and financial means to bring together and broker among two or more COP.Through the creation of a joint platform, The Economist is engaging in configurational boundary work: Though a traditional outsider to ocean governance, it has emerged as a broker that not only frames the debate through organizing the event program, but also moderates the discussions through its staff.It also actively designs the structural spaces in which ocean governance takes place through deliberation.
Different to the space traveler, however, The Economist identifies both the blue economy and ocean health, seeing them as relating to one another and therefore acting as a bridge.The practice of bringing two communities together can be seen in the World Ocean Summit, where panels are mixed with participants from both communities and brought into interaction to try to align practices for the shared goal of sustainability for the sake of future economic exploitation.It does not speak to maritime security, though.The exclusion of the maritime security community can be seen in its sign-up form.Participants are asked to register their sector and interest, and maritime security as well as its subordinate groupings, such as enforcement, are absent.This demonstrates The Economist's own position-a bridge that has competence within two COP, but not in the third.

Institutional Guardians
Our third example is the United Nations Security Council (UNSC).Increasingly involved in ocean governance, it acts as a powerful guardian of the maritime security community.Through its debates (three of which have explicitly focused on maritime security; many others concerned maritime threats), it defines the boundaries of what should count as a security issue and what should not.Part of the Council debates have been environmental and blue economy issues, where the members disagree, however, on whether these should be evaluated as security issues.It follows that the Council as an entity in its own right carefully prevents a tooclose interaction between maritime security and other communities.

Conclusion
In this article, we have outlined a refined version of COP theory that allows for studying processes of global ordering.Compared to approaches such as regime, network, or field theory, this framework not only provides an account for the differentiation of global orders, but also generates a thicker understanding of the ordering processes underlying it.The COP framework goes beyond the relational empti-ness of network theory and the prioritization of competition in field theory by substantiating the quality of interaction and sources of coherence within orders.While it does not deny the existence of contestation within and between COP, it foregrounds the predominance of cooperative dynamics that evolve from joint projects and collective normative understandings.
Two decades ago, Adler already argued to understand "the international system as a collection of communities of practice" ( Adler 2005 , 14).As we have shown, instead of taking this argument forward, COP research has narrowly zoomed into understanding the micro-mechanisms of isolated communities.We have argued that in order to allow COP theory to speak to IR more generally, and to understand systemic levels, it is time to zoom out.Once this is appreciated, the COP framework can become a contender for offering new systemic interpretations of global order, complementing and challenging regime, network, or field theories.In terms of its core unit of analysis, COP research must leave behind the mere study of isolated communities and formally organized settings and instead focus on the interaction of a diversity of transnationally operating COP that often traverse formal organizations.
The framework of interactive forms that we have developed in this article allows studying the interaction of and spaces between COP.Focusing on the spaces in-between COP can help researchers learn about the specific mechanisms by which global ordering unfolds.We have laid out a range of six interactive forms, including different spaces (nested, overlapping, and peripheral) and agents (guardians, brokers, and space travelers).These provide a promising heuristic for empirical analyses.Through the case of ocean governance, we have showcased how this framework offers intriguing new opportunities for understanding the dynamics of ordering, thereby challenging extant frameworks.As a case that is hardly studied, but is global and transnational in focus, global ocean governance is a particularly interesting instance to advance the theorization of global order.Yet, the framework is equally useful at other scales, whether global or not.In advancing empirical research driven by the framework, the important next step is to investigate if and how particular spaces and agents are related to outcomes and produce global ordering effects that are complementary, conflictual, or cooperative.