Security community: A future for a troubled concept?

Deutsch’s security community

Karl Deutsch did not invent the concept of security community. ² However, he coined the concept – first in his Political Community at the International Level ³ – and, together with colleagues, operationalised and mobilised it in the book Political Community and the North Atlantic Area ⁴ in a way that immortalised it. Since immortalisation of ideas habitually involves the processes of more or less severe reinterpretation, and even a violent appropriation undertaken to further concrete political and research agendas, it will be useful, as a point of departure, to first recall how security community was conceived by Deutsch himself.

Deutsch’s security communities are political communities which eliminate war and the expectation of war within their boundaries. Political communities, and therefore also security communities, are social groups with processes of political communication, some machinery for rules enforcement, and the habits of compliance. What makes security communities a discrete category of political communities is a degree of their integration. The degree of integration that makes it meaningful to speak of a security community is reached when, as a result of a sufficient density and intensity of transactions, a sense of community is achieved and institutions and practices developed. Taken together, these factors ensure dependable expectations of peaceful change over extended period of time. The sense of community, Deutsch posits, is conditioned on a sufficient amount of transactions. These transactions produce mutual sympathy and loyalty, trust and consideration and responsiveness. Statements jointly produced by representatives of the political elite in forms of declarations or even treaties will not do, irrespective of how frequently articulated they may be. Finally, since integration can be measured and scaled, Deutsch proposes it is possible to determine when and where a security community exists (but not whether it will last, see below). It may then be possible also to differentiate between two distinct kinds of security community: amalgamated and pluralistic. ⁵

By pioneering the concept of security community, Deutsch and his colleagues consciously confronted the inside/outside dichotomy entrenched in the field of knowledge about security (and famously subjected to a genealogical treatment by Walker). ⁶ They challenged the idea, conditioned on the Westphalian simulation, that the state is associated with peace, and international politics with war, and that security outside the state, whether in objective or subjective terms, may not endure. ⁷ Their intervention, in other words, consisted of opening the space for imagining that dependable expectations of peaceful change may exist also beyond the state – that the state is not the only conceivable security community there is, that by extension an organised government (in the traditional sense) is neither a necessary nor a sufficient condition of peace and, last but not least, that empirical evidence can be mustered to demonstrate that this is the case.

Indeed, the state continues to feature prominently in Deutsch and his collaborators’ thinking about security. They define pluralistic security communities primarily as ‘clusters’ of political units – that is, states – rather than, for example, as networks of transnational (or societal) transactions to which agents would be ‘hooking up’ ⁸ but without producing new durable social associations as a consequence. This shows that Deutsch’s was not a radical departure from state-centrism. What he was after was ‘zooming out’ from the Leviathan rather than cutting its head off.

An important consequence of this moderate restatement is a tension inherent in his security community theory. Pluralistic security communities are one type of security communities. Security communities are a kind of political communities, which are defined as social groups. But based on what Deutsch says about them, pluralistic security communities (clusters of states, rather than transnational associations) can hardly be conceived as social groups. Deutsch and his collaborators can indeed be credited with a new, theoretically grounded way of thinking about security beyond the state. But at the same time, at the immanent level, this way of thinking seems to contain a certain contradiction since the (pluralistic) security communities remain to be defined in static terms. ⁹

From a methodological perspective, Deutsch’s study of security communities takes place within the transactionalist framework that he devised. This approach called for an extensive, rigorous and hence time-consuming exploration and measurement of interactions (transactions) within and among societies, complemented by surveys of elite opinion and investigations of allocations of resources for war. Therefore, mobilisation of the concept of security community required a considerable, collective research effort. Only by means of such an effort, for example, Deutsch and his collaborators could come to a conclusion that instead of clear thresholds for the birth and death of a security community, there existed broad ‘zones of transition’ in which the communities would move, often there and back, and possibly over several decades.

Resurrection(s)

Despite its immortalisation and appropriation outside academia, in particular in the political discourse of North Atlantic Treaty Organization (NATO) (despite two members on its Southern flank, Turkey and Greece, assuming essentially war postures against each other), the concept of security community failed to produce an enduring research programme. Nevertheless, the concept did not fall into oblivion. In his famous article on the ‘pacifying’ role of the United States in Europe, for example, Josef Joffe states as a fact that Western Europe became a pluralistic security community in Deutsch’s sense of the word. ¹⁰ Joffe argues that Washington had paved the way for this development. But it would be only four decades after the concept’s original formulation when a momentous attempt to bring it back to academic life came with the publication of Security Communities (1998), a volume edited by Emanuel Adler and Michael Barnett. ¹¹ Suggesting an instance of a constitutive interplay between power and knowledge, ¹² their move came in the period when the notion of security community received a new purchase in the official discourse of NATO. The alliance was then more and more conceived of as a community in a new opposition to the previous state of being only an alliance. ¹³ The concept also entered the official discourse of another regional organisation, Association of Southeast Asian Nations (ASEAN), which later formally declared that it was a security community at the Bali summit (2003). ¹⁴ Adler and Barnett imply as much when they propose in their introduction that time has finally arrived to fully benefit from Deutsch’s observations which ‘seem particularly relevant in the present moment because of changes in global politics and international relations theory’. ¹⁵ However, their attempt at resurrecting the concept of security community, while the most ambitious so far and successful in terms of generating a following, is characterised by several difficulties. A short discussion of these difficulties that follows is not intended as a much belated review. ¹⁶ Instead, in walking the via negativa, the aim is to finally arrive at the positive conclusion about the continuing inspiration of security community as a concept for contemporary security studies.

Adler and Barnett’s appropriation of Deutsch is at once complex and superficial. It is complex theoretically and superficial methodologically. The complexity is a consequence of their colonisation of Deutsch, which consists of anachronistically remaking him into a proto-constructivist who was interested in ideational forces, shared norms and knowledge(s) but who was also a prisoner to the behaviourist epistemology that left little space to interpretative approaches. In developing their ‘reconstructed architecture’ for the study of security communities, Adler and Barnett aim to, simultaneously, ‘refine’ Deutsch’s original formulation of security community theory, ¹⁷ ‘correct for its [theoretical and methodological] shortcomings’, ¹⁸ and ‘provide further evidence of the potential insights of, the constructivist approach to international relations’. ¹⁹ The recent advances in International Relations theory (in addition to the changed nature of international politics) that make the resurrection of security communities as a concept so timely therefore turn out to consist, above all, of the development of their own via media constructivism.

This has several theoretical implications. First, in their definition of a (security) community, Adler and Barnett leave out communication as the foundation of Deutsch’s concept, and substitute it with a much more ambiguous and less tangible assemblage of common meanings, interactions and diffuse reciprocity. This is a major step away from Deutsch’s parsimonious model which features a few definitions and hypothesised causal relations. Second, while in its general form their definition of a pluralistic security community as a ‘transnational region comprised of sovereign states whose people maintain dependable expectations of peaceful change’ ²⁰ is simple, it is followed by introduction of imagined or cognitive regions that do not have to be spatially contiguous (a category in which the Transatlantic area can be included). ²¹ Therefore, to use their own example, a security community may exist between the United States and Israel. ²² It is beyond doubt that the two enjoy friendly relations, comparatively thick societal transactions between their countries, and that a deep security cooperation exists between them. It is another matter, however, to posit the existence of a ‘region’ as not exclusive, but still somehow (de)limited space that their societies intersubjectively share. The introduction of ‘cognitive regions’ makes the concept of security community yet more difficult to comprehend (and empirical validity of such intersubjectively shared geographical notion is disputable).

Third, Adler and Barnett develop a complex model of the emergence of security communities featuring a triad of precipitating factors. These are structural (power and knowledge) and process variables (transactions, organisations and social learning) and mutual trust and collective identity. However, elements of this complex model are underspecified. It is unclear, for example, how are the shared meanings to be recognised – instead, in a shift to a normative mode, it is posited that ‘meanings’ of liberalism and democracy must be shared in security communities. ²³ Nor is it clear which actors should be sharing them. In the definition of a pluralistic security community mentioned above, it is the people of sovereign states that form the community. In their continuous emphasis on the role of institutions, Adler and Barnett effectively focus exclusively on the social practice (such as social learning) of the political elites as architects and inhabitants of those institutions. The related problem is that while the importance of institutions need not be doubted, Adler and Barnett, their collaborators and many others who have used the concept recently limit their investigations to formal institutions (often empty shells), paying only limited attention to other sites and fields where potentially significant social practices take place.

To give further precision to the concept of security community, Adler and Barnett introduce the triad of types depending on the state of its evolution: nascent, ascendant and mature. ²⁴ However, the baseline for the nascent security community seems so low that such security communities can be observed (and indeed are, see below) almost everywhere, including at the global level, where Bruce Russett finds their traces. ²⁵ Moreover, the specification of indicators for the two other types of security communities Adler and Barnett introduce raises doubts about including such indicators as the perception of common threat (which also features among the triggering factors). The typology shows their progressivist liberal bias in projecting a linear roadmap to peace which predominantly consists of ever deeper formal institutionalisation. ²⁶ The final stage of the process of building a security community is therefore not existence of a ‘loosely coupled’ security community, irrespective of whether it meets the condition of dependable expectations of peaceful change, demonstrated inter alia in changes in unfortified borders and military planning. It is, instead, a community which in close integration of defence and security policies or free movement of populations and generally ‘international’ (but effectively regional) government cannot but resemble the design of a nation state. What Deutsch and his collaborators describe as a specific outcome of a historical process in which different political communities such as the North American colonies become closely integrated (‘amalgamated’) seems now to be proposed as the desirable end state of any regional security building process. The traditional liberal linkage between state and security is not only preserved but reproduced and reinforced in the form of a state of states, notwithstanding Adler and Barnett’s declaration of turning away from structures to interactions (and therefore from institutional to ‘individual’ integration). ²⁷ The Leviathan is not confronted, it is upgraded.

The methodological superficiality of Adler, Barnett and most of their colleagues’ attempts at the revival of the security community research agenda is not a necessary consequence of a particular theoretical perspective (via media constructivism), even if the theoretical complexity that comes with it makes the research design more challenging. It is rather a consequence of a failure to internalise the ethos of Deutschian research: empirical rigour, analytical depth and focus. Indeed, it seems that the depth of analysis was traded for the width as the authors are more interested in how the concept of security community ‘travels’ around the world. ²⁸ As a result, security community emerges as a floating signifier, that is, one which is free to absorb whatever meaning both the author and the reader will impose on it, and which as a consequence is a symbolically potent but unstable concept. The authors of the case studies collected in the volume are actually aware that their cases such as the Organization for Security and Co-Operation in Europe (OSCE) or the Gulf Cooperation Council (GCC) may in fact not be security communities. Yet, they seem to feel compelled to investigate those cases anyway. In such investigations, security community is turned into a kind of fetish, an object that is constantly evoked and attributed major significance. ²⁹ At the same time, as noted above, conceptually, security community in this rendition is a floating signifier, and therefore, it is related not to a stable ‘object’ but a variety of signifiers: from practices aiming to develop trust such as seminar diplomacy ³⁰ to the development of institutional identity. ³¹

Adler and Barnett’s claim that failure of the security community concept as articulated by Deutsch to beget a research programme was due to theoretical and methodological problems encoded in it may therefore be seen as somewhat ironic. It has to be admitted, however – whether or not this has been the result of a more favourable power constellation – that they have inspired a number of other publications that mobilise the concept with reference to Security Communities (while Deutsch is mentioned merely in passing, if at all). A survey of this literature shows that, by and large, it reproduces the problems at the heart of Adler and Barnett’s rendering of security community as a concept. In consequence, nascent security communities are found everywhere, usually with the baseline indicator in the form of an organisational (i.e. formal institutional) framework and the absence of war (rather than expectations of peaceful change). The existence of an organisation and the absence of war are simultaneously the independent and dependent variables, proving the presence of security, whose robust institutionalisation would only make the peace more ‘durable’. Poor specification of the concept in the constructivist rendition together with (paradoxically) the reliance on formal regional institutions means that existence of a security community is impossible to falsify, and that there is no way to demonstrate that a political project with the ambition to build a security community has failed. ³² In replication of the liberal progressivism engrained in Adler and Barnett’s theory, security community, defined primarily by complex institutional structure mirroring modern nation state design, is the utopia of regional security arrangement ‘towards’ which relevant actors – states, or their political elites – aspire or should aspire. Finally, security communities are often conceived instrumentally as serving states’ interests (including through coordination of common defence against external threats) rather than as results of existing dependable expectations of peaceful change among the concerned societies. Alternatively, the attention is refocused to include their external (rather than only internal) effects. ³³

This can be readily demonstrated on publications focusing on the Transatlantic area as the initial locus for the study of which the concept was developed. In the 1990s, and in line with the political discourse, it could be argued with little contestation (and without the need to assemble much empirical evidence) that NATO represented a security community, even as realists had qualms about Europe holding together. ³⁴ Shortly before 9/11, Robert Jervis could still argue that the security community would not disintegrate in foreseeable future (although its ongoing existence was contingent on peace and prosperity on which it rests). ³⁵ But over the last decade, the crises in the Transatlantic relations changed parameters of the debate. Around the Iraq War (2003), it could be forcefully argued, for example, that Europeans and Americans are somehow different species inhabiting different worlds. ³⁶ Broader differences over the Global War on Terror, as well as the perceived lack of interest of the Obama administration in Europe and his alleged appeasement towards Russia (accompanied by a shift of focus of American grand strategy towards the Pacific) compounded by renewed fears of Washington and Moscow’s condominium over the continent, together with more specific issues such as the most recent National Security Agency (NSA) and Government Communications Headquarters (GCHQ) spying scandals spurred debates whether the Transatlantic security community still exists. ³⁷ By and large, what participants in these debates speak of when they use the concept of security community is not expectations of peaceful change among societies in the Transatlantic area, but instead (a lack of) trust among governments, alliance cohesion and normative differences in the conduct of foreign and security policy towards the rest of the world. Therefore, security community primarily stands for (instrumentally conceived) a cooperative arrangement that at best is stabilised by common norms and identity. This seems to be the case even with the more recent publications. For example, Adrian Pabst finds new opportunities for the emergence of European security community in the US strategic shift (‘Pivot to Asia’) and advises that it would best serve Washington’s interests to encourage European autonomy while participating in the new security community, apparently conceived as an architecture for the provision of common goods. The participation in such a project would be a matter of simple political choice. ³⁸ Michael Rühle predicts, in his turn, that the most likely outcome of drivers such as globalisation, financial crisis and the Pivot to Asia will be a Transatlantic community with ‘lower ambitions’ – a statement which too makes sense only if security community is considered as a cooperative arrangement for states. ³⁹ The same understanding underlies the argument of Andrew Porter and Annegret Bendiek who demonstrate the existence of the Transatlantic security community despite imperfections in norm convergence observed in the area of counter-terrorism cooperation. ⁴⁰

In parallel to the debate about the (non)existence of the Transatlantic security community, the canonical status of ‘Norden’, its territorial segment traditionally conceived of (including by Deutsch himself) as paradigmatic and uncontested case of a successful security community, has also been challenged in several recent publications by authors such as Ole Wæver and Hakan Wiberg. Indeed, it is not disputed that a dramatic change in incidence of war occurred in this region following the Napoleonic Wars. There surely were conflicts, some of the most famous of which are the secession of Norway (1905) or the dispute concerning the Åland Islands after the World War I (WWI), but these have been cases of ‘non-wars’. However, it is problematic to attribute these outcomes to the existence of a security community which, in Deutsch’s conception, should emerge around a core and either a military alliance based on a common threat, or an economic alliance formed for a mutual benefit. Neither was the case in ‘Norden’ where institutionalisation of any similar kind has been weak. On the other hand, communication patterns together with cognitive and normative developments reinforcing the sense of a shared (if not common) identity can be detected (e.g. language, religion, ideology of ‘nordism’ initially promoted by elites but later internalised more broadly or norms of neutrality and pacifism). ⁴¹ This leads Wiberg to an interesting conclusion that the ‘Realist elements’ in Deutsch’s thought (which in light of what has been argued above can easily be identified with his residual state-centrism) make ‘Norden’ as a case problematic, whereas these more social factors can indeed be seen at play – even if alongside others such as fortuna. ⁴² But the debate whether security community is a useful analytical tool to investigate (non)incidence of war in ‘Norden’ seems to be ongoing. ⁴³ Some argue that its ‘model’ status as a security community actually prevents a closer scrutiny of the reasons why war seems to have banished from the region. ⁴⁴

As suggested above, the concept of security community is not limited to the Transatlantic area but, both in the works of Adler and Barnett’s collaborators and followers, it has indeed ‘travelled’: (predominantly ‘nascent’) security communities are now being discovered, or at least projected near everywhere. In Central Eurasia, Marc Lanteigne traces the development of the Shanghai Cooperation Organisation (SCO) from a consultative body into a security community based on growing institutional complexity. ⁴⁵ This is less surprising when considering that security community is actually taken to mean the emergence of SCO as a ‘strategic actor’. But the argument remains curious, not least given that the organisation includes Russia and China. Naison Ngoma claims that Southern African Development Community (SADC) is moving towards a security community, again mostly on institutional grounds. ⁴⁶ Rut Diamint makes the same case for South and Central America, observing an increased coordination among ministries. ⁴⁷ Oksana Antonenko and Carol Weaver both conclude that the Black Sea region is currently not a security community, but they suggest ways or conditions for it to become one. In the case of the former as a result of modified threat perceptions vis-a-vis Russia, and in the case of the latter through establishment of ‘balanced multipolarity’. ⁴⁸ Similarly, Sheryl Cross and her collaborators conclude that the Balkans is an ascendant security community despite the region’s recent troubled history and admitted lack of common trust – a lack of trust which in Bosnia and Kosovo is clearly present also on the (sub)state level. ⁴⁹

A special case of security community in academic literature is the ASEAN. Amitav Acharya used the concept in relation to ASEAN already in the early 1990s, that is, before Adler and Barnett’s intervention. He later embraced their theoretical framework for further, more detailed explorations of the constitutive effects of norms and intersubjective factors in the construction of security communities. ⁵⁰ Impressed by the progress the ASEAN has made from the dire conditions in the 1960s to the present, Acharya explains the (allegedly) changed status quo by development of a security community which he identifies as a type of security cooperation comparable to security regimes, collective defence and collective security. ⁵¹ Thus, he makes clear from the beginning that in his view, the security community is a state (as both institution and condition) of states, and the intersubjective identity relates to these states’ political elites. ⁵² This line of reasoning becomes even clearer when, recognising the domestic political conditions in the region, Acharya argues contra Adler and Barnett that while indeed shared values are necessary for a security community to exist, these do not have to be the values of democracy and liberalism. While there is nothing to a priori invalidate such argument – and Acharya is correct in maintaining that Adler and Barnett’s liberal bias is not present in Deutsch ⁵³ – from what follows, it emerges that the norm shared by the regional elites is the ‘illiberal peace’. ⁵⁴ The member countries’ elites form an unholy alliance of more or less authoritarian regimes based on the principle of non-intervention in each other’s domestic affairs. They attempt to cooperate, with variable success, to collectively withstand external pressures to reform domestically. The public sentiment, whatever it may be, is excluded from the analysis. Therefore, despite Acharya’s privileged voice convincing us of the opposite, the existence of the ASEAN as a security community too can be reasonably doubted (and has been doubted since the argument was first articulated), ⁵⁵ and this particular ‘community’ can be even radically reframed as a ‘failing realist security institution’. ⁵⁶

Potential of security community: three conditions

All the previous criticism notwithstanding, security community is a concept that contemporary security studies can draw much benefit from. However, for its potential to be fully demonstrated, the new security community research agenda must be removed from the theoretical complexity and methodological superficiality that has so far characterised its most ambitious ‘second coming’, which in turn has affected how the concept has been mobilised in the broader literature. This removal does not necessarily mean embracing Deutsch’s methodological assumptions about the possibility to measure and model social reality and the ways of doing it – that is, the ‘optics’ to confirm or disconfirm hypotheses – which to researchers engaging in either quantitative or qualitative inquiries may now seem out-dated. What is called for is instead the theoretical specification of the concept which would retain the potential for variability of its manifestations (most of all in terms of scale), which would restore the ethos of rigorous, transparent, collective and transdisciplinary research in its investigations, and which would be less centred on formal institutions and more on social practices across different sites and fields. ⁵⁷ Such research could be taken in several directions. One direction may involve heeding further Deutsch’s call for the synthesis of the sciences and humanities ⁵⁸ that Hayward Alker sought to respond to with his ‘humanistic methodologies’. ⁵⁹ Another (and not necessarily alternative) direction might involve international political sociology situated at the intersection of sociology and international relations that takes interest in political associations (and their frontiers) and security beyond the state, and does not privilege the gaze of/on the elites, but credits also subaltern practices and subjugated knowledges. ⁶⁰

Admittedly, the expansion of liberal (market) governmentality in the Western academia, privileging the ‘practical’ over the theoretical, quantity of publications over quality of research and to some extent also individual over collective research work, is a structural obstacle to any attempt at such restoration. At the same time, the research and development funding provided by the European Union, for example, may limit these effects if used to support similar projects that otherwise could not be funded from national sources. Deutsch’s ideas about research featured emergence of transnational research networks as means of communication among scholars from different countries that would advance the process of building and testing theories through better peer review, facilitate exchange of existing research or bring about better and more robust data banks. As such, these ideas are entirely commensurable with the ambition to develop a common European Research Area (ERA).

Second, the new research agenda must take seriously the challenge of the realist security paradigm and pursue it further than both Deutsch (‘zooming out’ from the state to a pluralistic security community as a cluster of states) and those who have followed him in investigations of security communities (envisioning this community as a ‘state of states’) have done so far. This is not to doctrinarily ignore national boundaries and their relevance in global politics, whatever imagined their ontological status may be. It is rather about striving for a fundamentally open mind capable of thinking about boundaries of security communities which may not necessarily coincide with national borders of one or more states. For example, when suggesting how their scenario-building method that instead of seeking precise point forecasts should broaden our possibilities to think about future, Bernstein and his colleagues imagine a future where national boundaries lose even more of their practical importance. ⁶¹ Their significance might be assumed by boundaries separating markets where security will be provided by private security companies but which will be discrete social and cultural, if not political, spaces.

Third, the agenda must expand the scope of research to seriously consider processes of security communities’ disintegration. Notwithstanding his positivist optimism – manifested, above all, in the belief that the accumulation of knowledge and the use of computerised modelling would make possible both reliable forecasting and policy steering – Deutsch was clear from the beginning that the process of social communication can cut both ways: it can lead to both integration and disintegration. In normative terms, he warns on these grounds that integration in systems with insufficient capacities to handle communication flows could have adverse political effects, and argues against the moves towards world government or a hasty integration in Europe. ⁶² In The Nerves of Government, Deutsch includes extensive discussion of systemic failure, and in Tides among Nations, he ventures as far as arguing that internationalism, constitutive of emergence of large groups with uneven distribution of communication, was conducive to societal fragmentation into translocal and local communication code users as well as those not belonging to either group. The communication produced within each group was only a ‘noise’ to the others. ⁶³

The more recent literature on security communities, however, does not engage with disintegration, or does so only in passing. Adler and Barnett do not discuss it almost at all, their near exclusive focus being on the security communities’ emergence and linear development into institutionally more complex forms that should guarantee permanence of common identity among populations (even though the processes of social learning seem limited to elites frequenting certain social sites). ⁶⁴ Acharya is a partial exception in this respect as indeed he discusses disintegration when laying down his theoretical assumptions about security community. ⁶⁵ Drawing on Deutsch’s argument that increased socialisation and interactions bring about strains and burdens, he introduces the concept of ‘decadent’ security community to complement Adler and Barnett’s triad. ⁶⁶ Acharya suggests trends that may affect a security community’s decay: resource scarcity, enlargement (creating a need for integration of new members with incompatible values), but also notes effects of globalisation creating winners and losers within existing security communities, and (confirming his ideological status quo bias) destabilisational impact of democracy. However, even Acharya does not use the concept of decadent community in his empirical inquiries which are focused on the ASEAN, an organisation about which he is unabashedly optimistic.

Conclusion: a new research agenda?

If these three conditions are met, the concept of security community may pave a way to a theoretically accessible alternative paradigm challenging the security pledge of the state and a solid critical vantage point from which to assess, for example, the practice of ‘phantom institutions’ building. ⁶⁷ Against the status quo presentism, it can furthermore offer a truly historical perspective in which political communities are not immortalised once (and as long as) their institutions are in place, but rather are seen as possibly ‘wavering’ and transient: always containing the possibility of crossing in either direction the line which makes and unmakes them as security communities. What could be, in more concrete terms, the research agenda of which the concept would be a founding stone?

In principle, there are few spatial limits that should define it (however, as noted above, it is dubious how much can be gained when ‘imagined’ regions are taken to open the possibility of communities at long distance). In this conclusion, just two suggestions are made. The first is to turn attention to Africa which offers a wealth of interesting cases of both substate and transborder ethnic, kinship and other (e.g. trade) communities that may, or may not, on closer inspection turn out to be security communities. There can be little doubt that such security communities will not be found to exist as coinciding with the territorial borders of, for instance, Somalia (a security regime may actually be found to exist here that regulates the clan violence, thus possibly rendering Somalia not a state defined by a hierarchical authority, but a ‘regulated anarchy’), Central African Republic (where the consequences of ‘unregulated anarchy’ are now being painfully felt), Libya (where the government relies for the provision of security on more than thousand militias, some formed on neighbourhood, some on political or ethnic principles, and refusing to give up their fiefdoms gained after Muammar Qaddafi’s fall) or other places where ‘predatory’ ⁶⁸ or ‘weak’ ⁶⁹ states are the norm. But analysis informed by the concept of security community allows for zooming in and out, or adjusting the optics, to look for security communities at a different scale than that of nation states and their clusters. Putting on this distinctive lens enables productively tracing trajectories of the rise and fall of such security communities.

The second proposal is seemingly orthodox: to focus on the Transatlantic area, the region where Deutsch and his collaborators made their initial historical investigations. The new research agenda should, however, be oriented towards the present, and partake and provide, in line with previous suggestions, conceptual clarity and empirical foundations missing in the current debate on the state of ‘security community’ in Europe or in the Atlantic area. Too often, arguments are obfuscated by using the concept in fetishising ways. The change should entail, first, empirically establishing whether such security community or communities indeed exist (for a case against the existence of a security community in Western Europe included, subversively, in Adler and Barnett’s volume, see Wæver’s contribution) ⁷⁰ and what situation they are in against the background of the reported rifts in Transatlantic relations and the Eurozone sovereign debt crisis. Second, it should include exploration of the processes of security community disintegration in other, historical cases to ascertain whether – provided that the security community exists in Europe – these might now be taking place. These processes currently deserve particular scholarly attention as, in the words of Thomas Risse, the ‘parochial vision’ of Europe seems to be gaining in strength as a result of Europeisation of the public spheres as communicative spaces, ⁷¹ threatening the continent with re-securitisation against the background of an ongoing risk of institutional fragmentation and demontage of the compensatory allocation system. ⁷² Such research should not be another in the line of securitising exercises that suggest the deepening integration as the only way to prevent repetition of Europe’s troubled past. After all, what the most recent crisis showed was that, concerning the fiscal compact, other choices did indeed exist besides further political integration and war. The new research agenda should instead be intended to broaden the horizons of political imagination, while seeking, in the words and the best tradition of Karl Deutsch, ‘greater competence and more compassion’. ⁷³