Existential security and the governance challenge: Confronting the antinomies of securitisation

Nathan Sears was a key voice in bringing the study of existential threats to humanity to the field of International Relations. His juxtaposition of national and existential security, and accounts of failures by major powers to adopt the latter frame, engage core questions of failures to securitize these potentially catastrophic threats and of great power (ir)responsibility. His work can also be read as a provocative entry point to a discussion of the governance challenge when existential security is at stake. It highlights the tensions between the necessity of mobilizing resources, building consensus, and creating legitimacy to address threats to humanity and doing so in a just and equitable way that acknowledges the unequal power dynamics in ‘speaking’ and ‘doing’ security. This debate potentially challenges some core premises of securitization theory. It also has enormous practical implications for addressing some of the gravest challenges of our time. Nathan Sears' concern for existential threats to humanity emerged from a deep sense of frustration and genuine puzzlement at the failure of the field of International Relations (IR) to take them seriously. He spent the early years of his dissertation project – partly captured in the article he published in this journal (Sears, 2020) – largely on making the case that a specific category of these threats caused by human agency ought to be a central concern of the field. This forum is already a testament to the inroads Nathan had made in persuading a new generation of scholars of the need for this work, even as his longer term ambition to develop new thinking and strategies to address these threats was tragically cut short. I appreciate the opportunity in this tribute to engage with his work to push forward that agenda. The irony was not lost on Nathan that his chosen field – motivated at its core by questions of security and survival – would be among the last to take such threats seriously. Philosophical and practical debates about existential threats to humanity are proliferating elsewhere in academic and popular writings, as others in this forum note. Similarly, Nathan found a ready audience of policy makers, being a winner in 2018 of Global Affairs Canada's International Policy Ideas Challenge competition. The following year he became Canada's Cadieux-Léger Fellow, a prestigious prize that Global Affairs used to place Nathan in its foresight division while he continued his PhD research. Meanwhile, real-world developments nearly unimaginable when Nathan began his dissertation put a series of exclamation points on the urgency of his research agenda: threats of nuclear war in Ukraine and fears of escalation; rapidly accelerating climate change and the litany of record heat and storms, wildfires, melting ice sheets, and warming oceans with their accompanying human impacts; and the unexpected leaps in advanced AI, that even a year ago Nathan's IR colleagues treated as highly speculative, now raising alarms even among its most prominent creators (Future of Life Institute, 2023; Kang, 2023). Nathan's realism also drove him to Securitisation Theory as the best way to bring these concerns to the IR field, but for the opposite reasons that motivated its original proponents (Buzan et al., 1998). Existential threats to humanity, he argued, demand a type of macrosecuritisation – a “higher order securitisation” – where the “referent object” is humankind and universalist claims about dangers to its survival legitimise extraordinary measures outside of normal politics and international relations (Sears, 2022; Buzan & Wæver, 2009, 259). His concern was failures to macrosecuritise, not the usual (often critical) focus on securitisation success (cf. Ruzicka, 2019). Great power consensus, he found, was the core condition for addressing these threats and macrosecuritisation was the means. Nathan's research also highlighted a dilemma that pushed securitisation theory outside its comfort zone of critical reflection on securitisations. It forced proponents to confront the uncomfortable possibility that some problems not inherently security related may be so potentially catastrophic as to justify or even require securitisation. But making such claims also necessitate engaging scholars who have highlighted the unequal power relationships among those who attempt to speak security, asking critical questions about the audiences thereby empowered, and assessing the consequence of securitisation when “humanity” is invoked as the referent object (e.g. Hansen, 2000; Sundaram et al., 2022). It is to these dilemmas I take up Nathan's work's invitation to engage. Hedley Bull, disenchanted with the Cold War superpowers' failures to sustain détente and act as “nuclear trustees” for humanity when they had the chance in the late 1970s, characterised them as the “Great Irresponsibles” (Bull, 1980). Sears' (2022) analysis of repeated failures of “macrosecuritisation” by the great powers echo Bull's assessment. Like Bull, Nathan, a committed Realist, held a deep belief in the moral responsibility of the powerful. His thesis rested, in a way, on claims to great power status being inextricably linked to their responsibility “to international society as a whole”, as Bull put it, and by extension to “existential” security when the threats to humanity demanded it. This idea of great power responsibility and interest in systemic stability or order is not, of course, limited to Realists. Nearly all branches of IR theory posit a responsibility among leading powers (who usually are the primary beneficiaries of the current system) to manage or support provision of collective goods and systemic security, and, to varying degrees, to bind themselves to rules and responsibilities to maintain order and legitimacy (Bukovansky et al., 2012; Ikenberry, 2017; Morgenthau, 1978, 24; Waltz, 1979: 194–210). Even constructivists who focus on how orders are negotiated, resisted and contested recognise that claiming recognition as great powers goes hand in hand with pursuing order-building goals and responsibilities (Loke, 2016). This commitment, however, embodies contradictions. On the one hand, Nathan's Realism led him to focus on the need to marshal the necessary material capabilities and extraordinary measures to address these threats to humanity and overcome the sources of great power rivalries that, at least in part, drove their destructive logic. He thus was particularly interested in great powers and their leaders as possessing the authority to “speak” security and “do” macrosecurity when they succeeded in replacing narratives of national securitisation with humanity securitisations. On the other hand, macrosecuritisations can mean empowering or imbuing with special rights or scope of legitimate action the very actors that scholars of global governance and critics of securitisation theory worry about. One worry is simply that there is little historical evidence that great powers can be relied upon to take their systemic responsibilities seriously, especially over time (e.g. Brown, 2004; Bull, 1980). If anything, recent scholarship has highlighted an equally common counter tendency to thwart attempts by other emerging or middle powers to promote or support more, or alternative, collectively responsible behaviour or institutional arrangements (e.g. Bernstein, 2020; Bukovansky et al., 2012; Gaskarth, 2017; Loke, 2016). Nathan and I had many discussions about the resulting limited variation on the “dependent variable” – macrosecuritisation almost always fails. More critically, successful macrosecuritisation may purport to speak for humanity but, in so doing, silence the actors or communities most likely to be suffering from insecurity. It may ignore their voices in the governance arrangements (Hansen, 2000; Sundaram et al., 2022) or reproduce power differentials, inequalities, racist or patriarchal structures, or even genocidal indifference. This worried Nathan in a way captured well in Belfield's contribution here. These antinomies and the dilemmas they posed are a key legacy of Nathan's work and highlight the stakes of his research agenda. On the first count, Nathan's arguments are compelling, and remain prescient. The empirical evidence that the structure of problems alone is neither self-evident nor sufficient to drive macrosecuritisation overwhelmingly supports the premise of securitisation theory that active construction is required. Indeed, the problematique of securitisation is usually understood as over, not under, construction and the destructive extraordinary practices of security and removal from normal politics that results. This default, however, greatly troubled Nathan. One of his great contributions was, in juxtaposing national and existential security, to shift attention of securitisation scholars to the implications of the securitisation frame. Macrosecuritisations may be appropriate and necessary, but anthropogenic existential threats are often not self-evident. Thus, when securitising actors “speak”, national securitisation frames often prove exceedingly difficult to overcome. The case of AI has, even since Nathan's death, become a textbook example. While the jury is still out on exactly what kind of threat AI poses, we are witnessing its national securitisation in exactly the way Nathan expected. He wrote that the US-AI strategy fundamentally aims to maintain “supremacy” (Sears, 2022, 197–203), quoting the 2019 Artificial Intelligence Strategy's explicit goal to recognise the technology as a means to “ensure an enduring competitive military advantage against those who threaten our security and safety” (US Department of Defense, 2019, 5). While the US military's embrace of AI is premised on great power technology competition and the threat to national security, the UK and European strategies seem driven by fears of relative decline while China has developed an ambitious strategy to be the world leader in AI for a variety of national and strategic purposes (Sears, 2022, 200). This contrasts with industry leaders' own warnings to lawmakers. As Sam Altman, chief executive of OpenAI, told a US Congressional hearing recently: “I think if this technology goes wrong, it can go quite wrong. And we want to be vocal about that… We want to work with the government to prevent that from happening” (Kang, 2023). In lieu of macrosecuritisation, seven major US-based companies have entered into a voluntary agreement that might be characterised as reflecting multiple framings of security: human (e.g. safeguards against threats to privacy, democracy or rights), national and existential. Most analysts agree these measures will have little practical effect (Shear et al., 2023). The accompanying statement from the White House notes its consultations on “voluntary commitments” with almost two dozen of its allies as well as support for several other international initiatives on AI safety and shared governance principles (White House, 2023). Yet, for all the corporate and government actors “speaking” macrosecurity, the only follow through with teeth is an executive order that nationally securitises the technology using extraordinary measures that will inevitably ramp up competition with China; it bars US citizens and companies from investments and other financial transactions with Chinese companies working on AI, quantum computing and other related technologies, including semiconductors. Meanwhile, both countries are continuing their major investments in national AI development strategies (The Guardian, 2023). Britain and the EU plan similar restrictions. This and other cases examined in Nathan's thesis – control of atomic energy and biological weapons – mark the first systematic responses to calls for a research programme on failed securitisation but go well beyond helping to understand conditions under which audiences may not accept securitisation moves. Nathan simply would not accept the view by even those calling for this research programme that some things – nuclear weapons, for example – are simply not “amenable to securitisation” (Ruzicka, 2019, 373). Nathan's concern with macrosecuritisation is clearly catching on, most prominently in the debate over climate change. Here, even one of the originators of securitisation theory, Ole Wæver, who has spent much of his career making the case that “ideally things should be dealt with on their own terms”, not securitised, is now making the case for macrosecuritisation of climate change as necessary because of the threat to humanity, even though he calls it a “dark hope” (Wæver, 2023). Nathan understood that identifying conditions of macrosecuritisation failure among great powers only constituted a first step, and his dissertation examined the rich empirical history of the complex governance responses that occurred in his cases. However, his measurement of “success” only against the metric of extraordinary collective measures led to some of our most intense exchanges over the appropriateness and consequences of alternative governance arrangements. I share with Belfield a sense that a logical extension of Nathan's research agenda is to reflect further on the complex and unpredictable ways governance has evolved even when partial macrosecuritisations occurred, and to critically consider the implications of alternative macrosecuritisations and governance futures. We especially debated climate change, which has since the late 1980s seen multiple securitisation attempts with vastly shifting narratives (Dalby, 2020). That history more often produced narratives of protection against climate change's consequences – as a source of acute conflict, migrations and other displacements, scarcity – than mobilising societal transformations required for just decarbonisation to prevent it or other means of responsibility for mitigating or remedying its harms. As the warnings and experience of climate change has shifted, so too have discursive moves towards securitisation more commonly invoking existential security narratives from a wide range of actors across positionalities and the political spectrum. However, here the tensions already noted in who has legitimacy to “speak” for humanity and critical questions about the consequences of securitisation moves could not be more evident. The proliferation in practice of “emergency” frames in climate change discourse illustrates this tension in a way that mirrors the more general logic of securitisation moves. Who gets to “speak” security is often the very issue at stake, even when emergency narratives claim to be in the service of “humanity” and survival. As Patterson et al. (2021, 841) write: Emergency frames may convey the urgency necessary to avoid further destruction of the environment … thereby opening up new political possibilities for consensual sustainability action … [They] may also resonate with longer term patterns of inclusion/exclusion. Emergency frames demanding sweeping action by the government could afford political coverage for unpopular actions … Yet, poor and/or minority communities may lose voice if emergency frames are oriented towards quick fixes. Specifically, the “climate emergency” frame may “empower a technocratic elite and undermine a plurality of goals and political creativity necessary to address justice and well-being concerns bound up in sustainability. Deploying emergency-as-strategy could normalise a pre-emptive logic that could close down debate and legitimise otherwise unpalatable options” (Patterson et al., 2021, 845). This is not an abstract debate. Concerns that macrosecuritisation will default to planetary solar radiation management (a form of geoengineering) are playing out in real time. They reflect concern about solar geoengineering's uncertain, unequal or possibly self-undermining effects. But even more so, they attack the legitimacy of great powers – or worse, the US as the world's current leader in solar geoengineering acting unilaterally – to speak for humanity absent new governance arrangements that confront “problematic structures that may currently exist in some institutions and regions” (Táíwò & Talati, 2022: 13). These debates are complex: some worry explicitly about increased “risks of militarisation or securitisation” (Stephens et al., 2021) while others worry more about incorporating relevant knowledge and voices, for example, from the global south, and more generally inclusive collective decision making (Táíwò & Talati, 2022). These are not debates “for” or “against” securitisation – those arguing for a plurality of voices assiduously avoid speaking securitisation or desecuritisation for others – but about the power dynamics often undertheorised in such moves. Despite the world of national securitisation he saw growing around him, Nathan remained hopeful that great powers would see taking responsibility for addressing existential threats to humankind as an opportunity to increase their own status and reputation in the process. “Great power rivalries”, he wrote in the closing sentences of his dissertation, could “work for and not against the security and survival of humankind”. Perhaps, an alternative might offer another way forward: “repoliticisation” of macrosecuritisations that brings the debate back into the political arena, which might both motivate responsibility of the kind Nathan craved and create or force space for a plurality of voices to speak security. Nathan might have appreciated this kind of provocation to securitisation theory that offers a middle, and I would argue necessary, if uncertain, way forward (Inoue, 2018; Kramarz et al., 2021; Paterson et al., 2021; Sundaram et al., 2022). That we have lost Nathan's voice in this important debate is a great loss to all. None. None. Steven Bernstein is Distinguished Professor of Global Environmental and Sustainability Governance at University of Toronto where he also co-directs the Environmental Governance Lab. His research interests include global governance and institutions, legitimacy and responsibility in global politics, and policy studies. He was Nathan Sears' thesis supervisor.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.