Is Securitisation a Natural and Useful Response to Existential Threats? Introducing the Idea of Peacification

Abstract

The need to decide whether to securitise an issue area that poses an existential threat, and then treat it as a security matter, is often regarded as a choice without positive alternatives. This article introduces an alternative framing: the “peacification” of issue areas that pose existential threats. It also demonstrates that there is variation in the levels of security and peace framing in authoritative speech. By measuring these levels in authoritative US presidential papers and comparing them with levels of success in the US efforts to reduce organised violence, the article falsifies the assumption that securitisation is the only realistic way to frame existential threats. Correlative evidence fails to support the view that the securitisation of organised violence is useful. This finding is highly significant for challenging the naturalness of mainstream security framing, which, due to the very grammar of “security,” is premised on political othering. After all, security is a property of an agent, while security issues traditionally arise from threats posed by another agent. By disaggregating the elements of these two framings, this article offers plausible explanations as to why, in general, peacification may be a better framing for the prevention of organised violence than securitization.

1. Introduction

The concept of security demands the definition of a referent object, while the traditional understanding of security also requires the identification of an intentional agent that poses an existential threat to that object. In this framing, the safety of individual countries, regions, alliances, or groups is threatened by geopolitical challengers, terrorists, or oppressive autocrats. The ability to deter, contain, stop, or destroy such dangerous actors is therefore seen as a means to ensure safety from violence for the actors in focus. Existential threats justify such measures, and it is thus considered natural for authorities to speak (inject via speech) issues of organised violence into the realm of security to justify extraordinary measures to repel such threats.

While securitisation theory generally views securitisation as something negative, it assumes that the choice to securitise or not is natural and without alternatives. Although the theory often criticises the move from a threat (or even an unfounded existential threat) to securitisation (see, for example, Buzan 2006), it does not offer an alternative. For instance, it does not suggest that it is possible to speak an issue area that constitutes an existential threat into the realm of peace/conflict rather than security/insecurity. Furthermore, while the theory reveals how an issue area is spoken into the security realm, it does not challenge the naturality of the way this realm is constituted—specifically, how defining something as a security issue justifies power-centric and secretive, rather than dialogical, approaches. Instead, the theory implicitly accepts the naturalness of securitisation as a response to existential threats. The very word “security” already prescribes responses to securitised issue areas; if security is the objective, we must define whose security is being discussed, and often also identify the agent causing the existential threat. It is precisely this naturalisation of the move from existential threat to an agent-centric security realm that needs to be questioned.

Instead of framing organised violence as an agent-focused security matter, we might consider viewing the referent object and the threatening agent as parts of a relationship that causes both parties to act in threatening ways. In this perspective, the security framing adopted by both parties could constitute the threatening relationship, as each side perceives the other as the threat whose behaviour must be changed by resorting to security measures. The crucial issue, then, is the jump from “existential threat” to efforts at “security”—a jump that must be problematised if we seek alternative ways to address existential threats.

To address this, the article introduces an alternative framing that may be more effective in reducing atrocious violence and its associated threats. Instead of “securitising” issue areas generally linked with organised violence, the article proposes the “peacification” of such issue areas. The existence of this alternative challenges the naturalisation of security framing in matters of organised violence. To demonstrate that this alternative is not merely a naïve idea unsuited to the real world, the article provides correlative evidence that undermines the assumption that security framing is associated with containing organised violence. On the contrary, higher levels of security framing are associated with higher levels of violence, while higher levels of peace framing—and a higher ratio of peace to security framings—are both correlatively associated with a reduction in violence.

For the empirical basis of the theoretical argument, the article focuses on US military operations in the post-Cold War era. The correlative evidence comes from a comparison of textual indicators of peacifying and securitising speech by the US president from 1993 to 2014, using data from the “Public Papers of the Presidents of the United States” (Clinton et al. 1993–2014) and statistics on fatalities of organised violence during the same period, drawn from UCDP data (Davies et al. 2024). The focus on US operations is motivated by the interest in understanding the impact of framings on violence while minimising the influence of intervening variables. Only US operations can seriously determine the outcomes of violence; focusing on smaller actors could yield more random results, as their influence may be marginal.

The focus is thus on the framing of a single major actor, even though the theoretical argument suggests that a relational framing produces better social realities for conflict prevention. The reason for focusing on the framing of one actor is to avoid accusations of tautological reasoning. If we assumed a relational perspective and argued that changing one’s construction of the reality and morality of conflict acts would automatically lead the opponent to adopt the same construction, this argument could be dismissed as circular. While adopting a common perspective undoubtedly facilitates the discovery of shared ground, proving this would be an analytical rather than an empirical task. It is far more challenging to demonstrate that a framing by a major power alone can have a positive effect on the level of violence, even when that power’s framing does not dictate the framings of others. If the alternative framing of a strong actor alone positively affects the level of violence, the argument that such a framing is unrealistic or naïve in an imperfect world can no longer hold.

The correlative relationships presented in this article do not prove causality, and designing a research strategy to test all alternative causal paths is beyond the scope of this study. Instead, the article introduces and operationalises the new concept of “peacification”, and provides plausible explanations for the problems inherent in security framing, suggesting why peace framing may be better suited to addressing existential threats related to organised violence. The strong correlative evidence linking high levels of security framing with high levels of fatalities from organised violence raises serious doubts about the naturalness of securitised framing. However, this evidence should not be treated as a definitive test of the theory of “peacification”. Testing this theoretical idea will be the next step in theory-building.

2. Theoretical Framework and Argument

According to a constructivist understanding of conflict, the way in which safety and violence are framed by one, some, or all actors in a conflict situation and its context creates social realities that may be peaceful or belligerent (Onuf 1989; Wendt 1998). Such social realities may then lead to more, fewer, or no fatalities of organised violence. The premise of this article is that, since the reality we create in our interpretation is our own creation, we should consider the usefulness of the social realities our interpretations bring about. We need to question interpretations and social constructs that are presented to us as natural and, instead, assess whether we find such naturalised social constructs useful and whether the social realities they produce are beneficial to us (Kivimäki 2016).

Starting from the normative premises of peace research, this article focuses on social constructs by evaluating their conduciveness to the prevention of organised violence. On the one hand, this article introduces an alternative construction of the reality of organised violence and demonstrates that the naturalised framing of organised violence varies in practical politics. On the other hand, it explores the social realities constituted by securitised and alternative framings of organised violence. This examination focuses on how the two different social realities enable conflict prevention, i.e., the reduction in fatalities by organised violence. If securitised framing is “realistic” and a peacified framing is naïve and idealistic, this will be demonstrated by the better performance of securitised framing in US military operations that attempt to contain violence and reduce fatalities of organised violence. This elaboration aims to present an argument for the plausibility of the claim that we should consider organised violence not as a security problem, but as a peace problem.

This article identifies a “security framing”, a result of a securitizing speech act, in which attention is focused on actors (actors whose security we prioritise, or actors that can threaten the security of those actors), and a “peace framing”, in which attention is focused on relationships. While security thinking can be soft and human-oriented, as in human security or environmental security, a perspective that focuses on security is, by its very grammar, focused on someone—an agent—whose security is of concern. Security of an individual from traditional threats, such as wars, or non-intentional threats, such as environmental catastrophes, is nevertheless non-relational security. At the same time peace framing is inherently relational, i.e., between agents, not something whose reference object is just one agent. Non-traditional, non-intentional existential threats are naturally outside the focus of this article, given that we cannot have peace with the environment or with economic conditions. Peace framing cannot be applied to issue areas of environmental security or threats from pandemics.

Peace is grammatically different, as no single agent can have peace alone; peace is always a relational quality, existing between agents. In peace framing, agency is acknowledged only as part of a relationship, sometimes even understood as being constituted by positions and relationships among multiple agents (Jackson and Sterling-Folker 2013; Qin 2018). If there is a threat of organised violence to state A from state B, a focus on relationships would require dialogue to prevent the worsening of the relationship to the level of war, as well as improving positive interdependence between people and businesses in these countries. While in a security framing, our security may depend on the characteristics of the Afghan state, Iran, or Israel, in a peace framing, Afghanistan, Israel, and Iran always act toward us in ways that depend on their relationship with us, not primarily on their individual characteristics. Galtung, Buzan and Hansen view these two framings as oppositional (Galtung 1969; Buzan and Hansen 2009).

In examining how framings of safety create social realities, this article aligns with Ole Wæver’s ideas on how securitising speech acts can constitute realities in which an issue area moves into the realm of security (Wæver 1995; Wæver 1998; Acharya 2006; Balzacq et al. 2016; Coen 2017; Heck and Schlag 2013; Kapur and Mabon 2018). While referring to an issue area related to existential threats to states, groups, and individuals as a security issue is known as securitisation, this article proposes an alternative, relational way of addressing existential threats, and calls it peacification.

In securitisation theory, declaring something an existential threat is assumed to necessitate the selfish, partisan approach of security practices. The theory assumes the move from “threat” to policies of “security”, rather than opening the option of moving from “existential threat” to policies of “peace”. After all, if one’s very existence is threatened, it is believed that one must defend oneself and one’s security (Wæver 1995). Security framing emphasises the agency of the one who is to be secured and, in traditional security issues, the one that is threatening. Whenever the threat is intentional (i.e., in traditional security threats), there is a need to change the calculus of the threatening agent and to adopt power-centric actions. There is also a need to stand firm in defence of the values we cherish: security is not something achieved through surrender, but rather, the protection of our values is intrinsic to our security (Acharya 2006). Thus, speech that weakens our country’s resolve to defend our values is dangerous for security.

Yet, willingness to talk about compromises may be necessary for peace. Moving from an existential threat to peace would, therefore, require different policies.

To sum up our theoretical argument, we present two framings that rely on two pathways for dealing with the emergency created by existential threats. Both framings lead to solutions to existential threats that can be effective, and thus neither should be seen as a natural framing. The effective causal paths from existential threat through security and peace framing to the mitigation of the threat are as follows.

When leaders securitise the existential threat that specific cases of organised violence somewhere constitute, the security realm deals with the threat by focusing on our side as the agent being threatened by the actions of the threatening agent. Since the focus is on agents as subjects and objects of a threat, it is natural to consider strategies that change the threatening agent’s behaviour, i.e., strategies based on the use of power. Since the focus is not on the relationship and the threatening behaviour of the other is interpreted within an agent-centric focus, we see the actions of the other as emanating from the characteristics of the other. Thus, there is a need to stand firm in the defence of political values and safety, resisting the other agent. For that, there is no need for dialogue with the threatening agent. When the threat emanates from a threatening agent, the security framing often—but not necessarily—demonises this agent, assuming it to be opportunistic rather than driven by a commitment to an understanding of fairness. As a result, security framing often leads to interpretations in which we not only need to safeguard our safety from the violence of these unreasonable, predatory agents, but also defend our political values against such agents. Thus, security practice involves not just the safeguarding of safety, but also the defence of political values and the enforcement of justice.

On the other hand, if a threat is framed as a peace/conflict issue, we focus on something that our action alone cannot fix. Since relationships always involve at least two agents, the control of the relationship involves at least two agents. Thus, dialogue and cooperation between those agents are necessary to address the threats emerging from the relationship. When the focus is on escalatory relationships that constitute an existential threat, this cooperation requires agreements that restrain actions threatening others and, in turn, provoke behaviours that threaten us. Instead of using power, a focus on relationships prescribes methods that restrain (provocative, threatening) power and methods that are, at the very least, neutral to power, i.e., they do not increase power (for example, rescuing people by offering asylum, etc.). The plausibility of this being an effective strategy for dealing with existential threats can be linked to the model of the security dilemma (Hertz 1950; Booth and Wheeler 2008), in which a focus on agent-centric individual security and the lack of relational focus leads automatically to socially suboptimal solutions. In addition to prescribing restraint, a focus on relationships rather than demonised agency suggests that disputes about values are real, rather than the opponent being unreasonable and us needing to enforce justice and our superior political values. Fixing relationships requires a willingness to compromise (Zartman 2019).

It is possible to present, as shown in

Table 1. Concepts of security framing and peace framing.

	Security Framing	Peace Framing
Focus of attention	Agent-Centric: Security of our side against the threatening agent. Our security is dissociated from the security of the threatening agent.	Relational: Peace between us and others. Our safety is associated with the safety of the other. Solutions are sought to fixing relationships that cause threats to the safety of both/all potential conflicting parties.
Method of safety	Power-Centric: Since the focus is on agents, mitigating threat requires change in the behaviour of the threatening agent; the method of safety is the use of power.	Power Negative/Neutral: Since the focus is on a relationships in which threatening moves of one agent provoke the other agent into threatening countermoves, the method of safety is based on moves and agreements that restraint provocative moves.
Relationship between safety and political values agent defends	Associated: Security means both safety of physical existence and the safety of important political values. Values need to be enforced on those that threaten our security to change their behaviour from threatening to benevolent.	Dissociated: Political values and measures to guarantee mutual safety from violence must be negotiated. Compromises are needed.

, three main differences between the security and peace framings, identifying the focus of analysis, the role of force, and the role of power.

Since the difference between security and peace framing is based on these three elements, we will first define them and integrate them into the existing literature. After that, we will then operationalise the three differences so that levels of securitisation and peacification can be measured in authoritative texts by the US president. The measurement and the elaboration of the effectiveness of each framing are based on the peacefulness of the realities the two different framings constitute. The intention is not to assess the moral fairness of such realities, as the investigation here is limited to the level of violence. This must be noted as a limitation of this study—it would be possible that one of the two framings produces fair and just terms of peace, while the other does not, and thus, the effectiveness of a framing in the reduction in fatalities of organised violence only refers to one of the objectives of security/peace practice, while leaving us with ignorance of how fair and just the peace that different framings produce is.

Furthermore, while revealing that peace framing is not unrealistic or naïve, we naturally do not assume that this framing would be optimal for each case of existential threat; it would be possible to imagine a case where violence is caused by one bad agent that cannot be negotiated with and that simply needs to be contained by means of military power. Instead, by looking at the track records of the two framings, we investigate whether authorities should securitise organised violence more or less, as track records are expected not to be absolute, but probabilistic.

3. Existing Literature on Agent-Centricity, Power-Centricity and Defence of Unyielding Political Values

3.1. Focus on Relations vs. on Agents

To identify the first difference between peace and security framing in the existing literature on conflicts, we must draw on the debate regarding agent-centric and relation-centric or relational understandings. Security studies concentrate on security, which is something possessed by agents. To identify security, it is grammatically necessary to identify the referent object of security. The critical debate in security studies of the 1990s and the 2000s was primarily focusing on the critique of the narrow focus in traditional approaches on the state as the referent object (Booth 2007), while the securitisation theory also warned against the securitisation of broader issue areas as it had militarising effects on those issue areas. This article suggests, however, that we should not focus on referent objects of security at all, but rather focus on relationships.

Relationships always exist between agents. The distinction between agent-centric and relational perspectives in our comparison of security and peace framing borrows from sociology, which considers the distinction between substantialism and relationalism as fundamental (Elias 1991; Emirbayer 1997; Jackson and Nexon 1999). According to relationalist sociology, we cannot focus on entities as ontologically independent of the relationships in which they are embedded: “The person, let alone aggregate entities such as the state, cannot be thought of as ontologically distinct from the social relations in which they are embedded” (Elias 1991, p. 25). The constructivist idea of mutual constitution (and thus the lack of independence of substances from their relationships) suggests considering agents at least partly as products of their relationships and interactions with others (Wendt 1987). This notion, in Peace Studies, indicates that our actions to address threats can fundamentally change the other actors in our security environment; therefore, if we do not consider these relationships when dealing with existential threats, we may create enemies.

Peace Studies has a mission that aims to promote peace and avoid conflict between actors, rather than seeking security for partisan actors (Galtung 1969). A peace perspective with a relational focus may go so far as to claim that agency is constituted by its relations with other agents. “The focus is not fixed on any particular individual but on the particular nature of relations [peaceful or belligerent] between individuals who interact with each other. The focus is on the relationship” (King 1985, p. 16). According to Yaqing Qin, relational framing considers the international system as “composed of ongoing relations, assumes international actors and actors-in-relations, and takes processes defined in terms of relations in motion as ontologically significant. It puts forward the logic of relationality, arguing that actors base their actions on relations in the first place” (Qin 2016, p. 33).

The lack of focus on such interactive relational processes is often absent in mainstream Western IR theory, which centres on calculative individual rationality (Wendt 1998). According to relational sociology, this may be due to the fact that Western European languages tend to exhibit a bias towards substantialist thinking (Elias 1970, pp. 111–12; Emirbayer 1997, p. 283; Jackson and Nexon 1999, p. 293).

In Security Studies, however, the focus on relationships is often considered naïve. Making peace with the enemy is difficult if we perceive the conflict situation as one where violence is purely due to the characteristics of the enemy—characteristics that are independent of the enemy’s relationship with us and others. Kaldor cites Major General Raymond T. Odierno, Commander of the US Army’s 4th Infantry Division, who claimed that 70–80% of captured insurgents in Iraq were ordinary criminals (Kaldor 2012, p. 165). Thus, resolving the security problem cannot be achieved by working on the relationship, let alone by addressing the grievances of both sides. As a result, Kaldor argues that “it may be impossible to obtain consent from both the local population and the warring parties. If an agreement has to be negotiated with a war criminal, then the credibility of the operation in the eyes of the local population may be damaged” (Kaldor 2012, p. 135). Thus, “what is needed is not peacekeeping [restraint, peace negotiation, or mediation] but the enforcement of cosmopolitan norms, i.e., the enforcement of international humanitarian and human rights law” (Kaldor 2012, pp. 124–25). Kaldor supports the Weinberger/Powell doctrine of overwhelming force in the production of security. Consequently, she naturally rejects one of the basic principles of United Nations peacekeeping, according to which operations must avoid the use of force except when necessary for self-protection (Kaldor 2012, pp. 128–29). Such law enforcement cannot be carried out “neutrally, as actors and illegitimate representatives of groups have to be enforced as it is breaking the law” (Kaldor 2012, p. 128).

3.2. Method of Self-Restraint vs. Method of Use of Power

If traditional existential threats are perceived agent-centrically, and thus the threat emanates from the behaviour of the threatening agent, it is natural that the solution would involve the use of power to change the behaviour of the threatening agent. Power, as the ability to change someone else’s behaviour (Dahl 1957, pp. 210–14), has been regarded as a discipline-defining concept in Political Science and Security Studies. While Morgenthau saw power as an end in itself, and argued that when states sometimes acted in the absence of the motive for power such actions were not, by definition, “of a political nature,” Waltz merely considered power as the principal means in political action (Morgenthau 2006, p. 27; Waltz 1988, pp. 615–16). Again, the peace framing radically challenges this naturalised conception of Security Studies—if relationships are in focus, it is logical to consider obsession for power or powerful means of dealing with existential threats. Rather than focusing on power as means, care for relationships requires self-restraint towards powerful acts that may threaten others, thereby provoking countermeasures that further escalate insecurity in the relationship.

When Syrians need to be protected by deterring, punishing, or preventing President Assad’s authoritarian violence, we are referring to power-centric protection. The idea of bringing order into the chaos of international relations by means of hegemonic leadership (Webb and Krasner 1989; Joseph 2002), humanitarian interventions (Bellamy 2015; Hehir 2010a), or overwhelming, norm-enforcing power (Kaldor 2012, p. 137) offers examples of security framing that uses power-centric methods of security.

The method is considered non-power-centric if the focus is on one’s own protective actions (power-neutral). For example, if Syrians are protected by granting asylum to Syrian refugees in our safe countries, this constitutes power-neutral protection. If the intention of the action is self-restraint or the establishment of regimes of mutual restraint (power-negative), then agreeing to end sanctions if Syria destroys its chemical weapons is a form of protection that reduces mutual power, and is therefore power-negative. Other examples of this approach include agreements on arms control and disarmament, which, instead of maximizing power, aim to reduce mutual power to destroy. Peace enforcement, as a practice, aims to change other agents’ behaviour, and is thus not only power-centric but also agent-focused. As such, it represents a practice that frames conflict prevention in terms of security rather than peace, despite the presence of the word “peace” in the name of the practice.

There are analyses of the benefits of mutual restraint and the non-use of power, despite the tradition of power-centricity in the discipline. Mearsheimer, Jackson, and Nexon derive their analysis of the benefits of restraining oneself from provocative security behaviours—behaviours that necessarily provoke reactions from potential enemies—from a clearly relational focus of analysis (Jackson and Nexon 1999; Mearsheimer 2019).

Muscular approaches, such as peace enforcement, have been criticized, and the idea of considering the prevention of violence as necessarily muscular has been especially criticized by feminist theories of international relations (Tickner 1992, p. 63; Enloe 2007, p. 40). There, in addition to the emancipatory aim of denaturalizing hegemonic masculine concepts of security, the assumption is that power-centric approaches are also detrimental to peace. Yet, despite decades of feminist critique of power-centricity, the first feminist empirical exploration of the relationship with power-centricity is rather recent (Kivimäki 2024).

3.3. Solutions Aiming at Compromises vs. Solutions Aiming at the Defence of Political Values

Security is not solely about physical safety. It is not what George Orwell mockingly suggested when he said that “the quickest way to end a war is to lose it” (Orwell 1946). Lippmann elegantly encapsulates the association between safety from war and political values in his definition of national security: “A nation is secure to the extent to which it is not in danger of having to sacrifice core values if it wishes to avoid war and is able, if challenged, to maintain them by victory in such a war” (Lippmann 1943, p. 51). While conceptually linking peace with terms of peace may be beneficial for bargaining leverage, such conceptual practices enable securitisation to permeate political values that serve as the conditions for non-violence against securitised organised violence. Security and safety become possible only if we also maintain the territory or governance system we demand in the dispute as a condition for our non-violence. Conceptually linking safety with political values provides very few opportunities for compromise, whereas avoiding instrumental violence—violence in war aimed at imposing one’s will on the other—requires compromises. The security framing of one side often prompts a similar framing by the opposing side. If both sides fail to recognize the conceptual distinction between ensuring safety from violence and refusing to compromise on political demands, such rigidity unavoidably leads to the onset or perpetuation of conflict. When conceptualization fails to help conflicting parties understand the difference between existential safety and the safety of their demands, war ultimately imparts this lesson upon them.

Once again, peace framing offers a radical challenge to the association of safety with political values. From a relational perspective of peace framing, it is unnecessary to associate political values with safety. Instead, it is preferable and acceptable to seek ways to facilitate compromises in political values to transition from violence to peace (Friedman 1990; Kriesberg 1998; Mitchell and Banks 1996). This does not imply a prescription for surrender. Since the peace perspective is relational rather than agent-centric, making compromises becomes an interactional task aimed at reconciling the core interests of both parties involved (Kelman 1972, 2008). Thus, rather than stubbornly defending one’s interests and dismissing the other’s as unreasonable, peace framing assumes that durable peace requires the basic needs (Mitchell 1990; Burton 1993) and core security interests (Galtung 1969; Azar 1990) of all actors in the relationship to be met. Political values remain important; however, (a) they are distinct from our safety, and (b) making our political values compatible with those of others is essential for social rationality. If a conflicting party adopts a peace framing that distinguishes between existential safety and political demands, it may invite its opponents to discuss the latter (demands) to ensure the former (safety) for both sides. Such an invitation also introduces the peace framing to the opponent: when political demands become a subject of dialogue to safeguard existential safety—i.e., when demands are negotiated to preserve existential safety—the practice fosters a separation, even in the opponent’s mind, between matters of safety and matters of demand. This creates space for conflict resolution, which security framing fails to achieve.

The association between safety and political demands or values that we consider as terms of our non-violence within security framing constitutes a social reality in which a willingness to compromise and sensitivity to the suffering of war are perceived as security threats. This is demonstrated in the fact that softness and even opinions that come too close to opponent’s demands in conflict are considered security threats. People and parties that represent such opinions or who interact in dialogue with the enemy about the political values that are at stake in the dispute, let alone those who care about the enemy’s suffering as a result of our side’s violence, are “in enemy’s pocket”, and they may unwillingly be the enemy’s “useful idiots”, “influence agents” or “assets”, meaning such people can be dealt with using security measures. This creates the starkest contradiction between actions to deal with organised violence in security and peace framing: many peacemakers are prosecuted as security threats. Chelsea Manning, Edward Snowden, and Julian Assange—frequent nominees for the Nobel Peace Prize—have all been found guilty of security-related crimes. Yet, from the perspective of how to avoid violence, people who demonstrate the possibilities of compromise, while also revealing the suffering in war, are necessary for compromises and, thus, for the reduction in organised violence.

4. Operationalising Securitising and Peacifying Speech

Our main argument is that there is an alternative to securitisation in the face of existential threats, and that this alternative—peacification—can sometimes serve as a better framing for reducing violence than securitisation. To support this claim, we must demonstrate that the levels of security and peace framing vary in authoritative speech. To do so, we will need to operationalise peace and security framing in authoritative speech. This will show that peacification is not merely a theoretical concept, but an existing opportunity. To argue that peace framing may sometimes be better for reducing fatalities from organised violence, we will also need to design a strategy to measure levels of organised violence in the context of a comparison between discourse and outcomes. In the next chapter, we will operationalise the two framings, and in the following chapter, we will operationalise the level of organised violence. If the exploration—using the operationalisation of authoritative texts and the comparison of levels of securitisation and peacification in such texts with the levels of fatalities from organised violence—reveals no evidence that peace framing, as a naïve, unrealistic, or idealistic approach, systematically produces poor results or increases organised violence by creating opportunities for bad actors, then we can consider abandoning peacification and treating securitisation as the “natural” response to existential threats.

The method of comparing framing and its ability to reduce organised violence assumes that peace and security framings coexist to varying degrees in US presidential utterances. Therefore, it is necessary to create a scale for peace and security framings and study their temporal variations (as well as their variations between countries when considering power-centricity). Additionally, a framing may not be entirely peace-oriented or entirely security-focused, but rather a mix of both to some extent. Hence, it is possible to measure both the level of peace or security framing and the ratio between the two framings.

When measuring the monthly levels of peace and security framings, it is important to avoid conflating the prevalence of a framing with measurements of how often the president speaks. For instance, if, in January, the US president utters only six sentences, all related to security, and is otherwise silent due to extensive involvement in security operations, but in February speaks verry frequently, primarily about peace efforts but sometimes also about security, it might erroneously appear that the frequency of security-oriented utterances is higher in February. In reality, the president’s sole focus in January was on security, while in February, though the president discussed peace, the greater volume of speech led to more security-related words being used. To avoid this pitfall, the frequency of each indicator of a framing is divided by the total number of monthly codings.

We begin by measuring the relational or agent-centric focus in framing existential threats, using the words “peace” and “security” as proxies for the two framings.1 While a focus on good relationships can be expressed in various ways, a relational framing in matters of organised violence must employ the word “peace”, which inherently refers to a relationship. Similarly, the word “security” must be uttered if organised violence is talked of in an agent-centric manner. Both words refer to a positive state of affairs, signifying low or no fatalities from organised violence. Thus, the selection of relational and agent-centric words is balanced, and neither word is, in its meaning, more likely to refer to escalating violence. Therefore, the relative frequency and ratio of these two words indicate the extent to which the US president frames international politics in relational or agent-centric terms.

Our analysis of power-centricity focuses on sentences containing the verb “protect” and its derivatives. While the words “peace” and “security” are automatically associated with the two framings, power-centricity cannot be studied without defining the context in which power is used. Firstly, power can serve purposes other than protection from existential threats—for example, we speak of presidential powers. Therefore, we need the specific context of protection, which is defined by focusing exclusively on sentences containing the word “protect” and analysing the methods associated with this verb. Moreover, the data used for this analysis (based on definitions in Kivimäki 2019) are limited to sentences about protection related to international security affairs, specifically the protection of people. This excludes sentences where protection refers, for example, to the protection of an argument or a legal interpretation.

A sentence is considered power-centric if the means of protection involve preventing or deterring the actions of the agent that harms or threatens the people to be protected, or rewarding the threatening agent for refraining from such harm. It is also power-centric if the method involves weakening, incapacitating, or destroying the threatening agent. Conversely, protection is not power-centric if it is achieved through actions that do not affect the threatening agent (a power-neutral strategy, such as offering asylum) or through agreements that restrain both parties—ourselves and the threatening agent—by reducing the range of options or the power of both (a power-negative strategy, such as arms control agreements). Protection is also not power-centric if it focuses on agentless threats, such as environmental catastrophes.

Power-centricity is then measured both by counting monthly power-centric sentences with the word protection divided by all sentences with that word, and by using the ratio of non-power-centric protection sentences to power-centric protection sentences.

To measure the level of the stubborn defence of values versus the willingness to compromise, we focus on words that reveal how political values are perceived. If political values are considered natural or universally agreed upon, they must be “enforced” rather than negotiated. Failure to react to violations of these values is seen as undermining a value system regarded as common and universal, necessitating defence rather than negotiation (Hehir 2010b, pp. 218–19; Bellamy 2005). In this narrative, rogue actors must be held accountable, and norms of non-violence enforced upon them.

However, vocabulary indicative of the security framing, where the enforcement of norms is essential and compromise is unnecessary, presents challenges. The analysis of US texts reveals that such unyieldingness with regard to political values is based, in the securitizing speech, on the idea of considering alternative political values as unreasonable and morally non-negotiable. Often, violence in such contexts is viewed simply as between innocent people and predators. On the other hand, unyieldingness in security framing is based on the function of the military to heroically defend the “rule-based international order” or the rights of these innocent people. Consequently, it would be possible to measure the righteous, moralistic side of the defence of values by measuring the relative frequency of the word “innocent”, while measuring the heroic militaristic side of unyieldingness by measuring the relative frequency of the word “military”.

Thus, a high frequency of the word “military” indicates security framing, while a low frequency suggests a more peace-focused framing. Here, it was not possible to find good terms that would indicate a peace framing, and thus the relative lack of indications of association between existential safety and political demands must suffice as an indication of dissociation between the two. This is not optimal, but must suffice for this elaboration, which establishes the plausible pathways from the two framings to the increase or decrease in organised violence.

The three different aspects of peacified or securitised approaches to existential threats –relationality/agent-centricity, the role of power, and the willingness to compromise/defending political values—form the elements of two coherent framings. However, it is possible to conceive of peace or security framings that strongly emphasise only one or two of these elements while disregarding the others. Thus, we do not need to expect a perfect association between the different elements of the same framing.

Table 2. Operationalisation of security framing and peace framing.

	Peace Framing	Security Framing
Relationality vs. Agent-Centricity
Relative frequency of “Peace” (stemmed)	x
Relative frequency of “Security” (stemmed)		x
Frequency of “peace” divided by “security”	x
Non-Power-Centricity vs. Power-Centricity
Non-power-centric utterances of protection divided by power-centric utterances	x
Association vs. Dissociation Between Safety and Political Values/Demands
Relative frequency of “Innocent” (stemmed)		x
Relative frequency of “Military” (stemmed)		x

summarises the assumptions about the words that belong to each framing in our elaboration.

We conduct computer-assisted textual analysis using the NVivo R 1.6 package to identify the monthly levels of security and peace framings in each document in a collection of authoritative US texts. In this way, we can measure the variation, over time (and between countries),2 in authoritative US texts. For the textual material, we will use US presidential papers (Clinton et al. 1993–2014), which can be regarded as the most authoritative speech acts constituting the social realities of US foreign policy.

An earlier dataset reveals the monthly frequency of the words “military” and “innocent” in US presidential papers (Kivimäki 2019). For this article, it was necessary to add the monthly frequencies of the remaining indicative words and create an updated dataset based on the coding of methods of protection (related to power-centricity). Furthermore, to make the data relational (to avoid bias based on the monthly variation in the volume of textual material), the data on absolute frequencies were adjusted by dividing the monthly frequency of each indicative word by the sum of all monthly frequencies.

To avoid measuring events related to the turbulence surrounding the collapse of the Cold War order and to focus on the new, post-Cold War era, data collection for the post-Cold War period began at the start of 1993. The publication of US presidential data lags behind, while the datasets used began three years before the writing of this article. Thus, the textual data from US presidential papers extends only to the year 2014. Consequently, the full data focus on 21 years, starting from the beginning of 1993. The data are freely available at the University of Bath Research Data Archive at https://researchdata.bath.ac.uk/id/eprint/1466, accessed on 12 January 2025 (Kivimäki 2025).

5. Operationalising the Level of Violence

Regarding the level of violence, in this article, we assess the levels of fatalities of organised violence in conflicts in which the US participates.

Table 3. US protective military operations.

Country	Started	Ended
Serbia/Kosovo	24 March 1999	10 June 1999
Afghanistan	26 September 2001	31 August 2021
Pakistan	19 December 2001	7 October 2017
Iraq	20 March 2003	31 December 2011
Somalia	8 January 2007	Not ended
Yemen	18 March 2008	4 February 2021
Libya	19 March 2011	30 October 2011
Mali	13 January 2013	Not ended
Syria	4 July 2014	Not ended
Iraq	8 August 2014	27 July 2021
Libya	13 November 2015	30 October 2019

shows the military operations the US has been involved in during the post-Cold War era. US participation in UN-led peacekeeping is not considered, as there, the leadership is represented by the UN, and thus US framings are not necessarily reflected in US actions.

These operations are defined as those that an earlier version of UCDP data identifies as instances wherein the US acts as a supporting conflicting party with troops in the country or conducts deadly drone operations from outside the target country’s territory, and in which the US is not acting under UN authorisation as part of a UN peacekeeping mission (if it were, then UN documents would inform the framing of the operation).

Violence in these “US wars” is measured using UCDP data (Davies et al. 2024), with the assumption that US participation in even one of a country’s conflicts affects fatalities in all of them, rather than focusing solely on the specific wars wherein the US has direct military involvement. For example, if US forces engage in a fight against Gaddafi’s government in Libya, the weakening of the government caused by this involvement also affects the wars the government is fighting with other armed opposition groups, even if the US does not directly participate in these additional conflicts.

As an indicator of violence, it would be possible to examine each month and focus on the fatalities of wars in which the US participates during those months. However, this approach is vulnerable to two criticisms. First, in a specific month, if the US initiates participation in a conflict that previously caused one million monthly fatalities but reduces these fatalities to 100,000, this indicator might misleadingly suggest that US war fatalities have increased by 100,000, even though US involvement reduced overall fatalities by 900,000. Thus, while the monthly fatalities of organised violence in US wars indicate how much war participation the US has had, they do not indicate US contributions to the reduction in such violence.

Second, such an indicator could be criticised for failing to account for how the use of force may initially increase violence before ultimately reducing it. It could be argued that a forceful approach during the initial months is effective in the long term and that this tough, security-oriented strategy at the outset of an operation should be considered a positive contribution to an operation’s ultimate success. If we used the indicator described above, we would have to conclude that the first months of the operation were less successful than the later months. However, without the actions taken during the initial months, the success achieved in subsequent months may not have been possible.

The indicator of success for US conflict prevention used in this study addresses this issue by assigning the same monthly success rate for each month of the operation. This success rate is calculated as the difference between the monthly average fatalities during the entire operation and the monthly average fatalities during the three years prior to the start of the US operation. For example, if a conflict causes an average of 10,000 fatalities per month during the three years before the US operation, and a US operation lasts for five months—beginning with a high escalation that doubles the pre-operation fatalities but then reducing fatalities to zero until the last month, when they return to 10,000—the average number of fatalities during the operation is 6000. Therefore, the success rate of the US operation during each of the five months is counted as 4000 lives saved per month. This success rate is then compared to the levels of the different aspects of security and peace framing.

6. Is Peacification Naïve? Is Security Framing the Only Realist Choice in the World of Atrocity Criminals?

A qualitative analysis of US presidential utterances reveals that US conflict operations tend to predominantly follow the logic of security framing. The quantitative analysis of the relational and agent-centric framings also reveals the dominance of this element of securitising speech. The word “security” appears 30% more frequently in US presidential vocabulary than the word “peace”. In most interventions, presidential rhetoric reflects a narrative in which conflicts are attributed to bad actors (terrorists, dictators, atrocity criminals) who must be deterred and eliminated. The assumed characteristics of “the other” make security framing appear realistic, while peace framing is often portrayed as naïve.

The underlying assumption is that these bad actors are aware they are committing unjustified atrocities for personal or group gain, and that the use of force and deterrence can dissuade such behaviour. Negotiating compromises—a relational strategy—is deemed ineffective because the US perceives its adversaries as criminals and opportunists. All this would make us assume that the securitisation of existential threats from organised violence in the US is not very successful in general.

However, the elaboration of the indicators for different aspects of peace and security framings clearly demonstrates variations in the levels of the two framings. This indicates that the securitization of existential threats is not the natural response that research has often assumed it to be. The development of various aspects of US security framing during the observation period is illustrated in Figure 1, Figure 2, Figure 3 and Figure 4. Figure 1 indicates the development of relationality, Figure 2 the level of emancipation of power-centric bias, Figure 3 the level of military defence specific political demands and Figure 4 the righteous moralistic defence of political demands. In these figures, the dots represent monthly values for framing indicators, while the Lowess curve is used to depict non-linear trends.

Relational thinking (Figure 1), which had been on the rise until the onset of the “humanitarian interventionism” beginning with the Kosovo operation in 1999, declined rapidly, but began to recover following the failure of the humanitarian intervention in Libya during the latter half of President Obama’s term. Power-centricity (Figure 2) and righteous unyieldingness (Figure 4) also declined or remained relatively unchanged until the onset of the War on Terror after September 2001, at which point both framings gained dramatic prominence. While power-centricity continued to rise, righteous unyieldingness peaked during the final year of President George W. Bush’s administration, and began to decline during Obama’s presidency.

Military-oriented unyieldingness (Figure 3) decreased during President Clinton’s tenure but began to rise between the Kosovo intervention and September 2001, increasing almost exponentially until the end of the observation period.

While it seems clear that the elements of security framing have generally been strengthened with the US assumption of a more unilateral role, with unilateral interventions and the disregard of cooperative UN leadership in global security governance, it is equally clear that the US’s ability to reduce fatalities of organised violence has deteriorated with this increase in the prominence of security framing. Using the above developed indicator of success, we can see that the trend in the level of success in the US’s protective operations is declining, and that in general, the more active the US is with its primarily security-framed protective military operations, the more host countries of US operations lose lives (see Figure 5).

Thus, the mega-trends seem to support our two suspicions towards framing existential threats as security issues, and suggest on the one hand that the generally securitising discourse of US policies is not efficient, and that peace framing could be more conducive to the reduction in violence. In general, US military interventions have not reduced levels of violence, while the increase in the prominence of security framing at the end of President Clinton’s tenure, and especially after the beginning of the War on Terror, seems to be associated with a failure to prevent violence.

A qualitative analysis of the US presidential discourse suggests a way to understand this. First, the agent-centricity of security-framing leads to diagnoses according to which the problem of violence is caused by evil agents, not by deteriorating relationships (with the Muslim world, for example); this diagnosis leads to prescriptions of forceful, power-centric and unyielding policies. President Bush’s policies after the terrorist strike in Washington and New York on the 11 September 2001 exemplifies this logic:

“These are people that you just cannot reason with. You can’t negotiate with them. Therapy is not going to work with them. [Laughter] They’re coldblooded people”.

(Bush 2005, p. 1317)

However, an examination of the discourse of these adversaries reveals that they consider themselves legitimate actors, and interpret US aggression as evidence of the violent nature of the order the US seeks to impose. As a result, they view resistance as necessary. For Saddam Hussein, US unilateral military actions and the use of force were seen as criminal acts that justified violent resistance. The foreign powers’ non-Muslim identity added a religious dimension to the resistance against the “infidel” aggressors:

“Each one of us in the family of the faithful, patient, oppressed Iraq by its evil enemies must remember and not forget that these days will add to earning you the glory you deserve before God”.

(Hussein 2003, p. 1)

Since the US security framing identified evil opponents as the cause of threat rather than the disputes and bad relationships, the general logic especially during the War of Terror emphasised the use of military power to change the ways of these evil enemies:

“To meet these emerging threats, I called for a significant increase in our defense budget... American military strength is the key to preserving peace...”

(Bush 2002, p. vi)

However, given the reversed perceptions within the good-guys-versus-bad-guys framework, US enemies perceived the use of power as an incitement and motivation for violence, rather than as a deterrent. Instead of interpreting US violence relationally, as a response to the actions of dictators or terrorists, the lack of dialogue led them to conclude that “what they did was motivated by the desire to attack and take over with oppression and aggression. However, what the lion of Islam does is to rightfully defend the faith and one’s self and what is sacred” (Osama Bin Laden 2016, p. 3). Thus, power acted as a motivator rather than a deterrent for violence.

Furthermore, associating political demands with safety blurred the distinction between the two, making it impossible to recognize which compromises could end violence. Instead of focusing on the political demands of the enemy, security framing leads the US president to think that the means the enemy uses represent its objective: “They want to create such havoc on our TV screens by killing innocent people that the American people...” (Bush 2007, p. 1519). Yet terrorists, for example, did not seek US compromises on the acceptability of civilian targeting; terror, for them, was a means rather than an objective. Even while insisting on the prohibition of civilian targeting, the US could have pursued compromises in its policies, such as alleviating suffocating sanctions and refraining from military interventions that were not legally mandated by the UN Security Council. Thus, focusing on and defining the opponent based on their violent methods (used by both the US and its enemies) rendered compromises unattainable, as compromises need to focus on the objectives of the opponent, not their means of war. This conflation may also explain why, at a macro level, security framing appeared to be associated with an increase in violence.

However, a monthly investigation of variations gives more detailed results than the overly general observations of the mega-trends. During the 252 months under investigation, the relative word-frequency of “peace” (peace frequency divided by the frequency of all the indicators of the two framings) demonstrates clear variation, indicating that US presidents sometimes do choose to peacify organised violence, while at other times securitising it. Correlation tests show a moderate and statistically very significant relationship between relationality, as indicated by the relative frequency of the word “peace”, and the success of US operations (0.3846, p < 0.00005, n = 252). The association is equally strong when considering the frequency of “peace” and the following month’s indicator of US success in reducing violence (0.3843, p < 0.00005, n = 251). Thus, relational framing predicts success in US efforts to prevent organised violence.

Similarly, security framing, measured by the relative frequency of the word “security”, is statistically very significantly and negatively associated with success, whether contemporaneous or lagged. The relationship is strongly negative when measured contemporaneously (−0.4481, p < 0.00005, n = 252), and even stronger when using a lagged success indicator (−0.4573, p < 0.00005, n = 252). Therefore, agent-centric framing predicts poor US success in reducing organised violence.

The association between the variable that reflects the ratio of relational peace framing to agent-centric security framing (frequency of “peace” divided by the frequency of “security”) and US success in reducing violence is even stronger (0.4921, p < 0.00005, n = 252). The lagging of the success indicator further strengthens this conclusion (0.4940, p < 0.00005, n = 251). Consequently, relationality in framing safety and violence emerges as a powerful predictor of success. This aligns with qualitative evidence, supported by texts from US opponents, suggesting that the agent-centric, good-guys-versus-bad-guys approach characteristic of security thinking underlies the challenges in the US’s approach to violence.

This observation holds when examining power-centricity, considered an element of security framing. Power-centric authoritative speech also appears to distinguish US operations from cases where the US does not engage in wars. Power-centric speech dominates presidential clauses of cosmopolitan protection in 55% of cases, meaning US presidential utterances of the word “protect” are associated with a method requiring changes in someone else’s behaviour in 55% of instances. In areas where the US ultimately conducted military interventions, the power-centricity of presidential speech varied between 74% and 100%. In other words, nearly always, when the president spoke of protection, it involved changing someone else’s behaviour to achieve protective objectives. Thus, either power-centricity leads to escalation, or the increase in violence leads to power-centricity. The investigation of the association between power-centricity and US success in the reduction in fatalities of violence using lagged indicators shows that the association with power-centricity to failure is strong. The level of power centrality is statistically very significantly, though only weakly, associated with higher levels of violence during operations. The relationship between power-centric discourse and violence is bidirectional, as demonstrated by correlation tests using lagged variables. However, power-centric discourse tends to follow violence more than the reverse (see

Table 4. Association of elements of securitising and peacifying speech with US success in the reduction in fatalities of organised violence.

	Association with Fatalities of Organised Violence
Relationality vs. Agent-Centricity
Relative frequency of “Peace”	0.3846/0.3843 1, n = 252/251, p < 0.00005
Relative frequency of “Security” (stemmed)	−0.4481/−0.4573, n = 252/251, p < 0.00005
Frequency of “Peace” divided by “Security”	0.4921/0.4419, n = 252/251, p < 0.00005
Non-Power-Centricity vs. Power-Centricity
Non-power-centric divided by power centric	0.2438/0.2236, n = 187/186, p = 0.0008/0.0022
Association vs. Dissociation Between Safety and Political Values/Demands
Relative frequency of “Innocent” (stemmed)	−0.2699/−0.2718, n = 252/251, p < 0.00005
Relative frequency of “Military” (stemmed)	−0.3037/−0.2763, n = 252/251, p < 0.00005

¹ The contemporaneous association between discourse and violence is mentioned first, and the association between discourse and lagged violence is second. Thus 0.3846 is the correlation coefficient for the contemporaneous association and 0.3843 the lagged association, and 252 is the number of contemporaneous observations, while 251 is the number of lagged observations.

).

The prominence of the word “military” in presidential speech shows a moderate and statistically very significant negative association with success in reducing fatalities from violence. Additionally, the frequency of the word “innocent” appears to be a systematic, albeit weak, predictor of failure in reducing violence (see Table 4). Thus, an approach prominently involving the stubborn defence of values predicts failure in conflict prevention, though not as strongly as indicators of agent-centricity and power-centricity.

The American demonisation of the values of “dictators” and “terrorists” and the idealisation of their “innocent” victims—such as some groups opposed to the rule of Bashar al-Assad or Moammar Gaddafi—illustrates the failure of this overly simplistic framing. This framing justifies violence, but since the demonised enemies do not perceive the situation in the same way, US violence in the defence of values deemed to protect “the innocent” may provoke further violence rather than deterring it.

Table 4 summarises the quantitative results of this study regarding the main indicators of peacification and securitisation, and their relationship with changes in fatalities from violence.

7. Conclusions

This article introduces an alternative framing and speech act for situations in which an international agent addresses existentially threatening organised violence. Earlier conceptual and theoretical literature, as well as political practice, has assumed that such situations naturally create opportunities for those in power to securitise issue areas that pose a threat through securitising speech. However, this article reveals that viewing this as a natural way of addressing existential threats is based on three naturalised but unsustainable premises.

First, while the discussion in Security Studies has debated who should be considered the referent object of security, and while the Copenhagen School of Securitisation has participated in this debate by suggesting restraint in expanding the security realm to cover more referent objects, this study shows that the entire idea of focusing on agents as referent objects of security can be challenged. It is also possible—and it appears that it may, indeed, be beneficial—to focus on relationships rather than agents whose security must be guaranteed. It seems evident that focusing on relationships gives rise to more peaceful social realities. Thus, prioritising relationships over the security of referent objects may represent a more effective framing for preventing organised violence.

Secondly, while classical Realists consider power a value in itself, and Neorealist analysis views it as the main means in international security relations, this study demonstrates that power should not be seen as the natural, prime means of preventing organised violence. A focus on power is a logical consequence of agent-centric security thinking. However, it seems that approaches to protecting people from organised violence that emphasise power are associated with failure, whereas approaches grounded in restraint rather than power—something that logically follows from a relational perspective—are associated with success.

Finally, the association of political values with safety in security framing tends to extend securitisation to political values, and to justify policies in which conciliatory, dialogical, and understanding approaches to adversaries are met with security measures. This is the area where security and peace framing most violently clash in politics. Militant stubbornness and the insistence on treating one’s values as universal and non-negotiable (the attitude that we must win first before negotiating and before the conflict can be ended) do not naturally arise from existential threats. Instead, such an element of security framing also correlates with a failure in addressing the existential threats posed by organised violence. Peace requires compromises, and a willingness to negotiate, rather than strength in defending one’s own values. This openness tends to foster processes of mutual concessions and peace.

While this article provides only an initial exploration of the relationship between success in violence reduction and these two alternative framings, it nonetheless supports the argument that peace framing should not be dismissed as naïve or impractical in situations of existential threat. Demonstrating that peace framing actually causes success, and that securitisation causes failure, would require more systematic research than the mere exploration of correlative relationships. However, given the clear correlative association between levels of security framing and increases in violence, it seems unlikely that securitisation could be shown to cause success in conflict prevention. Such proof would require the establishment of a positive, rather than a negative, correlative relationship between securitisation and success.