The dual-use security dilemma and the social construction of insecurity

ABSTRACT In this article I introduce the concept of the “dual-use security dilemma,” specifically through elaborating on two main aspects that shape this dilemma. First, inspired by traditional security scholarship, I focus on the spiral dynamics of actors responding to the insecurities raised by dual-use technologies that affect this type of dilemma. Second, I further develop a securitization reading of the traditional security dilemma, tracing how social constructions of insecurities and the justification of extraordinary measures affect the dynamics of the security dilemma. Combining these two aspects, I suggest that enunciators shape the dynamics of the dual-use security dilemma by using specific rationales of insecurity to mobilize support for measures against opponents holding dual-use technologies, whose response further fuels insecurity that spirals over time. The innovative theoretical and policy implications of this research become especially important given the rise of dual-use cyber technologies.

using or adapting the technology for harmful purposes, even over the long term, creates insecurity and may spur policy makers to take measures to address the threat. Similar to the traditional security dilemma, these measures can influence an opponent's behavior, leading it to acquire military means or even to make military use of the dual-use technology. Furthermore, and also like the traditional security dilemma, these dynamics may spiral, aggravating the actors' security. Even actors whose aim was to develop technology solely for civil purposes may find themselves in an arms race or a costly situation. While militarily arming a civilian project can be very expensive (for both sides), failing to respond to a military project because of its civilian characteristics can also be very costly.
From a constructivist approach, the dilemma is shaped by how the insecurity is constructed. Insecurity is not given, but rather is socially produced through actors calling attention to certain issues and mobilizing support to address them. Moreover, how a technology is applied-for peaceful or military purposes-is also socially determined. This means a technology's dualusage is influenced not only by its technical characteristics, but also by international norms, by how it is conceived, and by how knowledge is mediated by socio-scientific agents. Seeing a technology as dual-use allows an actor to construct it as a(n) (existential) threat, even if the technology is developed for civil purposes.
The dilemma, then, is whether or not to securitize. While securitizing a potential challenge carries the advantage of preparing for a threat and mobilizing support for certain policy measures, it can also contribute to a spiral insecurity dynamic. Once a technology has been acknowledged as dualuse, an enunciator can take advantage of the accompanying uncertainty of how the technology will be used. They can securitize an opponent's dualuse technology to justify taking extraordinary measures to address itregardless of the opponent's intentions around developing the technology's military/harmful applications. Furthermore, the very act of securitizing an opponent's technology without taking any extraordinary measures-can alarm an opponent, who then initiates a securitizing reaction. All these may result in a securitizing chain reaction that decreases both actors' security.
Acknowledging the dual-use security dilemma's dynamics and the social constructions shaping it is important for three main reasons: First, it provides a way to better capture a number of aspects overlooked in the traditional security dilemma, such as actors' economic considerations, the involvement of non-state actors, and threats that go beyond military conflicts. These aspects, as I elaborate further below, are crucial, for example, to understating the chain reaction between the American government and Huawei (and Chinese government) regarding the deployment of 5G networks. Second, the suggested concept is especially promising for enhancing the study of cyber security. For example, it opens up recent scholarly thought on the security dilemma to examine cyber security (e.g., Dunn Cavelty, 2014) as well as emerging research on cyber securitization (e.g., Hansen & Nissenbaum, 2009). By extending the discussion to include the dual-use security dilemma and the social elements affecting its dynamics, we can compare the emerging cyber technologies with other dual-use technologies, such as space and nuclear technologies.
Third, this paper further develops the social constructivist approach to the security dilemma. Relying on a number of works examining these dynamics (Olesker, 2018;Roe, 2005;Watson, 2013;and especially, Van Rythoven, 2020), I elaborate on how the "traditional" security dilemma is affected by social constructions, securitizing moves and countersecuritizing moves. Reading the security dilemma through securitization theory does not refute traditional theoretical approaches to these dynamics, but it adds another layer of understanding of them. It not only establishes the foundation for exploring the dual-use security dilemma but also illuminates the dynamics of the traditional security dilemma.

Dual-use and the dual-use security dilemma
Dual-use Dual-use refers to items that can be used for both military/harmful purposes and civil purposes (Atlas & Dando, 2006;Lin, 2016, ftn. 1, p. 113). 1 These include not just the technology but also its developmental research and resultant products/artefacts (Forge, 2010, p. 112). Dual-use technologies can be categorized according to their transformability from civil to military/ harmful applications. For example, computer technology can be used without any modification for both military/harmful and civil use; whereas other technologies require adaptation to be used for both purposes (Molas-Gallart, 1997, pp. 370-371). Another way to categorize dual-use technologies concerns their original use. Some technologies were originally developed for military/harmful use but were adapted to civil purposes, such as GPS, while other technologies originally developed for civil use but were adapted to military/harmful purposes such as computers chips (Brandt, 1994, pp. 359-360;Lachow, 1995, p. 127;Forge, 2010, p. 117).
Some production of dual-use technologies can decrease development costs (Molas-Gallart, 1997, p. 376) or contribute to advancing human needs and values (Miller & Selgelid, 2008, p. 9). But they can also have disadvantages, primarily due to their potential harmful effects. This is why Miller and Selgelid assert that dual-use technologies poses an ethical dilemma for researchers and those who authorize, assist, or impede research (Miller & Selgelid, 2008, p. 9).

Dual-use security dilemma and current scholarship
The existence of dual-use phenomena-research, technologies, and artefacts -intensifies uncertainties in the international arena. Actors do not know whether their opponents will use their technology for defensive or offensive purposes, or, more fundamentally, whether the technology is intended for harmful use at all. While this dilemma overlaps with two main research areas, this paper is a first attempt to theorize the dual-use security dilemma as a distinct dynamic of security dilemma.
One area of research concerns the study of the "dual-use dilemma." Unlike the dual-use security dilemma, the dual-use dilemma does not refer to an international (security) dilemma (where the behavior of one actor affects the (in)security of another), but instead to an intrastate ethical dilemma. In the latter, actors need to balance between different valuesmainly between the advantages of a technology's civil applications and its potential harmful uses (e.g., Atlas & Dando, 2006;Evan & Hays, 2006, p. 106;Miller & Selgelid, 2008, p. 9;National Research Council, 2004, p. 19). The dual-use dilemma also extends to capture strategic international dynamics for suppliers of dual-use technology such as nuclear technology. In such situations, the dual-use dilemma stems from the strategic advantages in helping another country versus the fear the country will take advantage of this assistance and develop a military nuclear program (Fuhrmann, 2012, p. 2, 12). According to Fuhrmann (2012, p. 101), uncertainties play a prominent role in the dual-use dilemma. However, this dilemma differs from the security dilemma and the suggested concept of the dual-use security dilemma, where insecurity spirals between two rivals.
A second area of research is scholarship exploring how dual-use technologies affect the dynamics of the security dilemma. While these studies explore interactions between rivals, they do not develop broader frameworks to trace the dynamics related to different technologies, nor do they acknowledge the unique considerations around these technologies and the threats they pose. Some scholars focus on specific dual-use technologies. Johnson-Freese (2007, p. 6, 244-245), for example, briefly considers the dual-use characteristics of space technology that affect the (traditional) security dilemma. Similarly, a number of scholars employ the security dilemma concept to explore the cyber dual-use technology. Some have debated the unique implications of cyber technology for the offense-defense balance (Saltzman, 2013;Shaheen, 2013; but see Gartzke & Lindsay, 2015, pp. 322-326;Garfinkel & Dafoe, 2019, pp. 750-751;Slayton, 2017). Others use the traditional security dilemma's spiral of insecurity concept to explore dynamics distinctive to cyber domain (Buchanan, 2017;Dunn Cavelty, 2014).
Some even more specifically connect the latency and uncertainties of dualuse technologies, such as nuclear and space, to the security dilemma. Glaser provides a useful point of departure by suggesting that actors' power is a key factor affecting the security dilemma. Power, according to him, is "the ratio of states' resources that can be converted into military assets" (Glaser, 2010, p. 76). From this perspective, dual-use technologies can become a source of latent power affecting the security dilemma by shaping rivals' calculations. As Volpe (2019) argues, given the dual-use nature of nuclear technology, "the accumulation of latency-even for peaceful energy purposes-sets the stage for the security dilemma by threatening others with the prospect of proliferation." This uncertainty further affects opponents' responses (Volpe, 2019, p. 824) and could foment nuclear proliferation (Fuhrmann & Tkach, 2015, p. 444). Similarly, Peoples (2011, p. 76) wonders whether the nonmilitary uses of outer space technologies "generate dynamics of military competition due to the latent dual-use potential of many commercial space technologies." While I rely on these important studies to develop my own conceptualization, my research takes them further in a number of ways. First, I extend their theoretical frameworks to trace the general dynamics that can be applied to different dual-use technologies (space, nuclear, cyber, etc.) and that take place between international actors responding to each other. Second, rather than seeing the dual-use characteristics as elements affecting the security dilemma and its intensity (e.g., Volpe, 2019, pp. 816-817), I argue that there are a number of advantages in viewing it as a distinct dilemma. Such acknowledgment, as further discussed below, allows us to capture different threats that initiate and fuel the dilemma and the different kinds of actors involved, and it also opens up comparisons among different dual-use technologies and among different factors affecting this specific dilemma. Third, I explore the previously underdeveloped issue that these dynamics of insecurity in general and of dual-use technology in particular are shaped by social constructions.

The dynamics of the dual-use security dilemma
The dynamics of the dual-use security dilemma stem from the uncertainties that accompany these technologies. According to the traditional security dilemma, actors must arm themselves and form alliances because they cannot know whether an opponent will use its means for defensive or offensive purposes. But even arming for defensive purposes enhances the threat to the opponent and thus may lead to a spiral: resulting in an arms race and, in turn, leading to war (Booth & Nicholas, 2008, pp. 1-41;Jervis, 1978;Roe, 2005, pp. 8-24;Tang, 2009;Glaser, 2010, pp. 55-92;Van Rythoven, 2020, pp. 481-483).
Dual-use technologies create another layer of uncertainty. Actors cannot know the intentions of other actors that develop, acquire, or produce these technologies and artefacts: that is, whether they have military/harmful purposes or civil purposes. This uncertainty is created even when actors simply initiate advanced research on dual-use technologies without acquiring specific dual-use products. Actors' civil purposes concerning dual-use technologies affect the dynamics significantly, adding a number of considerations that are overlooked in the traditional dilemma. These include the fulfilment of social needs and economic incentives, such as the interests of trading partners.
In a dual-use security dilemma, Actor B fears that Actor A is developing a dual-use technology that may have harmful applications. Given the uncertainty and the potential grave risk the technology may pose, Actor B may need to act to address the challenge. But actors must prepare well in advance due to the resources and time required to address potential longterm technological challenges.
Foremost, actors may try to advance arms control agreements to decrease the intensity of the dilemma (e.g., Glaser, 2010, p. 88; see also Booth & Nicholas, 2008, pp. 89-91). Given the difficulty to achieve such agreements, as further discussed below, Actor B may also take one of following actions to address the potential threat Actor A poses: initiating a preventive strike (or developing strike capabilities); acquiring counter-measures to Actor A's technology; or developing similar dual-use technologies. Three strong rationales stand out for such responses: First, Actor B may need to ensure it has the capabilities to prevent Actor A's decisive achievement. Given the dualuse security dilemma, if Actor B fears its technology lags behind Actors A's technology, then actor B will invest in dual-use technology (or in preventive/defensive measures) despite uncertainty around whether Actor A intends its technology for military/harmful purposes. 2 Second, even if dual-use technologies are only employed for civil purposes, they signal state capabilities. For example, an advanced civil space program symbolizes military capabilities and indicates the means to launch military rockets missiles (Paikowsky, 2017, pp. 49-51). Third, given the prestige that accompanies the development of novel leading technologies (e.g., Gilady, 2018, pp. 122-123;Varadarajan, 2004), an actor that fails to develop a dual-use technology held by an opponent may find its status and identity challenged.
The response of Actor B to the threat posed by Actor A contributes to the spiral dynamic of the dilemma. Regardless of the precise reason or justification for the response, the actions Actor B takes may, in turn, pose a threat to Actor A. This spiral negatively affects both sides. At the most basic level, this dynamic forces both sides to heavily invest in technological development, and they may find themselves in a very expensive technological "arms race." 3 Furthermore, similar to the traditional security dilemma, the dualuse dilemma spiral can escalate into a violent conflict. This can occur in two ways: first, the dilemma may transform into a traditional security dilemma in which actors develop only the military applications of the technology; second, an actor may execute a preventive or preemptive strike to stop its opponent from developing the technology and in this way attempt to escape the dual-use security dilemma. Israeli discussions of attacking Iranian nuclear infrastructures exemplify this kind of reasoning (Lupovici, 2016); while there is some degree of uncertainty as whether military technology was being developed in Iranian nuclear sites, the fear that nuclear technology would be used for military purposes in the future supported such a move (e.g., Netanyahu, 2012).
The social constructions of the dual-use security dilemma Threats are not simply given. They are constructed and shaped through social dynamics. This does not mean that the objective elements that affect insecurity are irrelevant. Rather, these elements are mediated through interpretation and carry effects through discourse that identify the threat and justify the response (e.g., Lipschutz, 1995, pp. 1-23). Actors, intentionally and unintentionally, create social structures that produce (in)security and threats.
The dual-use dilemma is thus shaped by two interrelated social mechanisms. The first is the social production of the dual-use qualities of the technology. The second is the actors' construction of the technology of their opponents as an existential threat, which is embedded in a broader social context that includes the actors' relations, politics, identity, and history andcontinues previous securitizing moves. Through the combination of these, actors construct uncertainty over whether their opponent's technology is meant for civil or for military/harmful (and potentially existentially threatening) purposes, and thus they allocate resources to address it. This is even if the opponent has no plans for harmful or military application of the dual-use technology. The securitizing moves an actor takes mayaffect the security of other actors, who in turn may take additional measures and contribute to the spiraling escalation.

A securitization theory of the security dilemma
The concept of securitization captures the social dynamics of insecurity. Securitization is a process through which an intersubjective understanding that something is an existential threat is developed between a securitizing actor (enunciator) and audience. This intersubjective understanding allows the actor to call for exceptional measures to deal with the threat (Buzan et al., 1998; see also Buzan & Hansen, 2009, pp. 212-217;Gad & Lund Petersen, 2011). A crucial element of this process is the speech act of security, as it is the utterance itself that constitutes an issue as a security issue by labeling it as such (Buzan et al., 1998, pp. 26-27). A number of facilitating conditions related to the language used, the securitizing actor, and the "conditions historically associated with the threat" affect the speech act (Buzan et al., 1998, pp. 32-33;Floyd, 2010, p. 13). However, scholars have challenged the emphasis on the semantics of security, pointing to the importance of images and to external conditions affecting the securitizing move that are not related to language (e.g., Vuori, 2008;Williams, 2003).
It should also be noted that while the concept of securitization was originally developed to explore the constitutive elements of "new" security issues, such as migration, epidemics, and the environment (Balzacq & Guzzini, 2015;Buzan & Hansen, 2009, p. 215), scholars have begun using this theory to study "traditional" security issues (Hayes, 2013;Lupovici, 2019;Roe, 2008;Vuori, 2008). I follow this line of studies, and especially Van Rythoven's (2020) article. 4 Van Rythoven uses the concept of securitization dilemma to explore the unintended consequences of the security dilemma dynamics. Among other things, he aims to highlight how this concept bridges traditional and constructivist theories of security, and to justify epistemologically and conceptually the combining of the security dilemma with securitization scholarship. 5 Van Rythoven provides a promising point of departure to incorporate securitization theory into the study of the security dilemma. First, he defines the securitization dilemma as "a difficult choice where a securitizing move represents a powerful and attractive opportunity for political mobilization, but with the danger of perverse and unintended consequence" (2020, p. 479). In short, the focus of the dilemma is diverted to the question of whether to securitize or not. Second, securitization theory also allows the capturing of different non-traditional kinds of threat, and not just threats to sovereignty, that initiate the security dilemma for different kinds of actors (and not just to state actors) (Van Rythoven, 2020, p. 484).
I further elaborate on Van Rythoven's assertions by acknowledging the spiral dynamics of securitizing moves. For Van Rythoven, a key aspect of the dilemma is the unexpected consequences of a securitizing move. He highlights the uncertainty that surrounds securitizing moves, as enunciators cannot know whether the audience will accept them or what the impact will be of implementing the practices stemming from these moves (Van Rythoven, 2020, pp. 486-488). While all these uncertainties are important, another key, and underdeveloped, question is how securitizing moves affect other actors. I argue that this dynamic may create a spiral of insecurity where actors respond with a securitizing move to a securitizing move initiated by another actor, thus affecting both actors' security.
Important to understanding security dilemma dynamics is the acknowledgement that securitization is a process rather than a single act (Roe, 2008, pp. 617-618, 624;Watson, 2011, p. 10). Recognizing this allows us to consider the political context in which securitizing moves are embedded, as well as the political effects of securitizing moves (Buzan et al., 1998, p. 25;Guzzini, 2011, p. 331;Hayes, 2013, pp. 16-17). A number of scholars have extended these ideas by considering the different dynamics through which the same security issue is repeatedly securitized. They explore the concept of re-securitization, when an already-securitized issue is put back on the agenda (Mavelli, 2012, p. 190;McDonald, 2011) or attains a new climax (Lupovici, 2016). Scholars have also explored the international effects of securitizing moves. Beyond domestic consequences, securitizing moves' political effects and outcomes influence other actors, who respond to them as a result of either the rhetoric employed or the behavior triggered (Buzan et al., 1998, pp. 205-206; see also Lupovici, 2019;Waever, 2011, p. 476). Especially important for capturing dynamics around the chain reaction of a security dilemma is a counter-securitizing move, which occurs when actors resist a securitizing move. Actors can resist a securitizing move in different ways, 6 and against different elements of the targeted securitizing move (Stritzel & Chang, 2015, pp. 551-552). Although Stritzel and Chang (2015, p. 553) do not specifically connect their discussion to the security dilemma, they argue that "[a]n explicitly interactive reading of processes of securitization versus counter-securitization … can establish fruitful conversations specifically with military strategy." Thinking of securitizing moves as part of an iterating process that affects other (international) actors captures the essence of the security dilemma ( Table 1). The traditional security dilemma is fueled by the "objective" characteristics of the means acquired by the actors and focuses on whether to arm to prepare for an emerging threat, even though the act of arming may itself be threatening to the opponent. Conversely, the securitization approach to the security dilemma focuses on the actors' spiraling constructions of insecurity. From this perspective, a security dilemma erupts when an issue is securitized and counter-securitized by rival actors. An actor's securitizing move is constructed as an existential threat by another actor and thus it is spirally aggravated. A socially constructed challenge, that may refer to state sovereignty, but may stem from challenges in different sectors (e.g., military, societal). What is the dilemma To arm or not to arm Arming in response to the armament process of an opponent may fuel an arms race and confirm the threats for both sides, and thus escalate even if neither wants a war.
To securitize or not to securitize Securitizing move may help to gain domestic support and legitimacy for a desired policy; however, the move itself (regardless of execution of measures) results in securitizing and counter-securitizing moves that decrease both actors' security.
The securitization approach of the dilemma is based on the view that, on the one hand, a securitizing move may be advantageous for the securitizing actor by mobilizing support for extraordinary measures, legitimizing their response to an opponent's challenge or, alternatively, winning them some domestic political gains. On the other hand, the move itself may invite the opponent to make securitizing moves of its own. The result may be a spiral dynamic in which each actor conducts a securitizing move where they securitize the securitizing move taken by an opponent. The dilemma then is whether to securitize and enjoy the benefits of it while also facing the risk of negatively influencing other actors or providing tools for the opponent to perform securitizing moves of their own.
When the means Actor A holds or develops (or even the rhetoric related to these processes) are constructed by Actor B as an existential threat and Actor B can justify taking extraordinary measures to address the alleged insecurity, then Actor B contributes to the dynamics of the security dilemma. In turn, Actor A may securitize Actor B's response-that is, Actor B's securitizing move is securitized by Actor A. More specifically, Actor A may construct a number of things as an existential threat: (1) Actor B's construction (labeling the means acquired by Actor A as an existential threat); (2) Actor B's justification to take extraordinary measures; or (3) Actor B's execution of the measures legitimized through the securitizing move.
Securitizing of any of these issues by Actor A can lead to attempts to justify additional means Actor A should take and can expedite attaining the means of threat to Actor B that have already been justified. The securitized issue becomes part of the opponent's securitizing move: it provides a pretext for enunciators to perform their own securitizing move and it also gives both sides tools to further securitize the situation, and so on. This creates the dangerous spiral dynamic. Each securitizing move-and the ensuing extraordinary measures-becomes a strong justification for the opponent to re-securitize the opponent's moves and attain a new securitization climax.
A securitization theory of the dual-use security dilemma I build on a securitization reading of the traditional security dilemma to present the dual-use security dilemma (see Table 2). Basically, the dilemma in the latter is whether to securitize an opponent's dual-use technology. On the one hand, by securitizing these technologies, an actor can attain domestic support and legitimacy for policy measures, including measures addressing the (constructed) threat.
Conversely, the securitizing move itself, apart from any execution of these measures, may alarm an opponent or even provide it a pretext for its own securitizing moves, thus initiating a spiral dynamic of insecurity. This may especially be tragic in situations where the actor whose dual-use technology was first securitized had no plans to use the technology for military/harmful purposes.
I suggest that the dual-use security dilemma is shaped by two kinds of social dynamics. First, social processes shape whether and how a technology can be seen as dual-use. Second, once a technology is seen as dual-use, enunciators are able to rely on this knowledge to further securitize it.

Social attributes of dual-use technologies
Foremost, the production and distribution of knowledge play a part in shaping a technology as dual-use. Actors can come to acknowledge the peaceful or civil application of an originally military or harmful technology-as seen, for example, with GPS. Conversely, a social process can emphasize the military or harmful applications of a civil technology, weaponizing or militarizing it. For example, airplane technology can be used for military purposes (e.g., transporting military supplies) but can also be used to develop weapon systems such as fighter aircrafts. The result of such processes is the recognition that a technology carries dual-use characteristics. We see this in the weaponization and militarization of space technologies (Mowthorpe, 2004;and in Bormann & Sheehan, 2009), which have been constructed as means that are not limited to civil purposes. Knowledge of the military applications and harmful uses of space technologies-including existential threat to civilian infrastructures, commerce and societyallows enunciators to securitize space technologies (Peoples, 2011, pp. 85-90) and thus to make them dual-use technologies. Nonetheless, how technologies can be used is socially determined and interpreted by different actors. From this perspective, social constructions also affect whether and how technologies are seen as dual-use. For example, classifying items as dual-use is place-dependent and can change over time (Forge, 2010, p. 116). Unlike deterministic theories of technology that consider the challenges of dual-use as an inherent characteristic of the technology, constructivist approaches acknowledge that these are by definition affected by social forces and inspire competing interpretations (Tucker, 2012, pp. 29-30).
For constructivist scholars, social dynamics mediate the "objective" technical characteristics of a technology and its political impact (Felt et al., 2017). Enunciators are able to construct a certain technology in different ways-as a threat, as not threatening, as peaceful, or as related to both threats and peace. These constructions contribute to whether technology is seen as dual-use.
A number of scholars have started to explore how actors construct an opponent's acquirement or development of a dual-use technology as an existential threat (Hayes, 2013;Peoples, 2011). For example, an actor's acquisition of nuclear weapons does not necessarily represent an existential threat: in fact, for some actors, another actor's nuclear weapons could be constructed as a source of security because they provide extended deterrence (Hayes, 2013, p. 3, 12). Furthermore, a military technology not only can be constructed as unthreatening but can be pacified-constructed as having peaceful characteristics (Lupovici, 2013). For example, in the "atoms for peace" speech delivered by Eisenhower before the United Nations General Assembly, he presented the idea of developing nuclear technology "for the benefit of all mankind" (Eisenhower, 1953). This view can be seen as a process of pacifization targeting two international audiences: American allies and the Soviets. The word "peace" was thus attached to the word "atom" in order to downplay its negative connotations (Krige, 2006, p. 172, 176).
In a similar way, in the United States, Western Europe, and even the Soviet Union, "space" and space technology were framed as a platform for peace and peace enhancement. This framing, in contrast to the prominent calls to militarize outer space (e.g., Erickson, 2005, p. 105), aimed to enhance the state of peace by creating a manifesto for technological and scientific cooperation (Erickson, 2005;Sheehan, 2007).
It follows then that not only is a technology's use socially determined, but a technology can be constructed both as an existential threat and as related to peace. In such situations, a technology can be seen as a dual-use technology.
Social norms also mediate between the technical elements of technologies and how they can be used. These social attributes enable or disable seeing a technology as dual-use by normatively (de)legitimizing some of its various usages. For example, if there was a full taboo on the use of nuclear weapons-that is, their usage cannot be imagined for either a first strike or for retaliation-then nuclear technology would not be a dual-use technology but only a civilian technology. 7 Furthermore, even whether a technology has legitimate civil usages might be debated. Some countries, such as Norway, construct civil nuclear applications, such as nuclear power, as illegitimate (Muller et al., 2018, pp. 299-302), thus putting additional limitations on their considering it as a dual-use technology.
In a similar way, a military technology could have potential use for civil purposes; but if these applications seem inappropriate, it limits the ability to convincingly understand it as a dual-use. A fascinating example is Project Plowshare, undertaken in 1957 in the US. It aimed to use nuclear technology-and, specifically, harmful applications of this technology-for civilian purposes. The plans included using nuclear explosives to excavate harbors and canals (Kaufman, 2013, pp. 2-6;Kirsch, 2015, p. 3). The project was terminated in the 1970s due to lack of funding and legitimization, as well as challenges around risks, feasibility, and norms (Kaufman, 2013, pp. 5-6, 230-231;Kirsch, 2015, p. 6;Peoples, 2019, p. 290). In this respect, the dual-use of the technology was its biggest hurdle: "[H]ow to avoid a conflict between the desire to use the atom in the name of peace and a growing international movement to ban nuclear weapons tests" (Kaufman, 2013, p. 16).

The construction of the existential threat of dual-use technologies
Once a technology is established as dual-use, enunciators are able to securitize not only the technology and its potential harm, but also the uncertainty around how it can be used (see Figure 1). In such situations, even if the technology is allegedly developed only for civilian purposes, it can be treated as an existential threat. For example, Iran's civil nuclear technology can be seen as an existential threat (Lupovici, 2016;Nili, 2011) only if the international community has knowledge that one can transform a civil technology into nuclear weapons. This is why Prime Minister Netanyahu, in his address to the UN General Assembly, attempted to securitize the capability of transforming a civil nuclear program into a military one. He claimed, "I know that some in the international community think I'm exaggerating this threat, [but] A country that can enrich uranium to about 3.5% will also have the capability to enrich it to about 90%," and thus will be able to produce nuclear weapons (Netanyahu, 2013).
More specifically, enunciators may refer to one of the three main potential challenges of the opponent's dual-use technology: (1) It poses a strategic threat or opens a strategic gap between the two actors; (2) it signals the strategic advantage of the opponent; and (3) it challenges the actor's prestige, status, and identity, as the opponent takes the lead in advanced technologies. While I discussed these aspects of insecurity above concerning a more traditional view of the dilemma, each of them may in fact also provide a compelling securitizing plot justifying the need to take extraordinary measures concerning different security sectors (see Figure 1). 8 The first two are directly connected to the military sector, facilitating the building up of plots of insecurity around traditional security issues (Buzan et al., 1998, pp. 49-59). The third represents challenges to an actor's prestige, status, and identity concern threats in the societal sector (Buzan et al., 1998, pp. 119-140). For example, they challenge actors' ontological security (Mitzen, 2006;Steele, 2008) as well as their ability to maintain a coherent narrative of themselves as a scientific pioneering country or to uphold routines with significant (advanced) technological states. 9 Labeling the opponent threat, justifying extraordinary measures to address it (e.g., developing similar dual-use technology and its military/ harmful application or developing defensive or preventive capabilities), or executing these measures challenges the opponent's security or becomes a tool for the opponent to take its own securitizing moves. As with the traditional security dilemma, this dynamic results in securitizing and counter-securitizing moves-and, when the dilemma spirals, to re-securitizing moves and securitization climaxes (see Figure 1).
The combination of securitizing dynamics and dual-use technologies exacerbates the negative implications of the dual-use security dilemma. When actors construct dual-use technologies as an existential threat, the insecurity may spiral even if the actors hold no military/harmful means and have no plans to develop them. The eruption of this dynamic eventually spurs both actors to allocate resources to exploring and developing a technology's military/harmful application and risk its employment or escalation.
Furthermore, an actor will be able to use its opponent dual-use capabilities to initiate securitizing moves and justify taking extraordinary measures. As a result, this may incentivize actors to develop military/harmful applications of their dual-use technologies, fearing the opponent's means as well as its ability to continue securitizing the potential threat of the dualuse technology. All of this perpetuates the spiral dynamics.
Israel's securitization of the Iranian nuclear program demonstrates this dynamic. The Iranian civil nuclear program was securitized over its potential to transform into a military program and the accompanying rhetoric of Iranian officials on destroying Israel (Lupovici, 2016). Israeli securitization of this nuclear program became a challenge to Iran. From an Iranian perspective, a challenge stems from the ability of Israeli policy makers to continue securitizing Iran's nuclear program even if Iran decided to give up its military plans. The ability of Iran's opponents to mobilize support for measures that future securitizing moves would legitimize puts Iran at a disadvantage. Furthermore, once an actor uses the rhetoric of existential threat as part of their spiraling securitizing moves (which may serve political or ideological goals), breaking away from conflictual dynamics can become domestically costly. In other words, the dual-use security dilemma emphasizes another mechanism that burdens the ability to reach arms control agreements.

Cyber security and the dual-use security dilemma
A key contribution of combining the concept of dual-use security dilemma and securitization theory lies in its potential to advance the study of the traditional security dilemma and to consider a number of aspects it overlooks. These aspects are especially important in the study of cyber technologies. I discuss these aspects and then briefly illustrate them by tracing the conflictual dynamics between the United States and China/Huawei over 5G technologies.
First, the suggested framework allows us to extend the scope of research beyond military threats and incorporate various kinds of threats: not only civil applications of military technologies (e.g., nuclear power), but also technologies used for civil purposes and in civil sectors. For example, importing information technologies has been seen as a risk both in the United States and in China, given the possibility of their use to harvest sensitive information (Creemers, 2020, p. 116).
Second, this framework emphasizes the economic considerations, which are not usually a crucial element in the traditional security dilemma. While economic cost-benefits calculations do affect the traditional security dilemma (as entering an arms race can be expensive), economic incentives are often the cause of a dual-use security dilemma. For example, acquiring nuclear technology to diversify sources of energy may affect an opponents' sense of security. Furthermore, actors may trade dual-use technologies, adding complications to both sides (Klimburg, 2012, pp. 34-42). For example, the question of whether to acquire Chinese 5G technology was a dilemma for a number of countries, who considered the economic advantages on the one hand, and the potential threat that their networks might be taken advantage of on the other (Cartwright, 2020, p. 10).
Third, the framework allows us to incorporate the dynamics of non-state actors' involvement into the research. Cyber technologies are often developed, acquired, and used by civil non-state actors. In fact, some of the most powerful and important cyber actors are private corporations, whose databases, technologies, and means exceed those of many state actors (Carr, 2011;Valeriano & Maness, 2015, pp. 164-187). These actors are capable of causing harm in various ways (Sigholm, 2013, p. 11), and a number of non-state actors are cooperating with governments (Maurer, 2018). Not surprisingly, therefore, even civil cyber technologies can be securitized. For example, technologies that distribute information (e.g., popular social networks or search engines) can be constructed as threats (Facebook, Google) in different political contexts (Caravelli & Jones, 2019) and thus initiate and fuel the dynamic of dual-use security dilemma among different kinds of actors (states and non-state actors).
Finally, the concept of the dual-use security dilemma provides a useful point of departure to advance the study of cyber security by conducting comparisons among different dual-use technologies. Various factors and attributes not only account for varying degrees of the dilemma, but also help to demonstrate that the dilemma is not a deterministic outcome.
An interesting point of departure for this discussion is the conditions that affect the intensity of the traditional security dilemma. In addition to Jervis's (1978) offense-defense balance variables, Glaser suggests power and two information-related variables: that is, the information an actor holds about its rival's motivation and the information an actor holds about how the rival sees that actor's motivation, as well as the interconnections among all these factors (Glaser, 2010, pp. 72-92). In this respect, arms control agreements may serve as an important indication of the actors' intentions and thus as a means to limit the security dilemma (Glaser, 2010, p. 88; see also Booth & Nicholas, 2008, pp. 89-91).
These factors may inspire positivist scholars to generate explanatory arguments out of these assertions and apply them to the dual-use security dilemma: for example, how cyber actors' different meanings of power (Nye, 2011) guide assertions about the severity of the dilemma. Another promising route for positivist scholars is to replace the distinguishability of offense and defense applied to the traditional dilemma with the ease of transferring a civil technology into a weapon. Volpe (2019, p. 827) reveals that the ability to distinguish the civilian and military objects of a technology affects the intensity of the security dilemma (e.g., artificial intelligence, see Maas, 2019, p. 294). This element clearly demonstrates the advantage of comparing among different (dual-use) technologies.
The offense-defense factors can be further developed based on recent innovations in the research. Garfinkel and Dafoe (2019) show how this balance can shift over time due to scaling effects. This has important implications for the dual-use security dilemma since "more highly automated and digitally-oriented conflict suggests that scaling phenomena may become much more salient" (Garfinkel & Dafoe, 2019, p. 759). The fact that dualuse technologies like computing power are developed by private corporations further complicates these dynamics.
Government control is another factor scholars can develop to explore the intensity of the dual-use security dilemma. When in full control, governments have the means to limit the acquirement of the technology in their territory. While this element has limited effect on the traditional security dilemma, it is crucial in assessing the potential challenges of the dual-use security dilemma and the ability to narrate them. Among other things, when governments do not control the technology, the degree of uncertainty is higher. Furthermore, such situations provide effective tools for narrators to point to securitized situations where many kinds of actors hold and develop threatening technologies. For example, the many private actors and corporations involved in cyber technology indicate the huge economic incentives and opportunities in this domain. However, they also mean that government control over the development, use, and distribution of cyber technologies is limited compared to other dual-use technologies, such as space and especially nuclear (e.g., Saco, 1999). 10 Acknowledging social constructions may point to additional dynamics that affect the severity of the dual security dilemma, such as the weaponization and militarization of the technology. In recent decades cyber technologies have been weaponized and militarized (e.g., Deibert, 2003;Nissen, 2016), indicating or following securitizing moves. We see these moves, for example, in the establishment of cyber commands in various armies (Harrison, 2012, p. 53). These both result from and demonstrate how knowledge of cyber challenges and cyber violence has been institutionalized. The perception of cyber means as violent is not only a legal (or normative) question (Harrison, 2012, pp. 63-75), but a result of social dynamics that allow actors to agree upon this meaning and of enunciators' ability to frame these means in a certain way (Lawson, 2012). These dynamics thus not only shape how actors perceive cyber technology and cyber means, but also provide tools to further securitize them and in this way shape the dynamics of the dual-use security dilemma.

5G Networks and China/Huawei relations with the United States
It is too soon to fully trace the conflictual spiral dynamics between the U.S. government and China/Huawei around 5G technology. Nonetheless, looking into these dynamics demonstrates important aspects of the suggested framework. Some scholars trace the start of these dynamics to the Chinese government's attempts to securitize internet access and information security to protect the regime (Creemers, 2020, p. 130;Hansen & Nissenbaum, 2009, p. 1157Miao et al., 2020). In this context, in the 2010s, Chinese discourse "denounced the American hegemony in global cyberspace and surveillance … framing the USA as the world's largest cybersecurity threat" (Miao et al., 2020, p. 10). A main worry of China, aggravated by Snowden's leaks, was that it relied on U.S. dual-use technologies (Creemers, 2020, p. 108, 116). The securitizing moves of the U.S. threat emphasized the "urgent necessity of strengthening China's cybersecurity capacity to cope with the global risk" (Miao et al., 2020, p. 10). Various documents and speeches of top Chinese leaders echoed these securitizing moves to advance China's interests and cyber sovereignty (see in Creemers, 2020, p. 108, 113-114). 11 China's response was to advance self-reliance by substituting foreign suppliers with Chinese ones, such as Huawei. This move promised not only economic benefits but also political advantages (Creemers, 2020, p. 116).
However, these Chinese moves fueled the spiral of insecurity, led to counter-securitizing moves, and contributed to the deterioration of the relations between China and the United States (Creemers, 2020, p. 117). Among other things, network technology has been securitized in a number of official reports in the United States, with warnings against acquiring Chinese 5G equipment given Huawei's ties with the Chinese government (in Frieden, 2020, p. 102). The Trump Administration declared a formal "national emergency" with respect to the dual-use communication technology: foreign adversaries' ability to exploit vulnerabilities "with potentially catastrophic effects, and thereby constitutes an unusual and extraordinary threat to the national security, foreign policy, and economy" (White House, 2019). Based on this securitizing move, the U.S. government worked to limit the geo-economic space of Huawei in the United States and in other Western countries (Cartwright, 2020, p. 12).
While the U.S. administrations had some reasons to issue these warnings, experts have cast doubts over whether Huawei and the Chinese 5G equipment pose an existential threat and have asked whether excluding Huawei from American networks serves American interests (e.g., Rutkowski, 2020). Furthermore, as Inkster (2019, p. 108) notes, China's existential threat to the United States lies in its challenge to US technological dominance (Inkster, 2019, p. 108;Johnson, 2019, p. 19). 12 Huawei "potentially threatens US national security by enabling the same sort of surveillance programmes that the US is known to engage in" with its increasing global focus, thus also helping to exert Chinese power (Cartwright, 2020, p. 3, 9). Likewise, Johnson asserts that the U.S. perception of Chinese revisionist goals is fueled by its acquisition of AI related dual-use technologies-regardless of whether these perceptions are accurate (2019, p. 14). In fact, other countries were much slower to ban Huawei equipment despite US pressure (Baker & Chalmers, 2020), and thus avoided securitizing it. 13 In other words, not all agreed on whether Huawei poses an existential threat, which points to the social aspects and its non-deterministic characteristics.
The challenges to Huawei and China of United States securitizing moves and extraordinary measures were not just financial. They should also be understood in the context of the key role corporations such as Huawei (and ZTE) play in China's attempts to dominate information technology by providing intelligence-collection opportunities, as well as in China's aspirations to become a global power by supporting its cyber sovereignty (Cartwright, 2020, p. 11;Inkster, 2019, p. 108;Wang, 2020, pp. 288-289). These challenges also resulted in counter-securitizing moves evident in extreme narratives of threat within Huawei, seen, for example, in comparisons with the phoenix-of being burnt and reborn out of the ashes. American pressure on the corporation has been described as "a bird being tested by flames of international politics" (Wang, 2020, p. 288). This analogy is especially interesting in the broader context of the close ties between the company and the Chinese government, and thus also the Chinese perceptions and narratives.
The U.S. sanctions imposed on Huawei, as well as on ZTE, for violating intellectual property rights (and for violating sanctions against Iran) contributed to decoupling U.S. and Chinese cyber technologies. These in turn "fostered a greater sense of urgency in Beijing to enhance resilience, autonomy, and self-reliance" (Creemers, 2020, p. 118). As a result, Huawei secured alternative suppliers (Cartwright, 2020, p. 11;Wang, 2020, pp. 288-289). Furthermore, it also developed technologies that would substitute the American ones, such as launching an operating system that replaces Google Android (Creemers, 2020, p. 129).
The Chinese responses created economic challenges for the United States evident in the threats of Chinese sanctioning of companies that boycott Chinese products and in the loss of potential financial revenues (Creemers, 2020, p. 127). But the Chinese response and its reliance on alternatives to Western companies also threatens the U.S.'s technological aspiration for dominance. This is because such dominance relies on U.S. internet firms, which are "useful for the US in internationalising its power through surveillance programmes" (Cartwright, 2020, pp. 8-9). The growing share of Chinese competitors not only decreases U.S. dominance, but complicates its ability to harvest sensitive information (Cartwright, 2020, p. 8). While China's attempt to create a geo-economic space that helps its companies develop market dominance started before the limitations imposed by the Trump administration, the involvement of Huawei in 5G technology and U.S. responses accelerated and aggravated these dynamics. They resulted in concrete limitations on how Huawei can perform and it affected its supply chain as discussed above (Cartwright, 2020, p. 10, 12). In other words, both sides not only responded to the securitizing moves of each other, but continually re-securitized network security, negatively affecting their relations.

Conclusion
This article put forward the concept of the dual-use security dilemma. Actors experiencing this dilemma are uncertain not only whether an opponent intends to use its capabilities for offensive or defensive purposes, but also whether an opponent's technology is intended for civil or military purposes. Lookingatthesedynamicsfrombothatraditionalandaconstructivistperspective of security clarifies the challenges they pose for actors in the international arena.
I suggest that the dual-use security dilemma, with its focus on the social constructions of insecurity and their international effects, offers additional promising directions to study securitization and also carries further policy implications.

Securitization scholarship
The dual-use security dilemma provides a useful framework to elaborate on the emerging scholarship of cyber domain securitization (Dunn Cavelty, 2008;Eriksson, 2001;Hansen & Nissenbaum, 2009; see also Kallender & Hughes, 2017;Lacy & Prince, 2018). It allows the study of specific securitizing plots, in different security sectors, of different kinds of actors-states and non-state actors-as well as the relations between them.
Furthermore, studying how technological challenges are constructed as threats and initiate spiraling dynamics of insecurity is important for securitization scholarship, which has begun to grasp the significant implications of one actor's securitizing moves for other (international) actors (Lupovici, 2019;Van Rythoven, 2020;Waever, 2015, pp. 122-123). They may be seen as threats; but enunciators may also see them as an opportunity to pursue their own political goals. Focusing on the security dilemma from a securitization perspective provides scholars a direction to theorize the dynamics of securitizing, counter-securitizing, and resecuritizing moves.

Policy implications
This research carries a number of non-intuitive policy implications. First, when an actor emphasizes the civil or peaceful applications of a certain military/harmful technology, this does not necessarily decrease an opponent's fears. Not only does the actor face trust issues over whether such peaceful or civil programs are genuine, but, more importantly, from the perspective of the dual-use security dilemma, adding civil or peaceful applications to an existing military technology magnifies the uncertainty. In such situations, the new applications exacerbate the actor's uncertainty over how the technology will be employed. Using the lens of securitization framework to consider these dynamics emphasizes rival enunciators' ability to socially construct the insecurity derived from the uncertainties caused by dual-use technologies.
Another policy implication is that disarmament provides only a limited solution to the dual-use security dilemma. It cannot on its own prevent an enemy from transforming civil infrastructures or knowledge into military/ harmful means. In this respect, even if actors give up or decrease their military/harmful means, an actor still worry that its opponent will at some point in the future divert its civil technology to military/harmful use or take advantage of an actor's potential military/harmful capabilities by securitizing them and thus justify taking extraordinary (counter) measures.
From this perspective, arms control mechanisms, either combined with disarmament or on their own, are a much more effective way to increase actors' security and limit securitizing moves. While these cannot fully prevent actors from securitizing other actors' technologies, they may limit their effectiveness and provide tools for other actors to defend and respond to such attempts. Policy makers should, for this reason, be more aware of this dynamic. While they acknowledge the traditional security dilemma, they also must recognize the specific dynamics of the dual-use security dilemma and how the social construction of insecurity affects other actors. Some politicians are well aware of these dynamics and take advantage of them to advance their domestic political goals. However, those who would prefer to avoid them must carefully consider not only how they frame their own insecurities, but also what their opponents aim to achieve when they frame their own insecurities.
In fact, an important contribution of securitization theory is consideration of the security dilemma and the dual-use security dilemma. The framing of security is always a choice-and therefore always a concern of policy makers (Buzan et al., 1998, p. 32, 34;Van Rythoven, 2020, p. 479). As Van Rythoven (2020) asserts, "policymakers are not only responsible for the effects of security claims they anticipate, but also need to accept responsibility for unintended consequences of their actions" (p. 479). Both the intentional and the unintentional outcomes are part of the dynamics of the security dilemma and the dual-use security dilemma, and these are aggravated by securitizing moves and opponents' responses to them. In this respect, dual-use technologies pose a special challenge. When enunciators securitize them, they not only limit the social applications of the technology (and economic gains), but also fuel a dangerous dynamic that worsens the situation of rival actors.  (2004), I acknowledge that dual use consists not only of military applications but also of harmful applications. 2. This dynamic is especially prominent in the development of artificial intelligence technology. Even a small advantage in the decision-making abilities provided by this technology can be translated into a disproportionate strategic advantage (Payne, 2018, pp. 23-24) 3. It should also be emphasized that such an arms race between two actors may affect the security of additional actors dragging them into the race. 4. It should be noted that in addition to Van Rythoven, Watson (2013) and Olesker (2018) also study the concept of the securitization dilemma. However, while Van Rythoven uses it as a way to develop the research of the traditional security dilemma, Watson and Olesker are interested in a different dynamic where a securitizing move of one issue leads actors to face insecurity in another sector (Watson, 2013, p. 66), or may result in loss of legitimacy (Olesker, 2018, pp. 316-317). In this respect, Van Rythoven refers to a more international dilemma, a dilemma where two international actors (two different referent objects) face and gradually affect each other's security, and not an internal dilemma between two different values an actor holds (see also Van Rythoven, 2020, p. 479). 5. Van Rythoven convincingly shows how studies on securitization and on the security dilemma overlap in how they broaden the concept of security and acknowledge uncertainty, tragedy, and social elements (Van Rythoven, 2020, pp. 483-490). 6. The resistant actor may attempt to de-securitize the situation (e.g., in Vuori, 2008, p. 93); delegitimize it, or securitize it (Stritzel & Chang, 2015, p. 553). 7. This is somewhat different from Tannenwald's (2007) seminal work on the nuclear taboo. For her, the nuclear taboo is on the first use of a nuclear strike. 8. These securitizing moves can target both public audiences and private audiences within the state (e.g., elite, cabinet) (Roe, 2008, pp. 624-632;Salter, 2008, pp. 329-330, 334-336) 9. It should be noted that these three types of narratives do not cover all the potential plots for securitizing moves regarding opponents' dual-use technologies, but rather illustrate key ways to target audience support. 10. See for example, Lachow's (1995) discussion of the American GPS dilemma. 11. Chinese view of cyber sovereignty is not clearly defined but concerns both domestic governmental control over the domain, and international non-intervention in it (Creemers, 2020, p. 109;Miao et al., 2020, p. 11). 12. This perception also led the Pentagon to author several programs that aimed to protect US dominance (Johnson, 2019, p. 19) 13. Germany, for example, decided to apply the same security standards to all vendors (Rinke & Busvine, 2020), and thus avoided from specifically securitizing China or Huawei.