Introduction
Captivity is a complex phenomenon in international politics with a broad range of purposes, functions, and consequences. Existing research suggests that in holding individuals or groups of individuals captive, states and other political actors may, for example, seek to facilitate hostage or prisoner exchanges, or extract material rewards.Footnote 1 In addition to these reasons, scholarship on ‘human shields’ suggests that states, and other political actors, may also hold individuals captive for deterrence purposes. This practice shows that captivity may be a way in which political actors seek to deter other states from attacking them, particularly in the context of military conflict. The use of human shields is a rather traditional form of deterrence: actor A seeks to deter actor B from militarily attacking actor A out of fear either that the attack will kill the hostages used as human shields, or that A will kill the hostages in the event that actor B launches an attack.Footnote 2
This article argues that captivity may be used for deterrence purposes not only through the use of human shields and not only to deter traditional, military security threats emanating from other states. Building on research that has expanded the study of deterrence beyond the realm of traditional physical military security,Footnote 3 the article contends that captivity may be used by states to deter perceived threats to regime security, and that emotions are central to this process. In such cases, the targets of the deterrence efforts are not other states but rather individuals who are seen as threatening regime security through their activities and the narratives they spread. In this way, the kinds of captivity focused on differ from cases where individuals are detained to seek concessions from their home states, as in hostage diplomacy.
Earlier research has suggested that the emotion of fear is key to successful deterrence. Indeed, fear, or terror, is central to the etymology of the word ‘deterrence’, which can thus be broadly defined as preventing someone from action through fear of consequences. However, this article contends – in accordance with this special issue’s focus on captivity passions – that paying attention to other emotions than fear can help better understand deterrence dynamics, including deterrence failure. Specifically, because captivity, and deterrence, involves the denial of the captive’s agency and may trigger feelings of humiliation and shame, and related ontological insecurity, it can backfire as the target of deterrence might seek to act to regain agency.Footnote 4 Both captivity and deterrence are closely linked to agency. Captivity quite literally denies the target agency by keeping it detained. Successful deterrence also limits the target’s agency as it by definition means that the target refrains from some kind of behaviour or action. The article’s arguments are illustrated through empirical vignettes that show how China has sought to deter perceived threats to the Chinese regime, or party-state, in the form of foreign scholars, publishers, and NGO workers by holding such individuals captive.
The next section reviews scholarship on captivity in international politics. The section thereafter theoretically links deterrence to captivity and emotions. The sections that then follow provide empirical illustrations on China’s practice of detaining foreign scholars, publishers, and NGO workers. The final section presents the paper’s conclusions on the phenomenon of deterrence through captivity along with implications for future research.
Captivity in international politics
Political, or politicised, captivity is a common phenomenon in world politics. However, considering how common the practice is, the literature is nonetheless relatively limited and rarely treats politicised captivity as a single, broad phenomenon. Instead, a range of literatures cover related topics, such as terrorist kidnappings, hostage diplomacy, historical work on hostage taking, and the practice of using human shields.
While some scholars have explored how domestic actors in the state whose citizens have been held captive have sought to use kidnappings instrumentally for political objectives,Footnote 5 in much research on politicised captivity, questions related to why states and other actors hold individuals captive, and the functions and consequences of such practices, have occupied a central place. Historical scholarship has explored extensively the political role captivity has played throughout history, for example in antiquity and the Middle Ages when hostages were not just taken, but often given, for various reasons, for example as collateral to make sure that agreements and treaties were honoured.Footnote 6 In Arab history, the kidnapping of individuals from other tribes was a common feature of tribal conflict. In this context, abductions were often used to put pressure on enemies.Footnote 7
Much contemporary research has focused on kidnappings carried out by terrorists,Footnote 8 or the use by authoritarian states of ‘hostage diplomacy’, understood as the ‘taking of hostages under the guise of law for use as foreign policy leverage’.Footnote 9 Walt, for example, discusses in a brief article the question of why authoritarian states, such as China and Iran, engage in hostage taking even though it risks making them look cruel and could thereby undermine other foreign policy objectives. He provides five possible reasons. First, some captives may actually be guilty of what they are accused of. Second, there may be genuine differences in opinion concerning what is considered to be espionage or threats to national security. Authoritarian states that hold individuals captive may adopt a broader understanding of these categories. Third, individuals may be held captive because a citizen of the authoritarian state has been detained on criminal charges in the targeted state. Fourth, hostages may be taken as leverage to be used to effectuate a prisoner exchange or for broader leverage. Fifth, detention of foreigners may happen as part of domestic factional disputes where hard-liners seek to spoil efforts at improving bilateral relations. However, while Walt thus finds explanations for such arbitrary detention, he nonetheless concludes that he is ‘convinced that the costs of this kind of behavior usually outweigh the benefits’, and that in most cases, it ‘yields nothing more than reams of bad publicity’, thereby reaffirming his initial puzzlement.Footnote 10 The high-profile ‘Two Michaels’ and Wang Mengzhou case, which led to a prisoner swap between China and Canada, seems to support the suggestion that authoritarian states may detain foreign citizens in order to have their own citizens released in prisoner swaps.Footnote 11 The June 2024 prisoner swap between Sweden and Iran also exemplifies this pattern.Footnote 12 The same goes for the August 2024 prisoner deal through which eight Russians were swapped for sixteen individuals who were relocated from Russia to the United States and Germany.Footnote 13 These high-profile cases indicate that some authoritarian states have made it part of their coercive toolbox to use the detention of foreign nationals to bring about prisoner swaps. However, there are also cases where authoritarian states have taken captives seemingly without making any demands for prisoner swaps or other demands directed at the captive’s home country. In other words, there are cases where the motivations are less apparent. This article suggests that in some such cases, a deterrence dynamic operates.
The question of how captivity might be linked to deterrence has been given some, although relatively limited, attention. There is, for example, some research on how kidnappings, by terrorists, might be deterred.Footnote 14 In recent years, there has also been much discussion among policymakers and scholars concerning what measures might be taken to deter hostage taking.Footnote 15 It has, for example, been argued that in order to develop deterrence measures, it is necessary to understand the specific motivations that particular states might have for taking hostages.Footnote 16 There is also research on how middle powers might respond to hostage diplomacy by authoritarian states, including the possibilities for deterring such hostage taking through punishment.Footnote 17
In addition to debates on how to deter hostage taking, there is also some work on how captors use captivity for deterrence purposes. Gilbert, for example, focuses on armed groups engaged in civil war in Colombia who tax the local population. When locals do not pay taxes, the groups use kidnapping for ransom to punish them, but also to deter locals more generally from future tax evasion.Footnote 18 These groups, then, use deterrence through captivity for enforcement purposes, rather than to deter security threats. There is also scholarship that discusses whether and to what extent the use of human shields can function as a deterrent in war.Footnote 19 As Kinsella points out: ‘All forms of human shielding are informed by the same proposition: shielding is presumed to deter the opposing forces from exercising violence, or at the very least, to moderate that violence’.Footnote 20 However, the opposing forces may want to show that they are not deterred and thus still attack in order to discourage future would-be hostage-takers from using human shields.Footnote 21 The kind of deterrence that research focusing on the use of human shields is concerned with is traditional in the sense that it is related to deterring military attacks against traditional referent objects of security.
Tsukahira’s historical study of the Japanese Sankin kōtai (alternate attendance) system in Tokugawa Japan (1603–1868) also illustrates the use of captivity for deterrence purposes. Sankin kōtai was a system of political control whereby most of the more than 260 feudal lords (daimyō) had to spend alternate years in attendance in the capital Edo at the shōgun’s court. The families of the daimyō also had to reside permanently in Edo. Their presence in the capital rather than in the provinces functioned as a form of deterrence through political captivity. This institutionalised deterrence was a feature of the political order during the Tokugawa era. As Tsukahira puts it: ‘The enforced residence of their families in Edo acted as a deterrent to any attempt at rebellion, while the costly journeys to and from Edo and the expense of maintaining two establishments, one in the metropolis and the other in the domain, kept the lords financially, and hence militarily, weak’.Footnote 22 In the Tokugawa case, captivity was used to deter threats to regime security in the form of potential rebellion.
The present paper, while building on existing research, seeks to further extend the focus on captivity for deterrence purposes to non-traditional, non-military security issues. In doing so, it contributes to both the disparate literatures on political, or politicised, captivity discussed in this section and also research on deterrence, which is discussed in the next section.
Linking captivity and deterrence
After the end of World War II and with the advent of the nuclear age, deterrence became a central strategic concept. The centrality of deterrence in the practice of international politics was matched by a strong interest within the academic disciplines of international relations and security studies.Footnote 23 Broadly understood, deterrence involves preventing someone from action through fear of consequences.Footnote 24 As Morgan puts it, ‘The essence of deterrence is that one party prevents another from doing something the first party does not want by threatening to harm the other party seriously if it does’.Footnote 25 While this article focuses specifically on deterrence, there are closely related concepts, such as coercion and compellence. Schelling describes deterrence and compellence as belonging to the broader category of coercion. Both rely on the issuance of a threat, but while deterrence is about making someone refrain from doing something in the future out of fear of consequences, compellence involves making the target do something, or stop doing something, often by administering the punishment until the target stops.Footnote 26 Similarly, Markwica distinguishes deterrent threats, that is, threats issued to dissuade an actor from future behaviour, from coercive threats, which are issued to stop the target’s past or ongoing behaviour.Footnote 27 Deterrence and compellence, or coercive threats, sometimes overlap in practice, for example when the deterrer coerces the target to stop behaviour in the present with the long-term aim of dissuading the target from repeating the behaviour in the future.Footnote 28 This article focuses on deterrence, rather than compellence, since it is concerned primarily with effects on future behaviour.
For a long time, deterrence research focused on interstate deterrence in the context of nuclear strategy and conventional military conflict. In the last few decades, however, scholars have broadened the focus to include deterrence of asymmetrical threats, including non-state actors and issues such as terrorism as well as cyber and ethnic conflict.Footnote 29 Even more recently, scholarship has begun to appear on how democracies might be able to deter threats such as election interference, disinformation, and other kinds of hybrid warfare that threaten democracy.Footnote 30 The existence of such research suggests that just as the concept of security was broadened in the 1990s, deterrence has now also come to be broadened to an extent that goes far beyond its narrow understanding during the Cold War.Footnote 31 Recent scholarship has argued that actors wishing to deter may do so not only to advance physical security but also to confirm their identities as deterrers,Footnote 32 and to deter threats to their ontological security, or security of the self.Footnote 33 Relatedly, it has been shown that states such as China have sought to deter third-party states from criticising them for human rights violations by issuing threats and using diplomatic and economic sanctions.Footnote 34
The moves discussed above open up the deterrence research agenda to not only new referent objects of security but also a focus on new measures through which political actors seek to deter. Different referent objects may involve different repertoires of deterrence measures and different actors may have different deterrence repertoires at their disposal, due for example to differences when it comes to legal constraints in different states. Captivity, this article contends, is one such measure. If political actors can be assumed to use a variety of measures in order to deter threats to a variety of referent objects, one such measure may be the holding of individuals captive in order to dissuade those held captive, as well as other individuals, from carrying out undesired actions. When, in the context of a war, one of the parties uses captured individuals who belong to the other party as human shields, it is to deter the other party from carrying out military acts out of fear that those held as human shields might be killed in the attack. In such situations, we are concerned with an attempt to deter a traditional, military threat posed by another state.
However, captivity may also, as suggested above, be used to deter other, non-traditional threats, such as threats to regime security. In this context, captivity may function as part of state repression more broadly. In addition to captivity, such repression may encompass surveillance, harassment, bans, torture, and other forms of violence for the purposes of coercing and deterring actual and potential challengers to the regime from engaging in activities associated for example with freedom of speech and freedom of assembly.Footnote 35 In the case of China, context-sensitive scholarship on local understandings of ‘security’ has shown that Chinese elite understandings of security have tended to view the referent object of security as not the state but rather the Chinese party-state, or the regime.Footnote 36 Such an understanding of security as regime security is likely to be more common in non-democratic than in democratic states. Non-democratic regimes may be sensitive to activities by individuals, for example domestic and foreign civil society activists, academics, journalists, and publishers who construct narratives that might challenge and threaten the stories from which the deterrer state derives its legitimacy and ontological security.Footnote 37 Ontological security theory suggests that in addition to physical security, states and other actors construct biographical narratives that provide them with a sense of continuity and purpose.Footnote 38 Scholarship has demonstrated that states seek to protect and defend the narratives about themselves that make them ontologically secure.Footnote 39 An authoritarian deterrer state seeks to dissuade challenges to the narrative about the nation that it promotes domestically and thereby increase domestic cohesion and its own legitimacy and regime security.
Both authoritarian and (especially backsliding) democratic states may seek to deter the kinds of threats described above. For example, in democracies with backsliding tendencies both governmental and non-governmental actors may seek to pressure journalists, academics, or civil society groups who bring attention to sensitive issues in ways seen as threatening the national narrative as understood by the deterring actors.Footnote 40 In addition, it has been argued that democratic states have used the detainment of refugees and other immigrants to deter immigration.Footnote 41 However, democracies and authoritarian states arguably have different repertoires of deterrence at their disposal. While democratic and authoritarian states both use captivity as punishment for criminal acts as a way of deterring crime,Footnote 42 democracies are expected to do so within the context of the rule of law where suspects have the right to a fair trial. Authoritarian states, while often paying lip-service to the notion of the rule of law, may, by contrast, more often detain individuals outside the official legal system without pressing charges or based on trumped-up or vague charges. This kind of extra-legal captivity is part of the deterrence repertoires of some authoritarian states but is less likely to be part of the repertoires of democratic states, unless they are backsliding.Footnote 43
Non-democratic states may thus seek to use captivity to punish, compel, and deter those who offer an alternative to the regime, in terms of either alternative narratives about specific events, a more general alternative understanding of reality, an alternative national biographical narrative, and by extension ultimately an alternative political order. If through their activities, they offer or contribute to offering such alternatives, scholars, publishers, journalists, human rights activists, and other civil society actors may be targeted through detention or the threat thereof. Specific individuals belonging to a certain category may be detained both to punish that particular individual and to make them refrain from similar behaviour in the future, and to deter other individuals who are or might be engaged in such activities. In this way, deterrence functions by setting examples that can have self-disciplining effects on others than those directly held captive as the fear of being detained spreads among those who belong to the same category as the captive. From the deterring agent’s perspective, captivity is one way in which it seeks to instil fear in would-be challengers. Similar to the Colombian armed rebels who kidnap locals who avoid paying taxes to the armed groups, states may use captivity not only to retrospectively compel someone from behaviour seen as threatening to the state but also prospectively to deter such behaviour among a broader population.Footnote 44
One advantage, from the deterring actor’s perspective, of combining compellence and deterrence is that it increases credibility since the threat is occasionally carried out. Credibility has been a key concern in research on deterrence. Simply put, the fact that a deterring actor has the ability to do harm is not sufficient for it to successfully practice deterrence. For deterrence to work, other actors need to find the deterring actor’s threats credible. They need to believe that the deterring actor is both able and willing to do harm.Footnote 45 Since the threat is occasionally carried out, credibility is less of a problem for China’s deterrence through captivity of threats to regime security than it is for nuclear deterrence. If in the context of nuclear deterrence, a deterring actor carries out the threat, the result might be mutual destruction. When it comes to threats to regime security and the use of captivity to deter such threats, occasionally carrying out the threat by punishing some individuals shows other individuals belonging to the same category that the deterring actor is indeed willing to carry it out and thus that the threat is credible.
Credibility, in the context of deterrence, also depends on communication. If the target does not know of the deterrer’s willingness to carry out the threat, deterrence is less likely to be successful.Footnote 46 When it comes to deterrence through captivity, the intent to punish not only the particular individuals who are detained but also those belonging to the same category as that individual seems to be present especially when the detention of such individuals is widely publicised, for example through domestically and internationally televised forced confessions where the captives are put on display and made to confess their ‘crimes’. Such spectacles, used frequently by some authoritarian states such as China, guarantee that when, for example, civil society activists, academics, or publishers are detained, other civil society activists, academics, or publishers will become aware of it and possibly refrain from the kind of behaviour that the deterrer wishes to dissuade. Analytically, such communication can be treated as an indicator of intent to deter. If the objective were only to compel the detained individual to stop their activities at a particular point in time, rather than to deter both that person’s future behaviour, as well as a broader category of individuals more generally, there would be less of a point in broadly disseminating information about the detention.
Rather than explicitly outlining its emotional underpinnings, scholarship on deterrence traditionally tended to rely on assumptions according to which it was practiced by actors engaged in unemotional cost–benefit calculations. As Lupovici aptly puts it, it was assumed that such actors ‘would avoid challenging their opponents (the deterrer actor) if the costs of such attacks are higher than the gain they can achieve’.Footnote 47 In such understandings of deterrence, the role of emotions was limited. This is despite the fact that the etymology of deterrence makes it clear that fear, or terror, is central to its original meaning. It comes from the Latin word deterrēre, which consists of de-, meaning ‘from’ or ‘away’, and terrēre, a verb meaning ‘frighten’ but that is also the root of the more similar-sounding word ‘terror’. Deterrēre, then, means ‘to frighten from’.Footnote 48 However, even when its emotional foundations are referenced, as in definitions of deterrence as ‘the prevention from action by fear of the consequences’,Footnote 49 the precise role played by fear, let alone other emotions, was often not clarified.Footnote 50 More recent work, however, has begun to explore the roles of not only fear but also a range of other emotions in deterrence.Footnote 51 Lupovici, for example, shows how some states construct their identities as deterrers. Having a deterrer identity then means that when deterrence is successful it validates the deterrer’s identity, giving rise to positive emotions such as pride and a sense of ontological security. Conversely, when deterrence fails, such actors may instead become ontologically insecure and feel fear, frustration, humiliation, shame, nostalgia, and anxiety. Moreover, in constructing its own identity as a deterrer, such a state will do so in relation to an ‘other’ who is ascribed an identity as deterred.Footnote 52 Lupovici, however, is primarily concerned with the deterrer’s emotions, rather than with the emotions of those who are deterred or targeted in attempts to deter. Silincik and Duyvesteyn, however, discuss the emotions of various audiences. In doing so, they conduct a case study of the US assassination of the leader of the Iranian Revolutionary Guard Quds Force, Qasim Suleimani, an assassination that was explicitly said to have been carried out for deterrence purposes. They suggest that conservative Americans appeared to have felt happy following the successful assassination, while Democrats expressed anger, and some commentators expressed fear that Iran might retaliate. In Iran, the article suggests that the assassination triggered anger, as well as fear, but also happiness since it increased societal cohesion.Footnote 53
Even in cases where deterrence appears initially to be successful, this article suggests that there are nonetheless limitations to its strategic use as there is a possibility that it might backfire. As scholars working on coercion more broadly have demonstrated, by focusing on the phenomenon of reactance, when someone is subjected to attempts to limit their freedom of choice, it may trigger anger and reasoning that downplay the credibility of the threat, thus resulting in a desire to do the opposite of what the coercer requests.Footnote 54 Similarly, Markwica describes how the coercer faces a dilemma in having to induce enough fear for targets to alter their behaviour but at the same time avoid triggering the kind of defiant reaction that the coercive measures were meant to prevent in the first place.Footnote 55 While partly overlapping with the insights of such research, the framework proposed here focuses not on coercion but rather specifically on deterrence, and builds on ontological security theory and research on victimisation and agency. In order to understand the workings behind blowback to deterrence attempts, and arguably the larger question of deterrence success and failure, it is useful to take into account the ontological security of the deterred or targeted actor, along with related feelings of humiliation, shame, and anger. In this context, the distinction made in early ontological security scholarship between physical and ontological security,Footnote 56 and between fear and anxiety, is instructive.Footnote 57 Fear can be more intense than anxiety and is linked to specific dangers. Anxiety is more diffuse but often deeper and more closely linked to who we believe ourselves to be, that is, to our identity and self-esteem.Footnote 58 If one is deterred from taking a particular action, it is out of fear of the consequences of taking that action. However, a consequence of being deterred in such a way might be ontological insecurity and anxiety because in being deterred one denies oneself agency and might fail to act in accordance with one’s identity. Actions that are incongruent with one’s identity can cause feelings of shame.Footnote 59 As Steele puts it, ‘Individuals who break laws feel “guilty”, whereas no formalized rule needs to be compromised for shame to be produced’.Footnote 60 Shame, then, is more strongly related to identity than guilt is. In Anthony Giddens’s words, shame ‘bites at the roots of self-esteem’.Footnote 61 In other words, shame might gnaw at you after you have failed to act in accordance with your identity.
In the case of deterrence through captivity that takes the form of arbitrary detention, humiliation is also highly relevant. Humiliation is a strong emotion that results when a person is unjustly devalued or degraded.Footnote 62 Humiliation can result in reducing the humiliated party’s status and produce feelings of shame and lowering of self-respect. The act of humiliating someone may be motivated by contempt or the will to make them obedient and deferent in relation to the party that carries out the humiliating acts.Footnote 63 Key to the understanding of humiliation developed here is how it is related to agency and self-esteem. Research in social psychology has shown that having a strong sense of agency is closely associated with higher self-esteem.Footnote 64 Conversely, as Fernández et al. put it, ‘perceiving ourselves as mere passive victims of a perpetrator, incapable to respond to another’s humiliating act, would make us devaluate ourselves as less human, which would foster the experience of humiliation’.Footnote 65 Being humiliated, then, denies agency and tends to lower the victim’s self-esteem. This can cause feelings of powerlessness and lead to inaction.Footnote 66 For individuals who have been unjustly held captive by an authoritarian regime, the power disparities of the situation may make it difficult to defy the perpetrator while held captive, at least in the long run and as long as the captive hopes to be released. However, once released, the former captive may be able to regain a sense of agency and self-esteem by standing up to the perpetrator. Such agency can be exercised, for example, by speaking out about and narrating the humiliating experience.Footnote 67 As Pemberton, Aarten, and Mulder put it:
The damage of victimisation can be understood as a narrative rupture, which endangers a sense of control and continuity of one’s life story throughout time and with the social surroundings. The way victims attempt to make sense and meaning of their ordeal occurs in narrative modes of reasoning, while they adopt narrative means to regain agency and re-establish communion with their social context.Footnote 68
This suggests that one way in which those denied agency might regain it, and thereby regain a sense of ontological security, is by narrating their experiences. In addition, experimental research has shown that when those subjected to potentially humiliating situations react towards the perpetrator, and thereby exercise agency, they feel significantly less humiliated. Because agency is related to self-esteem, exercising agency by standing up to the aggressor mitigates feelings of humiliation and shame.Footnote 69
In sum, this article suggests that captivity can be used for deterrence purposes in order to deter what the deterrer views as threats to the ontological security narratives that underpin regime security. The empirical examples presented in the next section illustrate the plausibility of the overarching argument that captivity may be used by states to deter perceived threats posed by particular kinds of individuals to regime security, as well as the more specific propositions about how emotions are implicated in deterrence through captivity.Footnote 70 Having three brief illustrative case studies, rather than only one, makes it possible to see whether these are all characterised by similar dynamics. If we can see similar dynamics across cases, it arguably strengthens our reasons to believe that such dynamics are common. In this way, the article shows that states may use captivity to deter not only traditional military threats emanating from other states but also perceived threats to regime security posed by non-state actors, including individuals, and that paying attention to other emotions than fear can help better understand why these efforts are not always entirely successful. In order for the paper’s argument to hold, these illustrations need to show that these captivity cases have been broadly communicated and interpreted as deterrence by relevant audiences. It should also illustrate that deterrence efforts may backfire if the target of the deterrence efforts seeks to regain agency by showing that they are (no longer) deterred. Saying that deterrence can backfire, however, is not to say that it is necessarily unsuccessful. It is, however, a way of clarifying the role of emotions in the process whereby it can be partially successful in deterring some individuals, while still failing to deter others.
Deterrence through captivity in contemporary China
This section presents three empirical vignettes of the Chinese authorities’ use of captivity for deterrence purposes. It suggests that a key function of such detention is deterrence of what are considered threats to national security. The referent object of national security in the Chinese context is the party-state. This means that the objective is to secure the survival of not just the state but also the Chinese Communist Party as a key feature of the Chinese state. This understanding has, as Jonna Nyman pertinently puts it, ‘a profound effect on security policy, enabling a wider and more authoritarian range of policies and control, and shaping a more expansive understanding of threat: which includes citizens who do not support the party-state, in turn enabling the persecution of dissidents’.Footnote 71 An additional aspect of Chinese security politics is arguably that the government also seeks to provide ontological security for Chinese people by managing perceived threats to its preferred national biographical narrative.Footnote 72 The threats that captivity is used to deter are threats to the continuity of the narrative that the Chinese party-state promotes to provide Chinese people with a sense of continuity and purpose, and thereby also shore up its own legitimacy and strengthen regime security. Regime, or party-state, security is thus closely linked to ontological security. Deterrence through captivity is part of the broad range of policies and control used by the authoritarian Chinese state in its search for regime security. The targets of the deterrence efforts are thus not other states, but individuals who belong to a particular category. These individuals’ home states may of course seek to negotiate the release of these individuals, but these states are not the targets of the deterrence efforts. In this way, the kinds of captivity we are concerned with here differ from cases where individuals are detained and concessions are sought from their home states, such as in cases of hostage diplomacy.
Deterring academics
Hokkaido University Professor Iwatani Nobu was detained in China between 8 September and 15 November 2019 under the 2014 espionage law after his hotel room was searched by agents from the Ministry of State Security. Even though he was released after intense diplomatic efforts by the Japanese government, it does not seem as if the Japanese government was targeted. Instead, the target seems to be Iwatani himself, and other academics engaged in similar research. The specific reason for Iwatani’s detention was reportedly that he was in possession of what Chinese authorities described as state secrets. The state secrets he possessed were said to have been a book concerned with his area of research – the Sino–Japanese war of 1937–45 – purchased at a second-hand bookstore.Footnote 73
In recent years, numerous Japanese citizens have been detained in China. According to one source, seventeen Japanese citizens were detained on spy charges between 2015 and April 2023. A number of scholars based in Japan, in particular academics holding Chinese citizenship, have also been detained in China.Footnote 74 However, Iwatani’s detention is said to have had a particularly strong impact on Japanese academics. He was detained even though he had been invited to China by the Institute of Modern History at the Chinese Academy of Social Sciences (CASS). In addition, he was detained even though his research is not related to currently ongoing sensitive issues such as religion or national minorities and their efforts to gain independence. Instead, his research is on modern history, in particular the history of what in China is called the War of Resistance against Japan. The War of Resistance has long been a sensitive topic in Sino–Japanese relations, and the Chinese government derives its legitimacy in part from narratives about the war and has emphasised it as part of patriotic education since the early 1990s. As such, it has long been a securitised topic in the sense that actions or statements by the Japanese government that are seen as denying the official Chinese narrative about the war have been viewed as threats to Chinese collective memory.Footnote 75 In recent years, research on modern history has also become increasingly sensitive. Since Xi Jinping assumed power, the Chinese government has restricted access to historical archives.Footnote 76 It has also increasingly criticised what it labels ‘historical nihilism’, that is, history writing that in any way questions official Chinese Communist Party history, and by extension the Chinese government’s legitimacy. In 2016, a court in Beijing ordered historian Hong Zhenkuai to issue a public apology for having written an article said to have defamed five men who according to official historiography were heroes in the war against Japan.Footnote 77 In 2018, the Chinese government continued such efforts when it adopted the Heroes and Martyrs Protection Law, which criminalises the defamation of Chinese heroes and martyrs.Footnote 78 Against this background, Iwatani’s detention might be understood as an extension of the securitisation of Chinese collective memory to include empirical historical research being conducted by scholars based outside China.Footnote 79
While it is difficult to demonstrate that the Chinese authorities involved in the incident had a clear intent to deter scholars from conducting the kind of research Iwatani was engaged in, the reactions among scholars to this and similar incidents suggest that a consequence of such actions on behalf of Chinese authorities is indeed that many scholars are deterred from visiting and conducting fieldwork in China. For example, in an article commenting on the episode, Kawashima Shin, professor at Tokyo University, commented: ‘Many people [scholars] must have greeted the news of Iwatani’s arrest with the realization that it could happen to them as well’.Footnote 80 He further commented: ‘Chinese researchers specializing in Japan remain able to come to the country and freely pursue their data collection and other research activities here, but it seems likely to become harder for Japan’s China specialists to do the same when going the other direction. Another possibility is that Japanese researchers who want to go to China will need to ensure that their writings and statements align with the positions of the Chinese government and Communist Party of China.’Footnote 81 Others have similarly interpreted the incident as being part of ‘the Chinese Communist Party’s anxious attempts to assert narrative control over the way the country’s modern past is depicted and understood’.Footnote 82 These assessments echo the understanding of Chinese deterrence efforts as attempts to protect the Chinese government’s preferred biographical narrative. Either way, following the detention of Iwatani and other academics in China, several scholars have stated that they have cancelled planned visits to China or are hesitant to visit.Footnote 83 Many academics appear to have been deterred from conducting fieldwork in China because their peers have been detained.
This illustration suggests that captivity is used to compel specific scholars – in this case Iwatani – from research activities at a particular point in time. However, its deterrence function is arguably more significant than its compellence function, as it may deter not only the ongoing activities of the specifically targeted scholar but also the future research activities in China of that specific person as well as those of many other scholars. The risk of detention is associated with uncertainty and anxiety as scholars may or may not be detained when conducting fieldwork or other research activities in China. It might be objected to the account of deterrence through captivity presented here that deterrence requires the communication of a clear and credible threat, but that no such clear credible threat against scholars doing research has been issued. One could, however, regard the fact that scholars such as Iwatani have been detained as evidence that the threat is indeed credible. Even though an explicit threat has not been articulated against all scholars conducting research in China, one might still argue that the fact that it has been carried out against some scholars nonetheless makes it credible. Even though other scholars do not know for certain whether they too will be detained, it nonetheless raises the possibility that they could be targeted. For many scholars, that possibility appears to be sufficient for them to refrain from visiting China.
Deterring publishers
In late 2015, five men involved in the publishing business in Hong Kong – Lam Wing-kee, Gui Minhai, Lee Bo, Cheung Chi-ping, and Lui Bo – all disappeared. It was later revealed that they had all been detained for several months in China. The men were connected to the Mighty Current Media publishing house and its bookstore Causeway Bay Books, which had been acquired by the publishing house in 2014. Three of the men disappeared in mainland China, while Lee Bo, who has British citizenship, is said to have been spirited away from Hong Kong into mainland China, while one of them, Swedish citizen Gui Minhai, disappeared from his holiday home in Thailand. While the Swedish government has made efforts to have Gui released, his detention does not appear to be an effort to extract concessions from Sweden. The target seems to be Gui himself and other publishers. The fact that Gui was abducted from his holiday home in Thailand might be interpreted as signalling to other publishers that they too might be targeted even outside China’s borders.Footnote 84 Gui Minhai first appeared in a televised confession in January 2016 and then again in late February of the same year. At this point, the other men, with the exception of Lee Bo, also appeared in televised confessions, believed to have been forced. The televised confessions portrayed Gui as the ringleader behind the operation with the other confessors acting as supporting witnesses.Footnote 85 In 2020, Gui was sentenced to ten years in prison for ‘providing intelligence’ overseas.Footnote 86 Keeping some publishers, like Gui, in captivity for an extended period might be a way of communicating that there is a possibility that others might also be detained for not only short but also longer periods.
Mighty Current Media has published sensationalist and critical books, many of which are banned in mainland China. The books cover topics such as China’s political leaders, including their sex lives, corruption, and other controversial issues. One book is titled ‘Xi Jinping and his lovers’. Another book lists taboo topics forbidden in China, such as ‘press freedom’ and ‘civil society’.Footnote 87 According to one description of the books, they ‘mix rumour, speculation and outright fiction, spin stories about China’s elites’.Footnote 88 While the authenticity of some accounts put forward in the books may be questionable and difficult to confirm as they rely on anonymous sources, the high degree of censorship and tight hold on information in mainland China, as well as the often highly idealised portrayals of China’s political leaders, made the books popular among mainland Chinese. Causeway Bay Books, and other similar bookstores, thus received large numbers of mainland Chinese customers interested in gossip and speculation about top Chinese politicians.Footnote 89 One of the detainees, Lam Wing-kee, who had been detained in late October 2015, relates that while he was detained, he was told he had violated Chinese law by sending and delivering forbidden books to mainland China. During the interrogations, he was asked who had bought the books as well as who their anonymous authors were. In June 2016, he was finally allowed to return to Hong Kong, under the condition that he would fetch information about Mighty Current Media’s clients and authors and deliver that information to China. He would then get to continue to work at the bookstore, but only if he would report customers to the Chinese investigators. Lam initially agreed; however, a few days after his return to Hong Kong, he instead contacted a local politician who organised a press conference during which Lam spoke at length about his detention.Footnote 90 The press conference might be understood as an act through which Lam, after having had his agency denied during detention, regained agency and showed that he was not deterred. At the same time, it is clear that other publishers and owners of bookstores had indeed been deterred. As one article comments:
At the time of Lam’s abduction, banned books were everywhere in Hong Kong, sold throughout the city at big-box retailers, specialised cafes and corner convenience stores. Within days of his disappearance, they began to vanish, swept off shelves by mainland-owned shops and frightened independent booksellers. Authors were cowed into silence; presses refused to print sensitive material.Footnote 91
Similarly, there have been reports that in the months following the disappearances of the publishers and booksellers, people in Hong Kong involved in this industry have been deterred. Numerous bookshops have been closed down, publishers have stopped publishing political books, while printers have stopped printing them. As one publisher summarised it: ‘Everybody is scared’.Footnote 92
The narratives that Mighty Current Media published differ from the historical research published by scholars such as Iwatani in many respects, but arguably constitute a similar threat to the regime security of the Chinese party-state.
Deterring non-governmental organisations
Peter Dahlin, a Swedish citizen, worked for Chinese Urgent Action Working Group (China Action), an NGO that trained human rights lawyers and provided legal aid. He was detained between 3 and 26 January 2016 on several criminal charges, including foreign financing of illegal and subversive activities. The Swedish government worked hard to negotiate Dahlin’s release; however, there is no evidence suggesting that he was detained in order to target the Swedish government. Dahlin’s detention coincided with that of the Hong Kong publishers and booksellers, and took place against the background of a broader Chinese crackdown under Xi Jinping to strengthen authoritarian control by reigning in the media, limiting dissent, and ousting political opponents. More specifically, Dahlin’s detention was preceded by a clampdown in mid-2015, whereby a large number of human rights lawyers, including some who were linked to China Action, were detained. While such efforts by the Chinese government have primarily taken aim at domestic civil society, Dahlin was one of the first foreigners to be targeted.Footnote 93 As mentioned above, the publisher Gui Minhai is, of course, also a Swedish citizen, but the Chinese government has made it clear that it regards Gui as ‘a Chinese national first and foremost’.Footnote 94 The Chinese government thus views the two cases differently. Dahlin’s case, like Iwatani’s, suggests that the Chinese government regards domestic and foreign threats to the regime as interlinked and that it seeks to deter not only Chinese citizens seen as threats to the regime, but foreigners as well.
Dahlin’s arrest has been interpreted as not only a way of compelling Dahlin himself to stop his NGO work in China but also a more general deterrence effort against foreign NGOs active in China. As one charity director commented: ‘his arrest is a warning to all of us’.Footnote 95 Legal scholar Jerome A Cohen commented: ‘The authorities made their point, spreading intimidation and fear throughout both the domestic and foreign legal and NGO worlds’.Footnote 96 In April 2016, only a few months after Dahlin’s detention, China passed a new law on foreign NGOs that significantly limits the possibilities for such entities to operate independently in China.Footnote 97
On 19 January, a televised confession appeared on China Central Television (CCTV) in which Dahlin said he had ‘caused harm to the Chinese government’ and ‘hurt the feelings of the Chinese people’, a standard phrase often used by the Chinese government.Footnote 98 Dahlin himself has stated that the confession was forced. According to a report published by the NGO Safeguard Defenders, at least forty-five high-profile televised confessions were broadcasted between 2013 and 2018.Footnote 99 These televised confessions reach far beyond China’s borders since China Central Television channels are broadcasted elsewhere in the world as well, including in Europe.Footnote 100 Dahlin later commented that: ‘The point of these televised confessions is not only to deter the individual but to attack a larger group of people’.Footnote 101 He also stated that ‘they had used it to scaremonger foreigners, foreign NGOs, and their supporters’.Footnote 102 A human rights activist named Li, who was also detained and pressured to confess on camera, stated: ‘These TV confessions intimidate public intellectuals, they make everyone feel insecure, censor themselves, to never dare say anything or do anything against the Party. It’s a white terror.’Footnote 103 Others who have been detained have described feeling ‘humiliated’ and ‘traumatised’ by the experience. One person has said he feels ashamed about his confessions and that it made him angry.Footnote 104 Others describe the televised confessions as a ‘shame parade’,Footnote 105 and another detainee states that: ‘The Party wants the detainee to incriminate themselves, to humiliate the detainee and to make them look bad in front of the Chinese people, so they stop caring about the people who are detained. The detainees will lose their only source of moral support, and ultimately, they can only end up being destroyed by the Party.’Footnote 106 Peter Dahlin also describes having suffered from what he feared might be post-traumatic stress in the period after he had been released. He describes feelings of anxiety and not being able to relax properly, as well as getting angry when not being able to control things.Footnote 107
When successful, deterrence leads those targeted – whether those directly targeted or those to whom deterrence measures are meant to send signals – to refrain from doing what the deterrer considers threatening. However, targets do not always comply. They may also resist. Deterrence can backfire, leading the target of deterrence measures to double down on their actions or perform other actions in defiance of the deterrer in order to regain agency. While this seems to have been the case for Lam Wing-kee, the bookseller, Dahlin’s case similarly illustrates this point. After having been released and deported from China, Dahlin has repeatedly recounted his experiences as a captive. His narrative about being detained demonstrates that even when he was in captivity, he exercised agency. For example, when recounting how he was subjected to a lie detector test, he emphasised how he was still able to stick to his story: ‘I had defeated a lie detector. I couldn’t hold back the feeling that by extension I had defeated them.’Footnote 108 Several similar examples of how he exercised agency while being detained appear in his narrative. He mentions, for example, how he refused labelling his associates ‘criminals’ when his interrogators demanded he do so.Footnote 109 This illustrates how narrative can function as a means through which those who have been victimised and thereby had their agency denied can reclaim agency by emphasising in the narrative how one, even in a situation where one’s agency was clearly severely restricted, nonetheless resisted and exercised some agency.Footnote 110
In addition to narrating his experiences, following his release, Dahlin founded the NGO Safeguard Defenders. The NGO has published the book The People’s Republic of the Disappeared, which contains first-person accounts of victims of arbitrary detainment in China,Footnote 111 as well as the report Scripted and Staged: Behind the Scenes of China’s Forced TV Confessions,Footnote 112 which also contains the testimony of a number of victims of China’s televised forced confessions. Through such publications and other efforts, Safeguard Defenders have urged governments in other countries to take action against Chinese state-run media outlets that have broadcasted such confessions in those countries. For example, in 2023 Safeguard Defenders submitted a complaint to the French TV regulator ARCOM regarding CCTV’s broadcasts through the French satellite operator Eutelsat.Footnote 113 The testimony given by these victims, as well as Dahlin’s founding of Safeguard Defenders, can be regarded as a way of regaining the agency denied through captivity and forced confessions. As such, while China’s deterrence through captivity may be able to dissuade some of those it has targeted, this case illustrates how deterrence measures may backfire precisely because of how they humiliate those targeted and deny them agency. To show that one is (no longer) deterred can be a way of dealing with difficult emotions such as humiliation, shame, and anxiety and of restoring self-esteem, agency, and ontological security.
This, however, is not to say that the Chinese deterrence efforts have been unsuccessful. The fact that the Chinese government televises forced confessions can be seen as a way of communicating a credible threat to other foreign NGOs active in China. That forced confessions are indeed broadly televised, and it would have been possible to detain individuals seen as threats without televising it, further underscores the apparent intent to broadly communicate information about such detention incidents. It is also clear from the empirical analysis that many foreign NGOs and their staff have in fact been deterred.
Conclusion
This article has argued that states may use captivity to deter perceived threats posed by individuals, including foreigners, to regime security, and that emotions are central to this process. Holding certain categories of individuals, for example civil society activists, academics, journalists, and publishers captive, and widely publicising information about their detention, sends a message to other individuals belonging to those categories or who are engaged in similar activities. The article has suggested that states that use captivity for deterrence purposes do so in order to increase regime security and protect their preferred biographical narratives.
The empirical illustrations suggest that deterrence through captivity seeks to instil fear in the target and thereby dissuade it from unwanted action. At the same time, other emotions experienced by those who are held captive also matter. Arbitrary detention and the forced, and sometimes televised, confessions that at times accompany such detention humiliate and deny the agency of the victim. However, those who have been subjected to such treatment may, after having been released, seek to regain their agency, self-esteem, and related ontological security by showing that they are no longer deterred. They may thus tell the story of their captivity and seek to shame the deterrer. That deterrence is in this way related to the denial of agency may be particularly apparent in the case of deterrence through captivity. At the same time, it could be a more general feature of deterrence since those who are deterred will in one way or another and to some degree always be denied, or deny themselves, agency. Future research could consider this dynamic in relation to deterrence more broadly and compared with other kinds of deterrence measures. Do the same kinds of emotional dynamics related to the denial of agency that deterrence through captivity involves operate in the context of other types of deterrence as well?
Acknowledgements
Earlier versions of this article were presented at the University of Warwick in May 2024, at the EISA PEC conference in Lille in August 2024, at the Political Science Department’s research seminar at the Swedish Defence University in May 2025, and at project workshops for the special issue in which this article appears at Freie Universität Berlin in May 2024 and Northwestern University in March 2025. The author would like to thank all those who provided feedback on these occassions, as well as the two anonymous reviewers.
Competing interests
The author declares none.
Karl Gustafsson is Professor of International Relations at Stockholm University and Senior Lecturer in Political Science at the Swedish Defence University. His co-authored article ‘The Insecurity of Doing Research and the “So What Question” in Political Science: How to Develop More Compelling Research Problems by Facing Anxiety’ won the Jacqui Briggs Prize for best article published in European Political Science in 2024. His article ‘Memory Politics and Ontological Security in Sino–Japanese Relations’ won the Wang Gungwu Prize for best article published in Asian Studies Review in 2014, and his doctoral dissertation won the Stockholm University Association’s award for best dissertation in the Social Sciences in 2011. He has published peer-reviewed articles in journals such as International Studies Quarterly, Contemporary Security Policy, and European Journal of International Relations.
Open access
