How authoritarian rulers seek to legitimize repression: framing mass killings in Egypt and Uzbekistan

ABSTRACT How do authoritarian rulers legitimate repressive actions against their own citizens? Although most research depicts repression and legitimation as opposed strategies of political rule, justified coercion against some groups may generate legitimacy in the eyes of other parts of the population. Building upon this suggested link between legitimation and repression, this article studies the justifications of mass killings. To this end, framing theory is combined with recent research on the domestic and international dimensions of authoritarian rule. We contend that frames are directed towards specific audiences at home and abroad. Moreover, given the common threats at the global level and the diffusion of repressive tactics, we assume that learning processes influence discursive justifications of repression in authoritarian regimes. We provide an analysis of government rhetoric by comparing the protest crackdowns of Rabi’a ‘Adawiya Square in Egypt and Fergana Valley in Uzbekistan, taking into account the audiences and the sources of the frames that justify repression. In both cases, we find the terrorism frame to emerge as dominant.


Introduction
How do authoritarian rulers legitimate repressive actions against their own citizens? Most of the literature views legitimation and repression as diametrically opposed strategies of political rule, claiming that an increase in legitimation reduces the need for repression, while an increase in repression automatically decreases legitimacy. 1 We contend that to reduce the danger of delegitimizing political rule and thus to decrease the costs of repression, 2 authoritarian elites often put forward discursive justifications for the use of repression. Going further, besides lowering the cost of repression, successfully justified repression may, in some instances, even serve the purpose of creating legitimacy.
From an empirical perspective, legitimacy is an observable relation between the ruler and the ruled. Therefore, different societal groups may assess regime legitimacy in different ways. State repression in itself may generate support from certain segments of the population and/or from international actors. Thus, legitimation and repression can be perceived as two sides of the same coin. This article focuses on how repression is justified, with justification defined as the official explanation or defence of state action, as one part of the latter's discursive legitimation. We assume that the addressees of legitimation discourses extend beyond domestic audiences to the international sphere. We also take into account the sources of justifications, which may be historical experiences or international learning from global discourses. Recent studies on the international dimensions of authoritarian rule have highlighted patterns of diffusion and learning between autocracies.
Our theoretical approach integrates recent research on the domestic and international dimensions of authoritarian rule with insights from framing theory. To examine how repressive policies are justified, and to whom, we analyse the Uzbek and Egyptian governments' rhetoric during and after the repressive crackdowns of Andijon in 2005 and Rabi'a 'Adawiya Square in 2013. These extreme instances of high-intensity coercion, often described as domestic massacres, were highly visible and thus necessitated public justification. 3 We present a conceptual framework for studying the justification of repression, tracing how domestic and international sources of learning influenced the choice of frames. Moreover, we look at both domestic and international addressees to account for differences between frames. In the explorative empirical analysis, we analyse similarities and differences in the frames that Uzbek and Egyptian regime elites used to legitimize the repression of domestic protest movements. In the final sections, we formulate hypotheses based on our findings and outline agendas for future research.

How does justification work? Combining framing theory and autocracy research
When the security apparatus acts forcefully against certain groups, state representatives often present the target as a threat to the public. This mechanism can be analysed by using framing theory. Framing denotes the way in which issues are presented, what underlying narratives and what solutions are offered. 4 While framing analyses in political science often focus on social movements, we contribute to the literature centred on governmental frames. For authoritarian regimes, the literature often focuses on legitimation as the intended result of official framing. 5 Despite some insightful contributions, 6 governmental framing in authoritarian contexts is still undertheorized. Borrowing from research on the related concept of securitization and on authoritarianism, we introduce two specifications: First, in democratic contexts framing research often studies how governmental frames are adapted and transformed by journalists. 7 In autocracies with censorship and controlled media governments convey certain messages more directly to the public. Thus, we study the nature and origins of justification frames rather than their transmission.
Second, despite governmental control over the public sphere, the results of framing processes are not predetermined, as "frames do not have a single, universal effect" on the recipients. 8 Depending on who the target audience is, the framing of repression can have legitimizing or delegitimizing effects. Moreover, securitization literature suggests that audiences in autocracies do not necessarily comprise all citizens of a state. Instead, a security issue may be communicated to the elite or strategically important groups. 9 This is in line with more general research on authoritarian regimes' strategies. Legitimation as a relational category means that legitimacy claims are directed towards target groups, such as elites, the broader population, or different societal groups. 10 In this article, we focus on framing and its sources rather than on the success of framing endeavours.

What frames are used for justifying repression?
Few political scientists have studied the discursive justification of repression. 11 Dukalskis investigates popular protests in China in 1989, Myanmar in 2007, and Iran in 2009-2010, and finds that protesters were blamed as criminals for inciting disorder and for being manipulated by foreigners and "not committed to the regime's vision of legitimacy". We agree with his conclusion that authoritarian incumbents "attempt to endorse that repression in ways consistent with their legitimating messages". 12 To describe how legitimating messages in authoritarian regimes work more generally, March posits that [t]he main strategy is to define the entire state in relation to common goals, to define the goals and aspirations as virtually constitutive of the nation as such, and to equate the regime with the proper articulation and realization of those goals through the state apparatus. 13 A variety of goals or values in line with more general patterns of legitimation can be invoked to justify repression. Since personalist rule with demobilizing strategies has existed for a while in many authoritarian regimes, we expect justifications to revolve around stability. 14 Other justifications range from state-related arguments such as national unity, legality, security, 15 and public order to sociocultural aspects of tradition. Governments often label their opponents' activities as "harmful behaviour" threatening these values, which in Snow and Benford's terminology can be considered a diagnostic master frame. 16 Table 1 lists different underlying values that regime discourse claims to be endangered by corresponding, alleged oppositional behaviour.

Bridging the domestic and international levels: target audiences of justification and sources of learning
So far, the discussion has pointed to one factor that shapes justifications of repression, the target audiences of frames, as legitimation and framing are relational processes. There is another important factor determining which frames are used. Elites usually Regarding target audiences or addressees of justification frames, the domestic audience can consist of particular groups, such as the regime base, that are subject to selective legitimation attempts. However, frames can also target the broader population. Beyond these audiences, framing might address actors on the international levelfor example, neighbouring states, the international community, regional and international organizations, or diaspora populations.
The audiences need to accept governmental frames, so it can be assumed that elites choose their rhetoric strategically. However, one must avoid the tautological trap of presupposing that elites always use the frames that are likely to resonate, as framing literature shows that extreme or otherwise unsuitable frames are often used. In the exceptional context of public extra-judicial killings, credible justification is necessary. In line with Dukalskis' 17 argument, we expect justifications of repression to fit with broader legitimating strategies. For example, if a Western audience is a key addressee of legitimation, issues of importance to this addresseehuman rights, rule of law, and so onare likely to be broached.
Both the international and the domestic context are not only home to (potential) addressees of justification but often also inspire elites. Authoritarian elites learn from their past experiences. 18 On the global level, discourses on human rights and democracy became salient in the 1990s. A contrasting narrative of terrorism has been added to these narratives since 2001, and repressive tactics and their discursive justifications at both the regional and the international levels have diffused. These developments have shifted the repertoires of authoritarian leaders away from democratic windowdressing and towards more self-confident assertions of their power. We argue that traces of learning can be observed in authoritarian regimes' discourses.
The recent literature on authoritarian learning suggests that international linkages facilitate learning processes to optimize repression. 19 Linkages are especially strong between countries that are geographically proximate or similar in regime type. Yet, in security-related issues, robust linkages also exist between the West and the Arab world regarding the exchange of surveillance technology and military hardware. 20 Thus, sources of learning include neighbouring countries or other regional states, similar states elsewhere, and remote hegemonic players. The country's own past also offers important lessons. Finally, learning may relate to positive or negative examples; that is, learning from success or failure. In sum, we assume that frames are influenced by who the domestic and international addressees are as well as by the learning sources of authoritarian elites (see Table 2).

Research design and case selection
To generate insights on how repression is justified and derive hypotheses on causal factors, we conduct a paired comparison of two explorative case studies: Egypt and Table 2. Factors shaping justification patterns.

Domestic
International Addressees Targeting discourse towards regime base or all citizens Targeting discourse towards allies Sources Learning from history Learning from abroad Uzbekistan. In many ways, these authoritarian states are trendsetters and central to their regions, with the largest number of citizens in the Arab world and Central Asia, respectively. The countries are predominantly Muslim and have a poor record in regard to civil and political liberties. Both Egypt and Uzbekistan are key allies of Western powers in the global war on terror and have had phases of intense confrontation with oppositional Islamic movements, militant and peaceful. As different types of repression would require different narratives, we hold an important factor constant by comparing similar instances of repression. We choose highly visible, lethal, and extra-judicial repression, which necessitates great justification efforts, assuming that Egypt and Uzbekistan represent typical cases of justifications for repression that enable us to further investigate causal mechanisms. 21 We investigate each state's justification of its most brutal repressive incident in recent decades, both of which have been referred to as the respective country's "Tiananmen Square". 22 Despite the generally high level of repression that these countries have in common, the crackdowns in the Fergana Valley in Uzbekistan in 2005 and in Rabi'a 'Adawiya Square in Egypt in 2013, resulting in hundreds of deaths, were extraordinary and led to a ranking of 4/4 on the Political Terror Scale for the year of the mass repression. 23 The cases display many similarities, but also some differences in factors that we assume to impact on justification frames. Concerning international aspects, in the years preceding the respective crackdowns, relations to Western democracies were important to both countries, yet Uzbekistan also had close contacts with Russia and China. On the domestic level, the regime phase during which the repression incidents occurred differed. In Egypt, a new post-coup military regime tried to get rid of the former incumbents, whilst in Uzbekistan, Karimov had ruled for 16 years. These conditions offered differing possibilities for legitimation and thus necessities for justification. The empirical analysis explores how these similarities and differences influence justifications for repression.
In our qualitative frame analysis, the first task was to detect the frames employed. After having categorized the frames, we also investigated changes over time, such as shifts in "strategic narratives". 24 Our analysis drew on publicly available sources such as scholarly works, newspaper articles, blogs, and reports. Most importantly, we examined public statements made by officials before and in the aftermath of violent protests to extract the justifications presented to the public. In particular, we conducted a qualitative content analysis of a corpus of 15 speeches and press statements by the Uzbek President Karimov and Egypt's military chief al-Sisi complemented by officials' statements in media reports. They address various audiences in the period two months before and until two years after the respective repressive crackdown. The frames derived from our theoretical approach in Table 1 were treated as the initial codes. We then analysed the primary sources with regard to the occurrence and context of these frames and complemented them with inductively derived codes. Furthermore, the laws according to which protesters were prosecuted and the accusations made in court offer insights as to how repression was justified, and were complemented by information on legal proceedings in the case of arrest and subsequent trial. This allowed us to capture both the justification of concrete repressive actions and the broader engineered narratives of these events.

Empirical analysis
In the following section, we first describe the repression in the most relevant episodes of high-intensity coercion in Egypt in 2013 and in Uzbekistan in 2005. We then outline the frames that regime elites used to legitimize repression. We trace back the justifications to specific sources of learning and the respective target audiences.

Repression of the Muslim Brotherhood
The Muslim Brotherhood (MB) has a historically ambivalent relationship with the Egyptian state. Since 1952, state repression has alternated with periods of tolerance. Also, militant jihadism has challenged the state at various times, peaking in 1981 with the assassination of President Sadat and again during the 1990s. During times of greater toleration, the MB has played a role in formal politics through participation in parliamentary elections, while at other times its members have faced arrest. The 2011 uprising and Mubarak's ouster were a game changer. The MB was the best organized group and won the subsequent parliamentary elections with over 40% of the seats. It also won the presidential elections, and Muhammad Mursi took office in June 2012. However, the group's attempt to monopolize power and rule in an exclusive manner led to mass mobilization, which was fuelled by the army and resulted in a military coup on 30 June 2013 that was broadly supported by societal groups. The MB did not want to give up on its elected offices, and many of its supporters staged sit-ins; some also attacked security forces. The coup brought about a drastic reversal of the MB's status and MB politicians were arrested on a large scale. The most brutal instances of repression took place on 14 August 2013 with the forceful clearance of the Rabi'a 'Adawiya and Nahda Square protest sites in Cairo. Armoured vehicles, special forces from the Interior Ministry, and snipers on the roofs of military buildings fired at peaceful civilians, shooting and in part even burning an estimated 1,000 people to death. 25 Many protesters were arrested. In return, the US government immediately cancelled a joint military exercise and in October decided to withhold the delivery of some large-scale military systems. 26 Ironically, assistance for counterterrorism was not frozen, nor was the bulk of one billion in annual military aid. 27 The MB was outlawed in September 2013 and declared a terrorist organization in December. The repression ever since has drastically narrowed the public sphere and heavily affected the media, non-governmental organizations (NGOs), and universities. Army Commander-in-Chief and coup leader al-Sisi tightened his grip on power and was elected president in May 2014.

Repression under Karimov
Ever since independence in 1991, Uzbekistan's government has adopted a sceptical stance towards religion, in particular Islam. Then-President Karimov feared that pan-Islamic tendencies would threaten the citizens' loyalty towards the state, and religious organizations have faced tight restrictions, forcing them underground. 28 Nonofficial versions of Islam were referred to interchangeably as "extremism", "Islamism", "Salafism", "radicalism", "Wahhabism", and "Jihadism", and such labelling even extended to Protestant Christians. 29 In the 1990s, a militant Islamist movement, the Islamic Movement of Uzbekistan (IMU), formed in opposition to Karimov's rule. It joined the Taliban in Afghanistan and was crushed after the 2001 war. 30 Another Islamist group, Hizb ut-Tahrir, also sought to build an Islamic society, albeit by peaceful means. Finally, the so-called Akromiya movement is devoted to spiritual salvation. Regime agents blamed all of these organizations for perpetrating lethal attacks in Tashkent in 1999 and 2004. However, observers doubt the official narrative about Islamist terrorism, arguing that the bombings were part of a power struggle between different factions of the security apparatus. Uzbekistan gained new international allies after 2001 by joining the global war on terror, primarily against Afghanistan. The closure of the political sphere kept oppositional activists in exile or imprisoned facing torture.
In May 2005, security forces perpetrated a massacre in Andijon in the Fergana Valley. Local protests were first triggered by an unfair trial against 23 middle-class entrepreneurs who were charged with being members of the Akromiya. 31 The trial was monitored by 1,000 to 4,000 people outside, an unusual gathering of citizens. On the last day of the proceedings, the judges announced that the defendants would be imprisoned indefinitely without a verdict. Enraged family members and friends stormed the prison and freed them and others during the night. The group joined protesters in occupying the regional administration building in Bobur Square on 13 May and publicly denounced social and political injustice. Throughout the day, snipers and security forces from all branches killed hundreds of protesters. 32 Citizens from Andijon who tried to flee to the nearby Kyrgyz border were met with gunfire by Uzbek guards. 33 According to some sources, the military and intelligence agencies assumed control of the Ministry of Interior's anti-terror units, indicating a power struggle from which the former emerged victoriously. 34 After the Uzbek government failed to conduct an independent investigation as demanded by the United Nations (UN), 35 the European Union (EU) and US reacted by imposing sanctions that led to Uzbekistan's swift realignment with Russia through intensified collaboration in the Shanghai Cooperation Organization (SCO) and the Collective Security Treaty Organization. 36 Uzbekistan abolished the death penalty in 2008, although severe and occasionally lethal torture continues to this day. The number of political prisoners is estimated to be approximately 10,000.

Justification of repression in Egypt
Egyptian officials have made extensive use of the frames "security", "order", and "terrorism". Their main narrative was that the Rabi'a Square protesters consisted only of violent MB supporters, who were terrorists. In addition, the disruption of daily life and public order were contrasted with the security of state and citizens. Then-Prime Minister Beblawi "justified the use of force saying that Morsi loyalists had been inciting chaos around the country, 'terrorizing citizens, attacking public and private property'." 37 One justification strategy was to depict the police forces as defending their own lives, the state and its institutions. Therefore, officials highlighted and probably exaggerated the number of wounded policemen, while downplaying the number of victims among the protesters. 38 Interior Minister Mohamed Ibrahim employed the argument of selfdefence, explaining that the police forces "were surprised by protesters who started firing live ammunition" and added that "'clear instructions' were given to security forces to limit use of weapons to teargas after protesters had been told to leave by loudspeakers". 39 Framing protests as terrorism is a tried and tested narrative from the "war on terror". Even before the crackdown, on 24 July, al-Sisi had asked "all honest and trustworthy Egyptians" to "come out to give me the mandate and order that I confront violence and potential terrorism". 40 Afterwards, during a news conference on 17 August, presidential adviser Mostafa Hegazy said, "We are facing a war launched by extremist forces escalating every day to a terrorist war". 41 Similar wording dominated government discourse and legislative amendments. In the subsequent trials, the defendantsranging from high-ranking MB members to photojournalistswere accused of forming an armed gathering of more than five people that endangers public safety and security […], premeditated murder of security personnel, vandalizing private and public property, forcibly occupying buildings, obstructing traffic, terrorizing the public, and restricting citizens' right to freedom of movement and personal safety. 42 The public prosecution office held the MB liable for the violence that occurred during the sit-in. When al-Sisi declared his candidacy for presidency in a televised March 2014 speech, he made terrorism the most prominent issue, stating, My fellow citizens! We are threatened by the terrorists […]. It is true that this is my last day in uniform but I will fight every day for Egypt free of fear and terror.
[…] we'd rather die before Egyptians are terrorized. 43 Besides the frame of terrorism, the reference to a legal basis for repression seemed to be vital to Egyptian officials. While the crackdown was ongoing, government representatives declared that the crisis was being handled via "security measures within the framework of law". 44 Later, in an interview to Western audiences, al-Sisi referred to the rule of law and non-interference with judicial processes when questioned on his country's human rights record. 45 Depicting repressive measures as being in accordance with legislation has been a strategy of priority for judicializing repression since Mursi's ouster. 46 Initiatives such as the 2013 protest law, the anti-terror legislation, and the new constitution offer new regulations for judges to apply in unfair trials, and legalize administrative sanctions by police and military forces.

Justification of repression in Uzbekistan
The Uzbek government justified the Andijon repression claiming that it was acting in defence of the population against armed criminals. Karimov accused the protesters of using women and children as human shields. The main narrative focused on Islamism, which was becoming increasingly embedded in the global discourse on the war on terror. In a press conference on 14 May, the president blamed the Akromiya for the events and portrayed it as Hizb ut-Tahrir's local branch, although the two movements have no connection. Karimov drew a parallel to the unrest ensuing after the Tulip revolution in Kyrgyzstan two months prior, insisting on foreign funding and the participation of foreigners seeking to establish a "Muslim Caliphate". 47 He cited intercepted phone calls from protesters to Afghanistan and Kyrgyzstan as proof of international connections. The Foreign Ministry added the IMU to the list of suspects. 48 Only a year later, Karimov indirectly acknowledged socio-economic problems as a cause of the protests and dismissed the province's governor. 49 During the first press conference after the Andijon events Karimov did not accuse the protesters of being terrorists themselves, but over time the official rhetoric shifted towards this terminology. 50 In court, the charges filed against arrested activists included "terrorism, attacking the constitutional order, murder, the organization of a criminal band, mass disturbances, the taking of hostages, and illegal possession of arms and explosive materials". 51 Most of the defendants pleaded guilty, a testament to the horrible torture in Uzbek prisons, which can also extend to family members. 52 The forced testimonies reflected the authorities' confused approach that did not present a coherent narrative, but a vast conspiracy involvingin no particular orderthe BBC, RFE/RL, Chechen military instructors, NGOs, training camps in Kyrgyzstan, the U.S. Embassy in Tashkent, and extremists linked to Al-Qaeda [who] aimed to spark a Georgia/Ukraine/Kyrgyz-style revolution in Andijon in order to transform Uzbekistan into an Islamic state that would serve as the launching pad for a drive to establish a worldwide caliphate. 53 The defendants were not able to obtain their verdicts, rendering them unable to appeal. 54 Journalists and human rights activists whose accounts of the Andijon events deviated from the official story were defamed and persecuted as terrorism supporters. Independent news websites were shut down in order to erase alternative narratives in a carefully orchestrated "information black-out" intended to enforce the official version. 55 In addition to journalists, poets who wrote songs about the massacre were arrested for insulting the president. 56 Two citizens who possessed a cassette with the songs were handed even harsher sentences.
Not all repression was justified. As in Egypt, the Uzbek government downplayed the number of citizens killed to 187, well below the estimates from other sources of between 600 and 1,000 victims. Moreover, Karimov insisted that "not one peaceful citizen" had been shot, 57 especially not women and children. Witnesses reported that dead women and children were hidden from the public and that only male victims were made available for identification in makeshift morgues. 58 Justifications of repression reached their limits when it came to obviously innocent victims. Table 3 summarizes which frames officials used to justify repression against opponents in Egypt and Uzbekistan.

Justification frames: contents and context
In both cases, the central claim was that the protesters were armed criminals who had fired before the security forces opened fire. The danger to security also figured prominently. Interestingly, in both countries the terrorism frame used against civilians who were not only unarmed, but also not Islamists, gained importance over time and rendered it the most popular ex-post legitimation of repression. Karimov only implicitly mentioned terrorism in his press conference on Andijon, but amid international calls for an independent investigation, instead of "criminal behaviour", "terrorist aggression" became the major accusation in the trials and the surrounding discourse. 59 In Egypt, the terrorism frame was already introduced in preparation of the crackdown, and it loomed large in state-controlled coverage of the events. Whilst other frames, such as "disruption" or "foreign influence", vanished shortly after the repressive event, terrorism and extremism are the forms of harmful behaviour both governments stressed in the months and years after the crackdowns. After some months, frames about harmful behaviour subsided, while references to endangered positive values from our conceptual framework of justifications, such as security, unity, and so on, amplified in both cases. This could be interpreted as "normal" legitimation discourses setting in again after critical moments.

Domestic sources of learning
For tracing where the similar frames originate, the most obvious explanatory factor is that both Egypt and Uzbekistan had a history of successfully utilizing terrorism discourses. Officials thus resorted to established routines by labelling crackdowns as "anti-terrorism".
In Egypt, harsh repression accompanied by a security and terrorism discourse has previously targeted the Islamist opposition, under the pretext of combating jihadist militants. This happened after jihadists assassinated President Sadat in 1981, and in the so-called civil war between jihadists and the state during the 1990s. The pattern of first initiating a crackdown, then introducing certain narratives before altering legislation to convict the defendants of terrorism in military courts already existed back then.
During the repression against the MB in the 1990s, its members were frequently accused of attacking buildings and people. Under Mubarak, the term terrorism (irhāb) gained traction, and in 1992 the first Egyptian anti-terror law was passed. Although until 2011 many political opponents and bloggers faced administrative arrest for terrorism and/or drug charges, during the interim military rule of the Supreme Council of the Armed Forces (SCAF) in 2011-2012 the typical accusation against protesters became "thuggery". During SCAF's rule, the discourse shifted from "revolution support" towards stability and safety as ends in themselves, as observable in statements published on SCAF's official Facebook site. 60 Since the 2013 coup, the discourse centring on terrorism, security, and extremism has become a cover to justify not only the repression of MB activists but also the entire crackdown targeting the liberal opposition, the media, and NGOs.
The justification pattern used in Uzbekistan after the Andijon events also picked up previous discourses. A long-standing narrative is to accuse opposition movements of conspiring to destabilize the country and the region. Lumping different dissenting groups together is a trademark of Karimov's writings, where the words "opposition" and "radical" are usually used together "as if to suggest that […] all opposition is by definition radical". 61 Karimov's strategy was to insist that "Uzbek Islamism is wholly 'imported' from outside the country". 62 In February 1999, first the Lebanese Hizbullah then Hizb ut-Tahrir were blamed for car bombings in Tashkent, before the IMU and a secular politician in exile were indicted in court, invoking the unlikely scenario of militant jihadism in Afghanistan taking on Uzbekistan. 63 Both in the 1999 and the 2004 bombings, rival clans dominating different security services were suspected of being behind the violence. 64 Given the positive international recognition of counterterrorism efforts, "a kind of controlled terrorism, when duly ascribed to its Islamist adversaries, might be acceptable and even desirable for the regime". 65 Already in 2000, members of the political opposition were tried on charges of terrorism. The law defined terrorism as "socially dangerous wrong doing". 66 This definition sees terrorism even in ideological goals, such as the future establishment of a caliphate, without the exertion of violence. 67 After the IMU had been crushed in the Afghanistan war, Hizb ut-Tahrir was declared guilty of having organized the attacks of 2004. The government pursued its "plan to tie all forms of protest to terrorism" also after May 2005 in Andijon. 68 Some differences in the framings resulted from the domestic contexts, more specifically tried and tested discourses. The trials against protesters are a case in point. In Uzbekistan, prosecutors emphasized radicalism and foreign conspiracies. According to Horsman, this approach is rooted in Soviet legacy. He cites the "ideological and arbitrary definition of terrorism, the 'criminalisation' of the terrorist, the portrayal of external conspiracies against the state and the call for popular mobilisation against the threat", as well as the court trials, as displaying similarities to the political climate of the 1930s. 69 This resemblance could be a result of political socialization, as Karimov was the final Communist Party first secretary in Uzbekistan.
In Egypt, where the military depicts itself as the guarantor of national and state security, threats to state property were stressed. Furthermore, as "the rule of law" has been a widespread frame since President Sadat's rule, the rhetorical reference to the legality of repressive means was common.

International sources of learning
The prominent role terrorism assumed in the discourses of both countries is also a result of the diffusion of "war on terror" rhetoric and practices on the global level. Together with previous collaboration with the US in "fighting terror", international learning processes may have encouraged both governments to use this frame.
While the Uzbek government has always employed a tough approach towards Islamists, it enhanced its embryonic anti-terror discourse after 9/11 to receive financial support. Western partners have taken the alleged presence of Islamists seriously and supported Uzbekistan's anti-terror fight for geostrategic reasons. The international interest in a "stable" neighbour for Afghanistan diminished the danger of external criticism of Uzbekistan's policies and resulted in considerable Western assistance for police and intelligence services. 70 After the 2004 attacks, the government had referred to the perpetrators as representing "jamoats", which can be read as a reference to Egypt's al-Jama'a al-Islamiyya, the Egyptian state's opponent in the 1990s civil war. This deliberate act of framing resulted from learning from Egypt and was directed at the international community "with US backing for Egypt's 'war on terrorism' in mind". 71 One special counterterrorism unit that participated in the Andijon crackdown had received training in the US. 72 After the bombings in 1999 and 2004, the Uzbek government had learned it could get away with blaming terrorists for creating unrest without too many questions being asked, and it designed its rhetoric in 2005 accordingly. However, this lesson did not hold in the extreme case of the regime violence in Andijon, and both the EU and later the US imposed sanctions.
Egypt may have learned a lesson from events in Syria. There, by 2013, Assad partially succeeded in convincing the international community of the opposition's terrorist nature, not least by influencing circumstances to prompt a shift towards more extremism. 73 Given that even the fiercest repression can find some cover under a terrorism discourse, combined with domestic experiences and the general impression of the worldwide discursive success of the terrorism narrative, it is no wonder that the interim military government took up this frame to justify the Rabi'a 'Adawiya massacre.
When it comes to explaining differences in justifications, the divergent international partners play an important role. China and Russia, as major authoritarian powers with a communist past, are geographically proximate and have close linkages to Uzbekistan. The SCO provides a regional platform for cooperation on the security level, which was strengthened in the summer of 2005 after the West had called for an independent investigation. It is also a site for promoting the so-called "Shanghai Spirit", 74 norms that serve as an alternative to "Western" values. The fight against the "three evils"extremism, sectarianism, and terrorismtops the SCO agenda. 75 These issues also pervade Uzbek justifications for repression.
A primary goal of Uzbekistan's increased repression in 2005 was "diffusion-proofing" to avoid a colour revolution. Neighbouring Kyrgyzstan had been the latest and closest country where unrest had "resulted in the president's decision to vacate office". 76 Thus, in his above-mentioned press conference, Karimov refuted any similarity with the colour revolution protests, to the point of denying that people had assembled in a square. Here, the president tried to learn from others' failures.
In Egypt, specific international sources of learning were comparatively less relevant: the Arab League has been neither meaningful nor united enough to shape Egypt's framing. Quite to the contrary, some Arab monarchies followed Egypt's example of declaring the MB a terrorist organization. 77 When it comes to learning from failure, the Tahrir uprising was the most obvious, yet domestic source of learning.

Domestic addressees
Regarding similar frames, towards their own populations, officials rhetorically exploited negative examples of violence in the respective regional neighbourhoods to validate the need for repression. Instances of civil war and popular upheaval influenced the elite's threat perceptions, but they were primarily instrumentalized as justifications for harsh measures to avoid the spillover of conflicts. Regarding civil war, Tajikistan served as a cautionary tale for Uzbekistan, while Egypt could choose from a range of violent conflicts in the region. Both governments were able to point to the challenge of jihadism in their neighbourhood, most notably in Afghanistan and Syria.
A major concern to incumbents was to warn against the danger of protests, as demonstrated by the colour revolutions and the Arab uprisings. For Uzbekistan, Kyrgyzstan represented the key negative example, while for Egypt, its own recent history was a case in point. The latter is ironic, for even the military coined its 2013 coup the completion of the Tahrir "revolution". Demobilizing the domestic public was therefore a priority.
Diverging frames reflect the specific addressees in the respective contexts. In Egypt, anarchy and the disruption of daily life through the week-long occupation of a public square were prominent narratives. The highlighting of the right to "normal" life in Egypt points to the context of the lasting political mobilization since the Arab Spring and the profound changes that have imposed economic and emotional strain on many citizens.
In Uzbekistan, an important diversion strategy was the promotion of a nationalist "Uzbek" Islam as opposed to transnational forms of Islam seen as dangerous. Thus, the allegation of foreign influence in Andijon even amounted to the claim that Taliban fighters were present. Outlandish conspiracy theories and allegations of radicalism were unique to Uzbekistan.

International addressees
As mentioned before, the global war on terror has become a convenient blueprint for justifying state violence both in Uzbekistan and Egypt. The West has been an important addressee of anti-terrorism discourse in both states for some time. When the concrete crackdown incidents disappeared from public discourse, a more general narrative spread holding extremists responsible for imminent domestic collapse and regional instability. In an interview addressed to a US audience, al-Sisi offered his interpretation of the MB as "the godfather of all terrorist organizations". 78 In addition, with his speech before the UN General Assembly in October 2015, al-Sisi presented his window-dressing programme against terrorism. This is part of Egypt's discursive shift on the international stage: the topics extremism and terrorism now outweigh buzzwords of stability and peace that were typical in the 1990s and 2000s in speeches to the UN General Assembly. 79 The Uzbek government also pursued a sophisticated approach to transmit its own reading of events to an external audience. Its embassy in the US distributed a video in support of the official narrative to think tank scholars in an effort to win over Western academics. However, parts of this "documentary" were also used inside Uzbekistan in a propaganda video that was frequently broadcast on television in the summer of 2005. 80 While Uzbekistan's security cooperation involved diverse states, Egypt's close relations to the US were less balanced by other great powers. Thus, when both countries experienced pressure and scepticism from the Western and international communities after their crackdowns, reactions differed. Whilst the Uzbek president talked about "true and false friends" of his country following the imposition of US and EU sanctions in his speech at the SCO summit on 14 June 2006, his Egyptian counterpart lamented about the US not supporting his country enough. 81 The reactions by international addressees towards the crackdowns and the immediate justification efforts diverged. Russian officials and media took up the Uzbek government's framing, asserting that the dead of Andijon were "'militants, comprising bandits, Islamist radicals, and Taliban fighters' [who had] infiltrated from Afghanistan". 82 The fact that foreign influence was used to frame Uzbek protesters can be interpreted as a strategy directed at other international actors who do not really know what is going on inside the opaque politics of the country. 83 In Egypt, while using the argument of foreign influence is not alien to the al-Sisi government in other contexts, the frame was not used to justify the Rabi'a crackdown.
Rather, the Interior Ministry stressed its adherence to law in an interview with Western media, for example, when claiming "accordance with internationally accepted crowd-control standards". 84 The fact that the government is aware of international human rights observers was reflected in its rhetorical support for investigations of the incidents. However, the lack of results even led to interim President Adly Mansour calling for a renewed investigation, which the Justice Ministry denied. 85

Conclusion and outlook
This article has highlighted the links between legitimation and repression by analysing the justifications of the crackdowns on mass protests in Egypt and Uzbekistan in 2013 and 2005, respectively. We have argued that examining the addressees of official discourses and the sources of learning are central to understanding the governmental framing of opponents and repressive events. We suggest that combining the current research on authoritarianism with the framing approach provides scholars with conceptual and theoretical tools to investigate the discourses of autocratic rulers at such critical points in time, while also taking international influences into account. Our explorative analysis of two cases of brutal crackdown leads to innovative hypotheses that merit more thorough investigation in other contexts of governmental justifications.
Regarding the general frames, the justifications offered for the crackdowns were very similar. In both countries, the security and stability of the state and society were invoked, whereas protesters and their allegedly affiliated groups were accused of criminal behaviour, terrorism, extremism, violence, and disruption. The allegations brought forward echo Dukalskis' findings from other cases. 86 With regard to the timing of frames, we found that many frames referring to "harmful behaviour" were replaced by more vague "values" rhetoric over timewith extremism and terrorism being the crucial exception. We thus suggest to distinguish between temporally close and distant frames in future studies.
Whilst the terrorism frame is not a universal toolit was notably not used after the Tiananmen Square massacre 87it appears to be a popular justification in the current century. We hypothesize that terrorism and stability discourses are likely to be used under both extraordinary and "normal" conditions of contention, as the regime phases in our cases had little impact on the discourses. One important cause of the popularity of the terrorism frame is, in our reading, that it links concrete events and abstract risks to which both domestic and international addressees can relate.
Effects of international influences seem to play a large role in the choice of the general direction of frames besides their embeddedness in the respective regime's legitimation strategies. Although the rulers in both countries feared large-scale political protest, they claimed to face threats that tied in with the global diffusion of the war on terror discourse. Looking at the more specific justification frames, the "finetuning" of frames can be thought of as a second step in the elites' discourse strategy when they consider additional audiences and potential negative reactions from the international sphere. For example, the strong Egyptian focus on legality was to some extent directed towards a Western audience, whereas the foreign interference frame that was prominent in Uzbekistan echoed narratives that prevailed in its regional surroundings.
Our findings also support theoretical arguments on the international factors stabilizing authoritarian rule: Learning from abroad was triggered by geographical and system proximity, as epitomized by the negative Kyrgyz example. Uzbekistan made use of linkages to authoritarian great powers through regional organizations to underscore regime discourses and resolve isolation on the international scene. The analysis points to a case of learning a lesson that only partly held true for the extraordinary scale of repression: Despite adapting their frames to international addressees, both Uzbekistan and Egypt were subject to negative international reactionssanctions and freezing of military assistance.
These repercussions highlight the fact that, besides studying the contents of frames, another path worth pursuing would be to examine the effects of the legitimation of repression. The impact could range from military support or, if resonance is low, sanctions on the international level, to mobilizing and radicalizing effects in the domestic sphere. Harsh repression has often led to radicalization, turning initially false accusations into self-fulfilling prophecies. While the effects of repression itself are being increasingly studied, we argue that a government's justification of repression should receive more attention. The frames' (non-)resonance with the audience and counterframing efforts would be crucial topics of study. The counter-framing by opposition activists indicates that the question of where the "real" threat to security comes from is highly contested even under harsh authoritarianism. In his song about the Andijon massacre, Uzbek poet Dadaxon Hasanov sang: "We know who the terrorist is". 88 Notes 25. Prime Minister Hazem Beblawi spoke of a death toll "close to 1000" in Rabi'a Square and Nahda Square together, while the National Council for Human Rights counted 624 civilians plus eight policemen killed in Rabi'a Square. Other estimates are higher, including Human Rights Watch figures. The organization has stated that "a minimum of 817 people and more likely at least 1000" persons were killed (HRW, All According to Plan). 26. The Independent, 15 August 2013; Sharp, "Egypt." 27. CNN, 9 October 2013. 28. Ilkhamov, "Uzbek Islamism," 42. 29. Peyrouse, "Does Islam Challenge the Legitimacy of Uzbekistan's Government?," 2-4. 30. The US declared the IMU a terrorist organization in 2000. 31. As all information about the Akromiya as an organized body goes back to non-public sources the government presented in court rather than authentic material, the group's existence is questionable, see Kendzior, "Inventing Akromiya." 32. The death toll is highly contested.