What do you call the political upheaval? The conceptual linkage between revolutions and coups

ABSTRACT
 What kind of relationship exists between coup d’états and revolutions? Although these two phenomena have conventionally been studied separately, a close and intricate relationship exists between them on a conceptual level. Since “revolution” has a far more positive connotation than “coup,” the organizers and supporters of a political upheaval will attempt to frame it as a “revolution,” while the ousted leader and other opponents will try to frame it as a “coup,” Thus, each involved party uses the two labellings differently in accordance with their political incentives. Based on this understanding, the present article focuses on three types of events – “revolution,” “pure coup,” and “revolutionary coup”– and considers the preferences of various relevant actors –the military, protesters, the ousted leader, international actors, and the media– on how those events are to be labelled. The validity of the argument is tested using a newly constructed dataset on the labelling of political upheavals between 1975 and 2014 and case studies on Portugal (1974), Philippines (1986), and Burkina Faso (2014).


Introduction
In June 2013, following massive anti-regime protests that turned violent, the Egyptian Armed Forces detained President Mohamed Morsi and declared the suspension of the constitution. Although Morsi condemned the military's actions as "a full coup categorically rejected by all the free men of our nation" (The Washington Post, July 4, 2013), General Abdul Fattah al-Sisi justified it by stating that the Egyptian people have "called on us to perform public service and secure the essential protection of the demands of their revolution" (BBC, July 4, 2013), and that "the Egyptian military does not encourage coup d'états. The last coup was in the fifties" (The Washington Post, August 3, 2013).
This example illustrates the intricate situation in which certain actors label the same political upheaval as a "coup," while others label it a "revolution." This situation is not limited to the aforementioned cases but is extremely common in the phases of political change.
Conversely, in comparative politics, coups and revolutions have been conceptualized as different phenomena. Coups are defined as "illegal and overt attempts by the military or other elites within the state apparatus to unseat the sitting executive." 1 The important elements of this definition are: 1) the attempts are targeted at chief executives; 2) it is a part of the state apparatusthe military, in most casesthat becomes a threat to the incumbent regime, and 3) coups are inherently unconstitutional acts. A representative example of definitions for revolutions is "a forcible transfer of power over a state in the course of which at least two distinct blocs of contenders make incompatible claims to control the state, and a significant portion of the population subject to the state's jurisdiction acquiesces in the claims of each bloc." 2 Here, the minimum required elements to satisfy the definition would be: 1) irregular changes in the structure of political authority and 2) a high level of mass participation. 3 From the aforementioned definitions, although both are unconstitutional acts that occur outside of political systems, the two phenomena appear to be distinguishable from each other. A coup is a horizontal dispute over power, while a revolution is a vertical dispute performed through a popular uprising or mass protest. In many previous studies, coups and revolutions have been analysed separately. How can we understand a situation in which the two phenomena are considered distinct in the objective analysis of scholars but labelled in a mixed manner by the parties involved in these political upheavals?
In the context of political upheavals, the involved partiesthe public, military, ousted dictator, foreign nations, and othersmake choices concerning how to refer to the political upheaval in question. These choices reflect their political goals and intentions. Specifically, they choose whether to call the event a "revolution" or a "coup" (or similar terms). Therefore, since "revolution" has a more positive connotation than "coup" does, the organizers and supporters of a political upheaval will attempt to frame the event as a revolution, whereas those ousted during the event or who oppose it will frame it as a coup. Thus, the choice of whether to call the event a "revolution" or a "coup" is highly strategic and political. This is why the labelling and naming of political upheavals are complicated. The labelling is important in that it has the effect of conferring or stripping legitimacy on/off the seizure of power. Survey experiments have empirically shown that the labelling of a coup has the effect of reducing people's support for the transfer of power. 4 Thus, while labelling a political upheaval is politically important, existing research has not provided an exhaustive empirical analysis of how involved actors have labelled past political upheavals. Powell et al. (2022) pointed out that scholars use "coup" in a relatively narrow sense, whereas the general public, governments, and international organizations may use it in a broader sense. Although their findings are interesting, they did not empirically demonstrate a specific trend. In this article, we provide an exhaustive report on the labelling of political upheavals. First, we examined three types of political upheavals: coup, revolution, and revolutionary coup. One notable aspect of our analysis is the inclusion of the category of "revolutionary coup." 5 These political upheavals exhibit characteristics of both coups and revolutions, but as we will show later, they have outnumbered coups and revolutions in the period after 1975, highlighting their importance. Second, we examined how political upheavals were labelled by the military, citizens, ousted leaders, international actors, and media. This perspective of investigating the differences in labelling by each actor is novel in itself; however, by paying attention to ousted leaders, which has frequently been overlooked in previous studies, we scrutinized the political disputes over labelling and the strategic nature of labelling.
In doing the above, we constructed a dataset of how each actor labelled the 127 political upheavals between 1975 and 2014 and observe comprehensively how the labels "coup" and "revolution" were used in those cases. Thereafter, we analysed three illustrative cases. In this way, this article aims to empirically clarify how actors in political upheavals have used labels such as "revolution" and "coup." Specifically, we find that both citizens who initiate revolutions and militaries who initiate coups tend to label their political changes as "revolutions," while ousted leaders tend to denounce them as "coups." Additionally, international actors and media will use the label "revolution" when they welcome the regime change, and "coup" when they condemn it.
The contributions of this article are as follows. First, to the best of our knowledge, this is the first systematic study on the labelling of political upheavals. This does not mean that there have been no studies focusing on labelling. Marsteintredet and Malamud points out that coup-makers sometimes call their actions "revolution" in order to procure legitimacy. 6 Also, recognizing the importance of labelling to political upheavals, there are already studies that have investigated it in a comprehensive fashion. 7 However, while these analyses focused only on the coup and the military that initiated it, this article comprehensively collects data on labelling, including actors other than the military.
Second, this article newly shows the conceptual linkage between a coup d'état and revolution. Previous studies have pointed out the connection between revolutions and coups as phenomena, i.e. that protests tend to lead to coups. 8 However, despite being equally important in political processes, the conceptual aspect of this linkage has only been addressed by a few studies such as Marsteintredet and Malamud. 9 Third, it contributes to the existing research on the legitimization of regime change. There have been numerous studies on the legitimacy of established regimes in previous research, 10 and Grewal and Kureshi have conducted an excellent analysis on the legitimation through elections after a coup. 11 However, there has been relatively little research on legitimization at the point of seizure of power through unconstitutional means at an earlier stage.
Between "coup" and "revolution": labelling disputes among the involved actors This article analyses the phenomenon of diverse actors labelling the same political upheaval in various ways. The timing of the labelling can vary. For example, labelling may be retroactively carried out after a certain period of time has passed after the regime change. However, this article focuses on labelling during the political upheaval. In this section, we will organize how various actors choose different labels, according to the types of political upheavals. Specifically, we will divide the actors into five categories: military, citizens, ousted leaders, international actors, and media, and the types of political upheavals into three categories: coups, revolutions, and revolutionary coups.
First, as a premise, this article assumes that each actor uses labelling according to its own strategy or need. In other words, labelling is a type of framing, and it can greatly influence the impression people have of a phenomenon. 12 Thus, labelling can contribute to legitimizing or creating a positive impression of political upheaval or, conversely, de-legitimizing or creating a negative impression.
In this regard, both coups and revolutions involve the seizure of power through extraordinary or unconstitutional means. While there are ways to secure procedural legitimacy, such as by putting in place the constitutional successor to the toppled chief executive/head of state (especially in presidential regimes where the Vice President is used as an ally), additional efforts must be made to secure legitimacy compared to a normal transfer of power. Therefore, those who wish to justify itbe they newly empowered actors or foreign countries and media who welcome such regime changetry to supplement the lack of legitimacy by positive labelling. Conversely, those who wish to unjustify itbe they actors that have lost power or foreign countries and media that do not welcome such regime changetry to denounce such lack of institutional legitimacy through negative labelling.
So, what exactly do the various actors label political upheaval? What is positive labelling and what is negative labelling? In this article, we argue that whether a given political upheaval is referred to as a "coup" or a "revolution" is key when labelling it. This is because the connotations of these words differ greatly on a conceptual level.
Coups are an inherently irregular means of assuming power. They have been viewed negatively since at least the 1950s and regarded as destructive to democracy and constitutionalism. 13 For instance, former United Nations Secretary-General Kofi Annan, claimed that "all across the world, the evils of coups are becoming increasingly recognized by the coup-makers, once they become the victims of the next coup." 14 The situation with the term "revolution" differs greatly. It has positive connotations of popular participation and support, and its usage can provide democratic legitimacy to regime change. For example, in its coverage of the "Rose Revolution" in Georgia in November 2003, the media demonstrated popular frustration over declining standards of living as a reason why President Shevardnadze was ousted from power. This demonstrated that the connotations of the words "coup" and "revolution" differed greatly. 15 However, it is important to note that this does not necessarily apply to so-called ideological revolutions. Even though they are popular uprisings, the values upheld are ideological. In this sense, examples such as those seen in China and Cuba do not necessarily have positive connotations, depending on the political position of those who label them. Accordingly, to pre-empt the discussion in advance, the dataset created in this article also follows suit, dealing with the period after 1975, when few ideological revolutions took place and the majority were democratic revolutions. 16 Based on these assumptions, we organized each actor's labelling preferences for revolutions, coups, and revolutionary coups below. 17

Military
First, we considered labelling by militaries. For militaries that have launched coups d'état, the procurement of legitimacy is a particularly serious problem. This is because they are unable to justify their acts in the name of democratic legitimacy or the general will of the people. 18 Moreover, the success or failure in justifying their transfer of power significantly influences actual political processes. For example, support from the public and international community decisively affects the ultimate success and failure of coups. 19 Even if a coup is successful, failure to justify sufficient legitimacy might incite subsequent civil protests demanding democracy. 20 It is inherently difficult to perpetuate new post-coup regimes without support from the citizenry. 21 In this context, it is important for even people with guns who have already taken power to obtain legitimacy from the citizen.
Next, we would like to move the discussion to labelling as a means of procuring legitimacy. Ordinary observers and researchers recognize a coup to have occurred when a military intervenes to overthrow an existing regime and seize power. Such a case will be included in datasets on coups and coded as a coup in datasets on political upheavals. However, because of the aforementioned negative connotations, the military will prefer to frame the event as having been something other than a coup.
How, then, do militaries refer to their actions while denying that their assumption of power was through a coup? It is important for militaries that their actions be portrayed as having occurred in cooperation with the public or its support. Thus, it is useful for them to label their actions as a "revolution" or "popular uprising." For example, when Ethiopian General, Aman Mikael Andom, ousted Haile Selassie I and established a military regime in September 1974, he stated that "the people support us, and we have no need of force to go ahead with our revolution" (New York Times, May 5, 1974). Similarly, in the 1999 Guinea-Bissau case, the military justified their actions by saying that "this was not a coup; it was the clear will of the people" (Associated Press, May 11, 1999). The effect of labelling a coup as a revolution or popular uprising in order to acquire legitimacy has also been demonstrated empirically in a survey experiment by Grewal and Kureshi (2019).
Therefore, it can be said that when militaries conduct coups amidst mass protests, they prefer to frame the event as a revolution rather than a coup. Furthermore, even when these events occur in the absence of mass protests, militaries still attempt to do all they can to reject the label of "coup" and frame their actions as having popular support.
It should be noted that the military in this context includes civilian elites who participated in the coup. That is, these actors also label under the same preferences and strategies as the military. For example, in the 2012 coup in the Maldives, the mastermind, Vice President Mohammed Waheed Hassan, justified the political change by saying "It is wrong to describe the events as a coup" (The Independent (London), February 9, 2012).

Citizens
Next, we examined the behaviour of citizens. First, when regime change occurs solely through a popular uprising and the military plays no particular role in the process of overthrowing that regime, the public labels its own actions as "revolution." This is natural and uncontroversial. What should be considered is how the public interprets the action of the military. That is, how does the public label the actions of the military when the overthrow of a regime is exclusively carried out by the military (i.e. pure coups) or when a coup occurs in the midst of a popular uprising (i.e. revolutionary coups)?
When they view these actions favourably, members of the public will deem them a revolution and an expression of "people power." This is particularly true in cases where the coup was preceded by anti-government demonstrations. For example, the opposition parties who conducted demonstrations in the 1999 Ivorian coup said that "this is not a coup," and "this is a revolution supported by all the Ivorian people" (Africa News, January 4, 2000). Even in the absence of demonstrations, citizens who view the toppling of a regime by the military favourably will frame the political upheaval as a revolution or use similar language. For example, the 1989 Paraguayan coup was a pure coup in which General Andrés Rodríguez ousted long-time president, General Alfredo Stroessner. Nevertheless, politicians called the coup "a revolution for democracy" in speeches to crowds (The Times (London), February 4, 1989).
What we must consider more carefully are cases where the relations between the public and military are unfavourable. Even when a regime change is made possible by the military defecting from the former regime or adopting a neutral position during popular protests, the public will frame the military's actions negatively if the relationship between the military and the citizens deteriorates. In such cases, citizens may invoke "people power" and claim that they, not the military, were the leaders of the "revolution," or use the negative labelling of political upheavals to criticize the military, making statements such as "the revolution was stolen" or "the military's act is a coup."

Ousted leaders
Next, what labelling do those who have lost powertypically, dictatorsapply to political upheavals? Ousted leaders aim to delegitimize the new regime and the process of regime change. Extrapolating from what we have argued thus far, it is plausible that ousted leaders will claim that they were ousted not in a revolution but rather in a coup. Being ousted in a revolution implies that the loss of power was in accordance with the will of the people. However, by labelling the regime change as being the result of a coup, they imply that their loss of power was the result of irregular regime change by the military. Thus, while militaries attempt to avoid the "coup" labelling, ousted leaders prefer it.
Furthermore, it can be assumed that ousted leaders will attempt to delegitimize a political upheaval by labelling it as a "coup" even when the military does not act, i.e. when regime change occurs in a purely popular revolution that does not meet any criterion as a coup. For example, when President Askar Akayev was ousted in the 2005 Kyrgyz Tulip Revolution, he denounced the regime change as "an anti-constitutional coup" (BBC Monitoring Central Asia Unit, March 25, 2005).

International actors
There have been many studies on how international actors react to regime change, and it has been pointed out that their reaction can have a significant impact on stability and the implementation of elections after the regime change. 22 Therefore, it is very important whether the international community welcomes the regime change, and we will also consider the labelling by the international community in this article.
It is not only the domestic actors within a country who frame political upheavals; international actors also perform labelling in accordance with their political intentions. Thus, international actors who welcome the political upheaval will refer to it as a revolution rather than a coup. Naturally, international actors are not necessarily monolithic. As countries such as China and Russia place importance on the principle of international non-intervention and generally refrain from criticizing domestic political upheavals, we surmised that they likely prefer to avoid the "coup" labelling.
The supporters of a political upheaval do not use the word "coup." Instead, they often frame the event in ways that suggest it represents the will of the people, such as "revolution." In contrast, critics use the word "coup."

Media
Finally, we considered labelling by the media. This is rather complicated as the media serves as an actor for labelling, i.e. when the media supports a political upheaval, it may report that "a revolution has broken out." If they interpret the political upheaval as a negative event, they may focus on the actions of the military and inform the public that "a coup has occurred." Thus, how a political upheaval is labelled can vary among members of the media.
First, the differences between media will overlap to a certain degree with which country the media is from. Typically, Western media will often label an irregular seizure of power by the military as a "coup," whereas Chinese and Russian media will be less likely to make such labelling. For example, in Egypt's 2013 political upheaval, many Western media labelled it a "military coup" (The Independent, July 3, 2013) while Chinese and Russian media refrained from such labelling (Xinhua General News Service, July 3, 2013; TASS, July 3, 2013). Second, for local media in particular, their labelling will vary greatly depending on whether they welcome regime change. In some cases, civilian elites within the media will not merely report on the political upheaval but will be directly involved in it, in which case, naturally, the media will use positive labelling. Thus, the labelling of the media differs greatly depending on their stances, and it is important to note that even local media do not necessarily have the same preferences for labelling as citizens.
Simultaneously, the media serves as an audience targeted by the labelling conducted by other involved parties (as discussed thus far). Thus, it is conceivable that the content of reporting will change as members of the media are swayed by the labelling of other parties. For example, when the Turkish Prime Minister, Adnan Menderes was deposed by the military in 1960, the foreign media initially reported that a coup had occurred. However, influenced by the use of "revolution" by the military and its supporters, it gradually began to also use that term (New York Times, February 19, 1961).

Data on labellings
Next, we empirically showed, through the construction of our dataset, that each actor is extremely strategic in its labelling of political upheavals. This study focused on labelling in all pure coups, revolutionary coups, and revolutions between 1975 and 2014.
The reason for limiting the period to after 1975 is to restrict the revolutions being analysed to democratic revolutions and exclude ideological revolutions (e.g. those of the communist movement). First, in democratic revolutions, the word "revolution" is generally used in a positive context regardless of political standpoint. On the other hand, in ideological revolutions, the connotations can vary greatly depending on the person's political standpoint, and they can be used negatively. Second, most of the revolutions that took place after 1975 were democratic revolutions. The increase in democratization that Huntington identified as the "third wave" began in the mid-1970s, 23 therefore, the starting point for the analysis in this study was set accordingly in 1975. Furthermore, as a result, the analysis of coups is also limited to after 1975. As mentioned in the previous section, even if a regime change is objectively considered a coup, if the actors, such as the military, call it a "revolution," it is only limited to the period when "revolution" has a positive connotation.
The procedure for constructing the dataset is described below, with details available in the Online Appendix. First, in identifying the three aforementioned types of political upheaval, this study relied on existing representative datasets, including Powell (2012), Thyne (2011), SPEED, NAVCO, andGeddes et al. (2014) (see Online Appendix). As a result, we identified 54 revolutionary coups, 52 pure coups and 21 revolutions that occurred between 1975 and 2014.
Next, our dataset targeted four types of actors that label political upheavals: the military, ousted leaders, international actors (such as Western countries, regional organizations, China, and Russia), and the Western media. Here, a couple of caveats: first, we excluded data on labelling by citizens from the dataset. Although citizens are expected to have different preferences for labelling than the other three types of actors, antiregime protests occasionally lack leaders, making it difficult to systematically collect data. Second, media was limited to Western media. As mentioned earlier, the labelling can vary depending on which country the media is from and what political position it holds, so it is not possible to group miscellaneous media together, and it is difficult to systematically collect a sufficient amount of data for Chinese, Russian, and local media. Third, with respect to regional organizations, data were collected only for regional organizations in which the country experiencing the political upheaval was a member.
We collected data on labelling by these four types of actors in the following three items: whether each actor called political upheaval a "coup" or a "revolution," and whether they denied that it was "not a coup." As mentioned in the previous section, we also collected information on the "not coup" labelling, as we believe that coups are negatively connoted and therefore actors who wish to justify regime change have an incentive to avoid them. Typical of Western media, when an actor used multiple labellings to refer to one event, we counted each labelling separately. If the actor made any reference to political upheaval and did not use any of the labellings of "coup," "not coup," or "revolution," we treated it as "no labelling." If the actor made no reference to political upheaval, it was coded as "no reference." To examine the labelling of the four types of actors, we restricted targeted labelling to the two months following the political upheaval. For our sources, we principally used Western press materials, complementing them with Russian and Chinese media and other secondary materials (See Online Appendix).

Revolutionary coups
First, we examined the labelling of revolutionary coups (political upheavals in which the public and military were involved). This mutual involvement implies that they possess elements of both coup and revolution, indicating that their labelling easily becomes complicated. Table 1 shows the summary of each actor's labelling in revolutionary coups. These results indicated the validity of our argument. The characteristic results are explained below. Note that for labelling by each actor, the total value may exceed 100% because they may have multiple labelling for the same political upheaval, such as "not coup" and "revolution." 24 First, regarding labelling by militaries, of 54 cases of revolutionary coups, militaries referred to political upheavals as "not a coup" in 11 cases (20.4%), and a "revolution" in 10 cases (18.5%). This was regarded as distinctive to militaries as there was no single instance in which ousted leaders or international actors used this labelling. The labelling "coup" is also used by militaries in 8 cases (14.8%), but compared to the other actors, only militaries have this number smaller than the other labellings. Thus, when mass protests occurred, militaries were inclined to frame political upheavals as "not coups" or "revolutions." Next, we considered labellings by those who lost power. We identified 16 out of 54 cases (29.6%) in which ousted leaders called the political upheaval a "coup." Again, this number is notable when contrasted with the zero cases of both "not a coup" and a "revolution" labelling, and the fact that militaries referred to the same political upheaval as a "coup" in only 14.8% of the cases. Thus, it is immediately apparent that ousted leaders attempted to frame their loss of power as being due to a "coup" rather than a "revolution" even when the military acted in concert with the public, preferring the negative connotation of "coup." This showed that, as expected, ousted leaders sought to delegitimize these political upheavals.
Therefore, what do third parties call political upheavals? Western countries such as the United States, United Kingdom, and former colonial powers referred to half of these political upheavals as "coups" (27 out of 54 cases, 50.0%), a considerably higher percentage than the other actors. 25 This is predicable in that these countries tend to hold negative views of coups as events that are contrary to democratic principles. Noteworthily, there were no more than 4 cases (7.4%) in which China and Russia referred to these political upheavals as coups. We identified more cases (7) in which they referred to them as "domestic affairs." This result indicated that the international community is not monolithic, and there are differences in how its members label events.

Pure coups and revolutions
How do actors label political upheavals in the cases of "pure" coups (no revolutionary factors) and revolutions (where citizens, rather than the military, take the lead)? It would appear to be comparatively difficult for actors to engage in artificial labelling in cases where it is evident that either the military or public are uninvolved in the change of power. Thus, we particularly focused here on how involved parties legitimize or delegitimize these two types of political upheavals.
First, Table 2 illustrates a summary of pure coups. An unexpected finding was that militaries referred to the changes of regimes as "revolutions" even in archetypical cases of coups lacking revolutionary elements (6 out of 52 cases, 11.5%). This demonstrates decisively that militaries were inclined to frame political upheavals as "revolutions" in order to obtain popular support and avoid being negatively labelled as "coups." Next, we considered revolutions. Table 3 shows a summary of our findings. We particularly focused on the labelling used by ousted leaders. Although the military is rarely involved in political upheavals caused by revolutions, ousted leaders still tended to refer to such events as "coups." It was conceivable that the Western media is influenced by this labelling. The Western media referred to these political upheavals as coups in 5 of 21 cases; the other involved actor that did so was the ousted leaders. Thus, given how difficult it can be to ascertain the goings-on during political upheavals, the Western media is swayed by the labelling used by ousted leaders.
It is noted that a chi-square test and Fisher's exact test to examine whether there were any differences in the distribution of the labelling by the military for revolutionary coups, pure coups, and revolutions showed no significance (p > 0.1). Likewise, the test for the labelling by the ousted leader revealed statistically insignificant (p > 0.1). These results indicate that militaries or ousted leaders tend not to change their labelling, even which type of political upheaval happens.
The results show that various actors have labelled political upheavals according to their preferences and strategies, as we had assumed earlier. Militaries referred to coups as "revolutions," whether they were revolutionary coups or not. Ousted leaders tended to frame these events as "coups" even in cases of revolutionary coups and revolutions. Although the international community and Western media might have appeared, at first glance, to frame events solely based on facts, we observed a tendency for them to be swayed by the labelling employed by other involved actors. Our findings offer a significant contribution to the existing research, which considered coups and revolutions separately, by empirically demonstrating the linkage between the two types of political upheavals. Table 2. Each actor's labelling in pure coups (n = 52).

Illustrative cases
Through the construction of a dataset, we have quantitatively analysed the labelling of political upheavals by various actors. However, our quantitative analysis did not cover labelling by the public. The motivations behind the labelling of the actors remain unclear. Therefore, to complement these points, we examined the interactions between the labellings of actors using three illustrative cases, each of which is regarded as representative and analytically important: Burkina Faso (2014) as a revolutionary coup, Portugal (1974) as a pure coup, and Philippines (1986) as a revolution Table 4.

Burkina Faso (2014): a stolen revolution
The case of Burkina Faso in 2014 is a typical example of a coup occurring in the wake of mass protests, where a serious divide developed between the military and public. Thus, the dispute over labelling between the two groups was intense.
The catalyst for the political upheaval was President Blaise Compaore's proposal to amend the constitution to allow himself a constitutionally prohibited fifth term. Thousands of members of groups opposing the amendment began to demonstrate against it and demand Compaore's resignation. However, when the demonstrations turned violent, Compaore declared a state of emergency, an act which ended up inviting a military intervention. The President was detained by the military, and a provisional military government under Colonel Isaac Zida was established. Protesters participating in the demonstrations criticized this intervention, arguing that the military had stolen their victory (The Independent (London), November 3, 2014).
Thus, the main actors in the Burkina Faso case were the public and military. First, regarding the labelling by members of the public, given that this political upheaval was the mixed product of a popular uprising and a coup, it is noteworthy that citizens did  1, 2014). Thus, Benewende Sankara, an opposition leader, referred to Traore's move as a coup (The Nation (Thailand), November 2, 2014). Furthermore, those who criticized the military's actions as being a "coup" began to stress that the fall of the government had been due to a "people's revolution." Notably, they did not refer to their demonstrations as a "revolution" prior to the military intervention. For instance, an opposition leader described their demonstration as a sort of protest, saying "the people have decided to start a general popular resistance" (The Herald (Glasgow), October 29, 2014). However, once the military took over, they began to assert that their "revolution" was stolen by the military, and they had to take it back (Radio France Internationale, October 31, 2014). What had essentially been a coup came to be plausibly viewed by some as a popularly led change of government. A complicated array of labellings of the event were proposed, not just "coup" and "revolution" but also "stolen revolution." Next, we examined the labelling by the military, the other main actor in this case. Although opposition leaders and members of the public labelled the change of regime as a "stolen revolution," the military that carried out the overthrow of the government referred to this as not a coup but rather a "popular uprising." Zida, for example, justified the military's actions as being necessary to avoid anarchy and emphasized that "this is not a coup but a popular uprising" (International Business Times News, November 1, 2014).
Finally, regarding the labelling of the regime change by the Western media, the situation was heavily influenced by the dispute over labelling between the major parties. Certain outlets referred to it as a "revolution," while others used the term "coup." The Financial Times referred to the regime change as a "coup," saying "the army then staged appeared to be a coup" (Financial Times Online, October 31, 2014). Conversely, the Telegraph, influenced by the labelling used by anti-government groups, reported that Compaore "fell prey to the first successful popular revolution in the history of sub-Saharan Africa" (Telegraph.co.uk, October 31, 2014). Thus, the Western media's labellings reflected the complicated ongoing dispute between the military and public and swayed between the use of the terms "coup" and "revolution." On the other hand, the Russian and Chinese media, reflecting their governments' positions, did not label either a "coup" or "revolution." In particular, the Chinese media, despite the ouster of Compaore, who has adopted a pro-Taiwan policy, simply reported a comment by a Foreign Ministry spokeswoman that "China is closely following the situation in the country" (China Daily, November 4, 2014).
As seen above, the labelling of political upheaval in the Burkina Faso case varied greatly from actor to actor, with the military and public in particular showing conflict. This labelling dispute over political upheaval had a significant impact on the actual political process as well. Namely, the early departure of the new government due to the granting of legitimacy to the anti-military regime protests. In response to the hijacking of the "revolution" by Zida's military regime, opposition leaders immediately staged protests, issuing the statement "The victory born from this popular uprising belongs to the people. In no case can it be confiscated by the army" (Africa News, November 2, 2014). The protests, which were held under the slogan "The soldiers have stolen our revolution," succeeded in mobilizing thousands of people and forced the military regime to step down (The Times, November 3, 2014). Portugal (1974): from coup to revolution From a labelling perspective, the 1974 political upheaval in Portugal is a particularly interesting case. It is generally known as the "Carnation Revolution," but it was carried out solely by the military, and thus should be called a pure coup. In this regard, from the perspective of some scholars that define revolution as a period in which mass mobilization makes an effort to transform the political institution, 26 this transitional event may be seen as a revolution. 27 However, since this article focuses on the point at which regime change occurs, the upheaval itself is a pure coup. 28 Accordingly, this case is worth analysing because it shows how a pure coup can be labelled as a "revolution." First, it is clear from the following background that the 1974 political upheaval in Portugal was a case of pure coup, caused by military dissatisfaction with the regime in the absence of a large-scale protest by the citizens. Military grievances against the autocratic regime of the "Estado Novo" had begun to accumulate since roughly 1973, the result of exhaustion caused by chronic colonial warfare in Africa as well as governmental favouritism towards selected military officers. This culminated in the establishment of the Armed Forces Movement (MFA) by a group of young military officers to change the regime. 29 Finally, on 25 April 1974, the MFA organized a bloodless coup. Considering that the warfare had led to deteriorating economic conditions in the country, the public was also disgruntled with the regime. However, owing to the iron-fisted rule of the Estado Novo, large-scale protests had never occurred. 30 Below, we shed light on the labelling used by the military and public, as well as that of the Western media influenced by them.
After seizing power, the military established the National Salvation Junta led by General António de Spínola. In its official announcement, the military declared that it had "just achieved the most important civic mission in recent years" and stressed that its action was a reflection of the popular will; therefore, it avoided calling its act a "coup" (New York Times, April 26, 1974). Additionally, the military touted the establishment of democracy as the new government's policy and announced that it would hold popular elections in 12 months. 31 Given that the military had no intention to promote the democratization process, 32 these proclamations revealed that its intent was to justify its actions and garner popular support.
Members of the public welcomed the overthrow of the regime in a coup i.e. that the military and public enjoyed good relations during the political upheaval, and there was no rift between their labellings. Notably, the citizens regarded the event as "not a coup" but rather a "revolution," although the citizens had not staged any kind of mass protests. For example, Mário Soares, a leader of the Socialist Party who had been forced into exile because of the former regime's oppression, told Spínola that he "accepted what was thrust upon him by this revolution, and he has done a great thing for his people" (TIME, May 13, 1974). Additionally, people in the street reportedly celebrated the end of the autocratic Estado Novo, chanting "the Happy Revolution" (The Times, May 2, 1974). Thus, the citizens referred to the event as a revolution achieved in cooperation with the military.
This popular reaction was favourable for a military that wanted to avoid the labelling of "coup." The military also began to refer to the event as a revolution. For example, when the interim government issued its future policy programme on 11 June, a military officer reportedly stated that "we carried out the revolution in Portugal because we wanted to get out of Africa" (New York Times, June 12, 1974). Thus, because the citizens who supported the political upheaval framed the event as a revolution and the military who organized it also joined this position, the event has come to be widely regarded as a revolution, although it was unaccompanied by protests and exhibited none of the elements of a revolution.
Finally, the Western media was also influenced by the popular labelling of events. Immediately following the political upheaval, they reported the event as a coup in accordance with the facts (The Times, April 26, 1974). However, when citizens and the military began to call the political upheaval a revolution, the labelling by the Western media also changed noticeably. A passage from New York Times, for instance, illustrates this point; "It took the form of a coup. But with a bit more than a week gone by, what's happened in Portugal is beginning to seem like a revolution" (New York Times, May 5, 1974). The Western media served as an audience for the labellings proposed within the country; as the event, which would appear to have been a pure coup, began to be framed as a revolution, it deftly responded in kind. Afterwards, the Western media increasingly framed the event as a revolution, using terms such as "the April 25th revolution" (The Times, April 26, 1974) and "the revolution in April" (New York Times, June 27, 1974).
The labelling of the event as a "revolution" as described above had the effect of justifying the actions of the military. Immediately after the political upheaval, members of the public did not criticize the coup or the military-led interim government, but defended the military, saying "they are not the crooks that we all know," even though "nobody knows very much about the captains and majors who are the real force behind the coup" (New York Times, April 28, 1974). The jubilation and welcome for the political change were also seen outside the capital Lisbon, with tens of thousands of people coming to hear General Spínola's speech (New York Times, June 1, 1974).
Philippines (1986): a labelling dispute over "people power" Next, we analysed the 1986 People Power Revolution in the Philippines, using it as an example of a revolution. Pro-democracy forces led by Corazon Aquino launched a protest campaign against President Ferdinand E. Marcos after Marcos won the 7 February 1986 presidential election through electoral fraud. These protests gained momentum after being joined by the Reform the Armed Forces Movement (RAM), a military faction, and swelled to over a million people. The revolution was ultimately successful, and Marcos fled into exile.
The military and civilian groups developed favourable relations with each other and were able to work together as a single democratizing force and overthrow the government. It was when RAM leader Juan Ponce Enrile and Chief of Staff Fidel V. Ramos were almost arrested for plotting a coup that the army joined the pro-democracy forces. At that time, Ramos referred to his actions as a revolution and said, "this is not a coup; it is a revolution of the people" (New York Times, February 23, 1986). The army's actions were initially welcomed by the public. One opposition legislator indicated that the army acted in accordance with the desires of the public, saying that "it is clear that Enrile and Ramos are acting on behalf of the Filipino people" (United Press International, February 22, 1986).
However, the public never gave up its leading role in the political upheaval. Following the ousting of Marcos, citizens increasingly framed the political upheaval as an expression of "people power" and a "revolution" (New York Times, March 3, 1986). Aquino's brother-in-law, Agapito (the leader of a pro-democracy group), implied that it was Aquino, not the army, who was the head of the new government, saying "Ramos and Enrile are very brave, (…) But we now recognize Cory Aquino as commander-in-chief" (Christian Science Monitor (Boston, MA), February 24, 1986). For its part, the army emphasized that regime change was the result of the cooperation between the army and public, saying "We have been successfulnot so much because of military power but because of people power" (The San Diego Union-Tribune, February 25, 1986). Thus, although the citizens and military justified the political upheaval by deeming it a "revolution," a rift existed between them.
Rifts were also present between the ousted leader and the democratizing force that removed them. That implied that, while the public referred to the political upheaval as a revolution, Marcos, the ousted leader, deemed it a coup. He stressed that the army's overthrow of his government was illegal and publicly claimed to have been "deposed by a coup" (New York Times, April 10, 1986). He emphasized the popular support he enjoyed and asserted that he had not been ousted by the protests. Although it has been pointed out that the defection of the military is a decisive factor in the collapse of authoritarian regimes, 33 in the case of the Philippines, it was popular protests that enabled the escape of army officers whom the former regime could have presumably arrested and controlled. Therefore, this case demonstrated that, as previously mentioned, ousted leaders will frame political upheavals as coups as a means of delegitimizing the event even when few elements of a coup are present.
The media interpreted this political upheaval favourably, following domestic public opinion. The regime change was referred to as the "People Power Revolution" not only by domestic media outlets but also by Western outlets, which were expected to take a neutral stance. Thus, the media assessed the regime change positively and adopted the labelling of "revolution," following the terminology of the public.
Finally, many countries welcomed the political upheaval and ousting of Marcos, owing to the international community's negative opinion of electoral fraud and the way that Marcos initially responded to the protests. For example, while he did not use the term "revolution," the French prime minister indicated that the regime change was democratic and a reflection of the will of the people, saying that "France is at the side of all people who struggle, the side of Philippine democrats, and the side of Cory Aquino" (New York Times, February 26, 1986).

Conclusion
This study constructed the first comprehensive dataset and systematic case studies of how various domestic and international actors labelled political upheavals in terms of the legitimization/delegitimization of political changes. The results showed that even the same event could be labelled in a highly convoluted manner. Considering that the term "coup" had negative connotations, those who wished to delegitimize a political upheaval (such as ousted leaders and domestic and foreign groups with close ties to them) tended to use it as a labelling device to criticize the military or opposition groups responsible for it. In contrast, "revolution," a term with positive connotations, tended to be used by the military and pro-military civilians to emphasize the legitimacy of a political upheaval. We empirically demonstrated our argument through data analysis and case studies. The parties involved in political upheavals strategically labelled the events in cases of revolutionary coups and those of pure coups and revolutions. Scholars, as well as the media, are strongly influenced by these complicated political labellings by the involved actors, and this results in the distinction between coups and revolutions becoming extremely difficult.
For future directions of research, it would be promising to examine the influence of this conceptual linkage for labelling on actual political processes. The analysis in this article deliberately focuses on the point in time of political upheaval, but the success or failure of labelling by different actors and its effects are also important political phenomena. In particular, the dataset of labellings constructed for this study would be useful for future quantitative analyses that treat the labelling of political upheavals as an independent variable. For example, it might be possible to analyse the question of how and to what extent international reactions to political upheavals (sanctions or assistance) are influenced by the involved actors' labelling of the event. In addition, as mentioned in the case studies section, the labelling of political upheavals may also influence the political process of the new regime; for instance, the duration of the regime, or particularly in cases of military regimes, the length of time until a transition is made to civilian rule. Thus, the new perspective introduced by this study makes a relevant contribution to the research on political upheavals and a broader research field regarding the duration of political regimes.