“I Know Which Devil I Write for”: Two Types of Autonomy Among Czech Journalists Remaining in and Leaving the Prime Minister's Newspapers

This paper examines two different understandings of professional autonomy among journalists currently and formerly working at Mafra, a Czech media house acquired in 2013 by Andrej Babiš, who in 2017 became the Czech Prime Minister. We build on existing research of local trends in media ownership and journalistic autonomy to ask the following questions: What differentiated the experience of journalists who exited the organization after the ownership change from that of those who stayed put? How did the two groups understand professional journalistic autonomy? Based on the thematic analysis of twenty semistructured interviews with ten journalists who stayed in the media house after Babiš's acquisition and ten journalists who left, we argue that in the journalists’ narratives, the two decisions reflect two different notions of autonomy: autonomy-as-a-practice and autonomy-as-a-value. While our findings add to the scarce empirical research on journalists’ lived experiences of the region's mediascape marked by growing comingling and concentration of political, economic and media power, we also suggest that the autonomy-as-a-practice and journalists’ agency should be further studied as a possible way how to perform and promote journalistic autonomy even in illiberalizing contexts—in Central and Eastern Europe and beyond.


Introduction
Rebecca writes for Marfa, a media company that has belonged for the past few years to a trust maintained for the sitting Czech Prime Minister. Although Rebecca herself was never asked to write anything that would help the owner's political career, she could see how the newsroom had gradually turned into a crew of obedient newsworkers. "It is more 'a fish stinks from the head' thing," Rebecca says. "The professionals are leaving, and the second league is sneaking up." Rebecca thinks that the newcomers are "disgusting," obsequious, career obsessed. Nevertheless, she is still part of the team; leaving seems harder and harder along with the ever-growing shame that she and her colleagues are supposed to feel (according to the broader journalistic community). She also thinks it would be hypocritical to leave five years after Babiš bought the newspaper. Instead, she has developed and adopted a notion of journalistic autonomy she can stick to. "We're dancing on a pyre." Rebecca's complex situation is not exceptional. In 2013, the Czech businessman Andrej Babiš purchased what is now one of the biggest Czech media houses, Mafra. The businessman soon became a deputy in parliament (2013), then the Minister of Finance (2014), and then Prime Minister (2017). Given the owner's multiple social roles, the acquisition also further strengthened the re-established alliance of political and media power in the Central and Eastern European region (CEE) (Castro-Herrero et al. 2016;Sparks 2000;Stetka 2012). Some journalists working in the two major national newspapers that belong to the Mafra group perceived the power concentration to be inappropriate, casting a shadow over the news outlets' professionalism, autonomy, and integrity both from within and from an outside perspective, and either left immediately (see Hájek and Štefaniková 2014) or later, often in groups and waves and often speaking publicly about their reasons (see Mapping Media Freedom 2018). Exact numbers are unobtainable; based on our interviewees' estimates, about 100 of the presumed 400 newsworkers left between 2013 and 2018, allegedly because of the new owner 1 , and many others have been considering leaving.
In this paper, we look into the experiences and understandings of autonomy among journalists who left (leavers) compared to those who stayed in Mafra after its acquisition by Babiš (remainers). Specifically, we address the following questions: What differentiated the experience of journalists who exited the organization after the ownership change from that of those who stayed put? How did the two groups understand professional journalistic autonomy? While media theorists have been worried about the ever-growing media concentration and lack of pluralism in Central and Eastern Europe (Castro-Herrero et al. 2016;Örnebring 2012;Sparks 2000;Stetka 2010Stetka , 2012Waschková Císařová and Metyková 2015), empirical research on these trends and their consequences have been surprisingly scarce. At the same time, previous theoretical debates and scholarly research suggest that autonomy is not a monolithic value carved in stone. There has been an ongoing discussion of the conditional and contextual nature of journalistic autonomy and scholars have been calling for more empirical research into the diverse contextual factors shaping autonomy (Altschull 1997;Hughes et al. 2017;Sjøvaag 2013). This study investigating one case that is typical of the CEE media system, where "news media have often found themselves under combined pressure from both political elites and economic forces" (Stetka 2012: 434), thus seeks to add to the understanding of journalists' lived experiences of the region's mediascape, its current trends, and the notions of journalistic autonomy in the context of contemporary democratic deconsolidation/illiberalization of democracy in CEE (Surowiec and Štětka 2020). Our in-depth qualitative empirical approach also enables us to move beyond the often quantitative and general insights into the consequences of media ownership change (e.g., Castro-Herrero et al. 2016;Kostadinova 2015).
In the first part of the paper, we define our research problem by referring to the trend of strengthening alliances between political and media power in the CEE region (Castro-Herrero et al. 2016;Stetka 2012) and previous research on journalistic autonomy. In the second part, based on twenty interviews with ten leavers and ten remainers, we reconstruct the "point of friction" of the two groups. Building on Carpentier and Trioen's (2010) notion of a gap between objectivity-as-a-value and objectivity-as-a-practice, we argue that the main difference between the groups lies in their diverging notions of journalistic autonomy. While the leavers insist on (organizational) autonomy-as-a-value, the remainers embrace (individual) autonomy-as-a-practice.

Autonomy as a Practice, Agency, and Tactics
Autonomy-"the freedom to speak and publish, and freedom from interference in that activity" (Sjøvaag 2013: 156), i.e., "the degree of self-governance within the profession, and the extent to which the profession is independent of other societal institutions" (Örnebring 2013: 39)-is very often conceived as a relative value that to some extent is (or is not), but definitely should be, granted to individual journalists. Sjøvaag (2013) argues that this understanding of journalistic autonomy primarily in terms of the values imposed on reporters by the organizational (and social and political) setting reflects the traditional primacy of structure over agency in journalism and media studies. In turn, the understanding manifests itself in the lists and models of internal and external factors affecting-restraining or increasing-autonomy (Altschull 1997;Mellado and Humanes 2012;Nygren 2012;Skovsgaard 2014). The level of journalistic autonomy is supposed to be determined by the level of democracy, geographical location, government regulations and licenses, political parties, trade unions, interests groups, media laws, and lobbyists; publishers' interests, advertisers, market pressures, profit expectations, audience preferences; managerial and editorial decisions, newsroom policies, media types, media size; work routines, professional norms, codes, deadlines; informal feedback and discussion with colleagues, friends, and relatives. In the next section, we further expand upon one of the most relevant findings of previous research that journalists working in full democracies and in Western countries tend to enjoy more autonomy than their colleagues working in transitional, hybrid, and authoritarian regimes (Reich and Hanitzsch 2013).
Although some authors (e.g., Mellado and Humanes 2012) focus on factors influencing autonomy specific to individual journalists, such as experience, age, (editorial) position within the media environment, and the assignment of a news beat, not much attention has been paid to the role of journalists' agency (Giddens 1984;Sjøvaag 2013;Usher 2013). In other words, previous research and theory-interestingly, often based on journalists' self-declarations of perceived levels of autonomy-have rendered journalists somewhat passive consumers of an available level of autonomy that is shaped solely by factors external to the journalists' minds, intentions, interests, and will.
In this paper, we build, first, on Sjøvaag's (2013) critique of journalistic autonomy conceptualized as a positive and negative right and value. Based on empirical findings, we further develop her notion of autonomy as context-dependent, situational, fluid, moving, continuously adjusted, and, above all, performed (see also Couldry 2004;Hughes et al. 2017;Lauk and Harro-Loit 2017). Journalistic autonomy in CEE might not have the same tradition as in Western countries, particularly the United States (see Josephi 2012). In CEE and other posttransitional areas, the discursively constructed roles of journalism (see Vos 2017, 2018) tend to fall under the more active categories of advocates, participants, interpreters, mobilizers, and the like. Journalists in posttransitional countries feel the need to fulfill a particular mission in the young democracy, and thus take a "pro-active role to play in advancing democratic transformation, guiding and educating society and promoting justice" (Voltmer and Wasserman 2014: 190; see also Jakubowicz 1998). Gross (2004) observes that the roles in and relationships to society of CEE media are particularly intertwined with the cultures of their owners and other institutional constituents that, in CEE, are difficult to define. Our discussion of autonomy is set not only in this quasianomia and role uncertainty but also in specific social arrangements-under the condition of unprecedented power concentration that we further describe in the next section. By showing how autonomy is negotiated on an everyday basis, we look into whether, how, and what kind of journalistic autonomy can thrive even in difficult circumstances-i.e., how journalists can "take it" and perform it when the structures and others' interests restrict it (see Hughes et al. 2017).
Second, we borrow Carpentier and Trioen's criticism of objectivity as a onedimensional concept working as an object of desire that is impossible to reach rather than as a practice (Carpentier and Trioen 2010) and apply it to autonomy. According to the authors, objective reporting remains an unattainable horizon and an ideological construct that everyday journalistic practices can never fully meet. They thus distinguish between objectivity-as-a-value (a part of journalistic ideology) and objectivity-as-a-practice (what journalists actually do) and speak of an unbridgeable gap between the two. Previous research (Hanitzsch and Vos 2017;Josephi 2012;Skovsgaard 2014) suggests that there is a similar gap between other values/normative aspirations/self-perceptions and actual practices, including autonomies: "The situation of journalists worldwide shows that autonomy, also in western countries, remains an unattainable goal" (Josephi 2012: 482). In the empirical section below, we show how the difficult circumstances of newspapers owned by a powerful politician give rise to two versions of autonomy: autonomy-as-a-value and autonomy-as-a-practice.

Conditions of Autonomy: CEE Trends in Media Ownership
Research focusing on Central and Eastern Europe 2 shows that the postsocialist countries in the region are characterized by tight links between local investors, political parties, and the currently dominant media groups (e.g., Bairett 2015;Castro-Herrero et al. 2016;Örnebring 2012;Sparks 2000). Political elites use the media to avoid negative publicity, to win voters' support, and to defeat political rivals. The pattern of politicians interfering in privately owned media-and often the public service media as well-is clearly recognizable in Poland, Hungary, Slovakia, and Croatia (Metyková and Waschková Císařová 2009;Šimunjak 2020;Surowiec and Štětka 2020). There is also a clear trend since 2006 of a gradual withdrawal of foreign companies from the national markets and their replacement by local media tycoons with multiple nonmedia interests who concentrate media ownership (Němcová Tejkalová et al. 2015;Stetka 2010Stetka , 2012Waschková Císařová and Metyková 2015). National owners have become prominent not only in Czechia, but also in Bulgaria, Romania, and Latvia (Kostadinova 2015).
This context makes the curious case of Mr Babiš, who simply purchased one of the biggest media houses in the country, a relatively direct and overt, yet fairly typical, an example of the power concentration and control over journalism in the region. It is also a good example of the gradual "italianisation" of CEE media (Splichal 2000)-the process of putting significant parts of media systems under control by the leading political parties or politicians-and even their "Berlusconisation," meaning that "the mass media are monopolized by politicians and businessmen and used for their personal, political or business purposes exclusively" (Wyka 2007: 3). Indeed: Financial Times (Rohac 2017) and other news outlets have dubbed Babiš the Czech Berlusconi.
The "concentration of media ownership in the hands of a few oligarchs," including the those of the Prime Minister, "using their fortunes to buy newspapers in order to reinforce their influence," and related loss of autonomy and intimidation of investigative journalists has concerned organizations such as Reporters Without Borders (RSF), which, in its World Press Freedom Index 2020, ranked Czechia six positions lower than two years earlier (from a fortieth place in 2020 to thirty-fourth in 2018; Reporters Without Borders, 2018, 2020). Other CEE countries such as Slovakia, Hungary, Poland, and Romania dropped as well.
The trend of ownership concentration can have various and far-reaching consequences and is often seen by media scholars writing about CEE as the biggest threat to journalistic autonomy. The local owners' control over local media can weaken investigative journalism, undermine the watchdog function of the media (e.g., Örnebring 2012;Stetka 2012), and, in turn, do away with the scrutiny over political processes in CEE countries experiencing illiberal drifts (Šimunjak 2020). It can result in widespread self-censorship: "the political and economic pressures … have rendered self-restraint among journalists and editors an integral part of all the [CEE] media systems" (Schimpfössl et al. 2020: 6). The CEE journalists "know exactly what to report, what to omit, and how to advance their careers" (Schimpfössl and Yablokov 2020: 29). Importantly, this disquieting trend means that journalists have limited choices to leave their employer for another where they would enjoy significantly more autonomy. Data from the Worlds of Journalism study gathered between 2012 and 2014 (Hájek et al. 2015;Němcová Tejkalová et al. 2015) show that Czech journalists do feel the rising influence of owners upon their news organizations, and also indicate decreasing job satisfaction. The journalists' situation is further complicated by their poor collective bargaining position given by underdeveloped labor organizations (Castro-Herrero et al. 2016).
A more comprehensive analysis of the journalists' lived experience of the abovementioned trends and their professional autonomy is still lacking (Kostadinova 2015). In a forthcoming publication, we analyze the perceived challenges to journalistic autonomy stemming from ownership change and concentration (Waschková Císařová and Kotisova, forthcoming). Here, we want to take a step further and ask what differentiates the experiences of journalists who exited the organization after the ownership change from those who stayed put, and how the two groups understood journalistic autonomy.

Methods
To answer the research questions, we conducted a case study, that is, "a study of the occurrence of a phenomenon" (Czarniawska 2014: 21), the phenomenon being the CEE trend of media and political power concentration. The case is constituted by the Mafra publishing house which, as we argued in the previous section, epitomizes the trend; thus, we assume it is general and typical (albeit extreme), rather than specific. Mafra, one of the biggest media houses in the country, is comprised of two major national newspapers, a TV channel, magazines, weeklies, a free daily, and a radio station (Mafra 2019). We focused on journalists working in the two paid newspapers. While the daily Mladá fronta DNES (MFD) is the most widely read and bestselling quality newspaper in the Czech Republic, Lidové noviny (LN), founded in 1893, is the oldest Czech daily still in print and enjoys historically high prestige. In 2017, an amendment to the Czech Conflict of Interest Law (2019) banned politicians from controlling the media and the PM subsequently transferred his media activities to a trust fund (Transparency International 2018). This maneuver did not stop prominent journalists from leaving the two newspapers; the walkouts go on.
Between the autumn of 2018 and the end of 2019, we conducted semistructured interviews with twenty actors involved in the case in total: ten journalists who, at the time of the data gathering, worked in Mafra (remainers), and ten journalists who had left between 2013 and 2018 (leavers). Three of those who stayed at Mafra and six of those who left worked for LN, the rest worked for MFD. Because of the sensitivity of the research problem, we first used our previous research contacts to recruit the interviewees and then continued with snowball sampling. In this way, we reached a diverse sample of former and current Mafra journalists in terms of age (the interviewees were in their twenties to fifties), work locations (they worked both in the central newsrooms of LN and MFD in Prague and the various regional newsrooms), positions (editors, reporters, heads of sections, journalists from specialized departments), the length of experience in journalism (from three to about thirty years), and the number of years spent in the organization (between three and over twenty). Many of our interviewees were thus "old hands" with extensive institutional experience and memory. However, the sampling process resulted in a minor gender imbalance-nine remainers and five leavers were men. Other limits of the sample stem rather from the ethnic and educational homogeneity of the Czech journalistic field: all our interviewees were white Czechs and a vast majority of them were university educated, albeit not always with degrees in journalism.
The interviews consisted of narrative and loosely structured conversations between one of the authors and one of the journalists (see e.g., Arksey and Knight 1999) and were guided by a list of several open-ended questions, including inquiries about the interviewee's experience with the ownership change (if applicable), working conditions, and journalistic professional values and their feasibility in Mafra. The interviews started with obtaining interviewees' informed consent and lasted between forty and ninety minutes. They were conducted in Czech, recorded, and transcribed. Because of the sensitivity of the data (e.g., political opinions, beliefs, attitudes towards colleagues, management, the owner, and their practices), we anonymized the interviewees not only by changing their names (the remainers' pseudonyms start with "R," the leavers' with "L") but also by randomly changing their genders.
We analyzed the interviews using thematic analysis (Braun and Clarke 2006). The two ways of understanding autonomy were not conceptually predetermined. Instead, the analytical process, based on thorough coding of the interview transcripts, sought to identify the similarities and differences between the two groups by (re)constructing twenty-six concrete codes such as "newsroom culture," "stigma," "direct influence," "the importance of collective," and then putting them together into several general themes: changes in the newsroom after the ownership change; the decision to stay put or leave; the construction of professional boundaries. The last theme (including codes related to professional values) led us to identify the different notions of autonomy as the strongest line between the leavers and the remainers (while there are also economic and other factors-see the section Reasons to remain and the Stockholm syndrome). Throughout the process, we learned that the critical tone of previous research on the regional intermingling of political and media power does not justify the common public condemnation of the individual journalists staying in Mafra. In other words, we believe that some public criticism is necessary but should not be aimed at individuals en bloc.
It is important to stress that both groups' discourses on autonomy point to the building of professional boundaries rather than to an unproblematic, objective reality. This is to say that while the remainers believed that what they performed at Mafra was appropriate journalistic behavior, the leavers saw the developments in the media house as "over the line" and left (see Bishop 1999;Carlson 2015). Each group had its own understanding of what autonomy means, i.e., what counts as journalismwhat norms, values, roles, and practices professional journalism embraces-and what does not. Consequently, we do not investigate levels of autonomy as such, but rather the journalists' more or less genuine beliefs and stories about them.

(In)Direct Influence as a Shared Experience
There is a popular opinion that when a politician owns a media outlet, she or people close to her routinely call the newsroom to give orders-indeed, that the journalists are "Babiš's whores" (Leo). In this section, addressing the question of what differentiates the experiences of journalists who exited the organization after the ownership change from those who stayed put, we build on existing work on newsroom socialization (Breed 1955;Gravengaard and Rimestad 2014) to suggest things are much more complex than that.
The experience of the two groups of interviewees was not significantly different. Both the leavers and the remainers were familiar with moments when the owner exercised power over the newswork in a relatively direct manner, and at least three of these are publicly known. The first affair was the "Přibil case": anonymously released recordings of Babiš's call with MFD journalist Marek Přibil, in which they discussed a campaign against Babiš's then political opponent. The second is known as the "Olaf case": LN reporters published a piece about the European Anti-Fraud Office's report on the PM's subsidies fraud that obscured the content of the report. The third affair is known as the "orphans case": the online version of LN published, without any journalistic vetting, a text received from the Office of the Government and written by a fake doctor who supported the PM's position on the acceptance/refusal of a group of Syrian refugees.
However, the vast majority of both leavers and remainers experienced indirect rather than direct influence. This means that the owner's political/economic power was mediated by the middle managers who "filter, negotiate, and redistribute these influences to their subordinates, and in so doing, they might actually render the original source of influence invisible to individual journalists" (Hanitzsch and Mellado 2011: 421). This indirect control is also based on newsroom socialization "by osmosis" (Breed 1955;Gravengaard and Rimestad 2014). The indirect control by socialization is powerful and very frequent (Gans 2004). Typically, "[newsroom] policy is covert, due to the existence of ethical norms of journalism … the boss does not 'command'; the direction is more subtle." (Breed 1955: 327, 329) Perhaps because of previous professional socialization, the ownership change was a "shock" both for remainers and leavers (Ruth, Rachel, Luke), the "9/11 of Czech journalism" (Luke, Leo), causing sudden damage to the reputations of the newspapers and stigmatization of the Mafra journalists (see Harvey et al. 2017). Apart from the reputation damage, the new top management-agricultural "bumpkins" in "ill-fitting suits" (Leo) who, until 2013, "bred chicken" (Laura)-shuffled the managerial personnel, such that the new editor-in-chief of LN was a "good friend of Andrej Babiš" (Lucy). The new leaders and their managerial culture attract and provide career incentives to opportunistic like-minded newcomers with less prestigious education and experience but high ambition (Lucy), and are abused by the new management (Laura). To put it differently (Gravengaard and Rimestad 2014;Gans 2004), the newcomers were more willing to socialize, which made the direct corrections/control over their newswork superfluous (and the level of their autonomy irrelevant). The newsroom stopped acting as a team. Robin said that the employees do not know each other anymore: "I have closed myself off in a cosy bubble and don't feel the need to be friends with the people who came there recently," while noting that "before the gossip worked much better." Our interviewees almost unanimously agreed that due to the changes in organizational culture, both newspapers were more and more loyal to Babiš (Lee), so that government-related investigative reporting was practically ruled out for the Mafra journalists. Lawrence remarked that "when you are from Mafra, no one wants to give you any compromising material on Babiš" (see Stetka 2012;Örnebring 2012). The newsroom personnel started practicing negative self-censorship: "looking for something [bad] about ANO 3 instead of covering it up, to prove that [they enjoy professional autonomy]" (Ruth) (see . Likewise, Ryan stressed that although "a journalist shouldn't apply a double standard-be it Babiš, Fiala, Okamura, Zeman 4 , s/he shouldn't distort the information," he felt some colleagues were disproportionately critical of Babiš. This concurs with Gans' (2004: 252) finding that when feeling pressure, journalists sometimes tend to "seek out news with harmful publicity for the exerters of pressure."

Reasons to Remain and the "Stockholm Syndrome"
If the experience of the two groups did not significantly differ, what made half of our interviewees stay, and the other half leave?
For most, making the decision to leave was a long-term process. The journalists slowly moved closer and closer to the exit with the owner's growing influence, every new scandal, the ever-worsening loss of reputation, the increasing opinion mismatch between the leavers and the management, the resulting personal discomfort and dissonance, and, perhaps most importantly, tension stemming from deteriorating self-image: I was writing a text message to my wife that it wasn't possible anymore, that I'd need to quit, otherwise I'd feel like an idiot or, or just stupid. … Suddenly I had the impression that it was too much. It wasn't that someone would put pressure on me, it was rather that the pressure was growing and growing inside me. (Lee)

Each of us had a different threshold, in the sense of what one can bear. (Lisa)
The final decision to leave was apparently a very intimate step to resolve emotional dissonance. However, as we shall show in the next section, the threshold, the "last straw," the boundary was defined by each individual journalist's notions of professionalism and autonomy, and remainers and leavers made sense of their experience and its place within their notions of professionalism differently.
Somewhat surprisingly, the remainers were prevented from leaving first of all by their perceived individual professional autonomy, i.e., independence from newsroom pressures from managers and colleagues (Örnebring et al. 2016). They appreciated that they were allowed and encouraged to come up with their topics and compose articles exactly as they wanted. Therefore, they were able to maintain their moral integrity: I was telling myself, so what, whether I wanted to be here, whether the line had been crossed or not, and then I evaluated [the situation and concluded] that it had not, yet, that the degree of self-realization and autonomy is high. (Ria) Although our interviewees were mostly granted this individual autonomy, at times they needed to assert it on their own. Interestingly, individual professional autonomy was mentioned and appreciated equally by those who remained and those who left. In the next section, we will further explore the importance the two groups gave to individual autonomy and organizational autonomy-that is, the independence of the organization from external pressures (see Örnebring et al. 2016).
The second most common reason to stay was the impact that the interviewees, especially those working in the widely popular MFD, enjoyed: they felt their work was more meaningful than work in a highbrow news outlet.
Third, the interviewees indicated persisting identification with Mafra. They did not want to leave their professional groups, the team spirit, shared history, and "pubs" [parties at nearby restaurants]. "The life of a journalist is not enacted at … LinkedIn and these things, but in a pub," Raphael said. Rebecca expressed "gratitude to the newspaper" for nice memories. Moreover, Rebecca, like some others, was somewhat cynically curious about the company's future development: "I am going to stay to see how it works out, it's an incredible experience that I wouldn't get elsewhere." This is to say that the journalists' professional curiosity encompassed even their own situation.
Of course, there was also a range of more negative reasons preventing the journalists from leaving their secure positions, notably the "mortgage factor" (Raphael): financial and family commitments. Vice versa, some of the leavers admitted that they could afford to leave because they were young with no commitments, got another job offer, or had a partner who could support their family. Those working in regional newsrooms also considered the lack of alternative journalistic jobs in their regions; the older employees felt more insecure in the labor market. The importance of both journalists' individual economic conditions and larger trends such as the growing precarity of all newswork for "morality" is aptly expressed by Klyueva and Tsetsura's (2015: 21) observation that "poor economic conditions tend to diminish human dignity" and to compromise journalists' ethical standards, both professional and personal.
Finally, staying with the company was motivated by verbal violence, mistrust, and other unpleasant reactions from the journalistic community and the public towards Mafra newsworkers and the very existence of Mafra journalism. The remainers (Rebecca, Rachel, Robin) felt either energized by the circulation of "unfounded and exaggerated myths" (Rebecca) surrounding Mafra journalism, or ashamed and tired, or both: I … it rather motivates me, I have to say. Well, of course, I feel sorry and it annoys me. (Rachel) The effect of the mixed feelings was the desire to stay in the newsroom and fight for the good name of Mafra. As Ryan explained, "Discussing the hate speech binds the team." Lee called it a case of "Stockholm syndrome" and defined it as a situation when people who are tired of being consistently cast a slur upon for working there, … and suddenly there is a situation that you need to deal with, and the Stockholm syndrome manifests itself by being a bit more indulgent and by emphasizing the good things and sidelining the bad ones. And … Even though it's very personal and bothersome, the assaults from outside, you say "go fuck yourself, it's not like this, we're doing a good job here, no one influences us." The fact that most of the remainers did consider leaving too-mainly in 2013 and then whenever they learned about one of the three above-mentioned affairs (Přibil, Olaf, Orphans)-gives some credence to Lee's metaphor.
"Having the PM as the Owner Doesn't Mean we are bad Journalists" If the members of both groups cherished individual professional autonomy yet half of them left to be more autonomous from the owner, how did they differ in their understandings of autonomy? Or, what kind of autonomy did they value and need to perform their profession?
The remainers very often criticized the general trends in Czech media ownership and related partisanship. Ruth believed that Czech journalism was "much more activist than it had been" and the feeling that "now we need to direct the society in a way, systematically, intentionally" shifts the profession such "that it's not the dry, correct, totally objective journalism, but it's activist, commentating, coloured in a way"which is typical for CEE journalism (Gross 2004;Jakubowicz 1998;Voltmer and Wasserman 2014). For the remainers, there was no such thing as fully autonomous journalism in a context where all media are owned by one oligarch or another. Consequently, Ryan, Rahim, Ria, Richard, and Leo were disillusioned mainly by the fact that many of the country's influential media firms were owned by local businessmen-and that, therefore, there was nowhere to go. According to Ryan, one of the most problematic consequences of the recent developments was that many good journalists left [the mainstream media] for totally marginal news outlets. And they can be thumping their chest now for how independent they are, but almost no one reads them so that in the end the society is terribly beaten (Ryan).
Contrarily, Lisbeth, Laura, and Lucy found the oligarch with the strongest political interests in the country-Babiš-much worse than other media-controlling businessmen.
These divergent views provide a background to understanding the difference in the two groups' notions of autonomy. Most of the remainers-sceptical about the possibility of structurally granted autonomy (see Sjøvaag 2013)-stressed individual professional autonomy, i.e., independence from newsroom pressure from managers and colleagues (Örnebring et al. 2016). If someone ever did succeed in interfering in their work, they would allegedly leave. This was precisely the boundary line which they did not want to cross. Consequently, Richard believes that the sole fact that the owner is also the PM does not necessarily mean their journalistic performance is bad. A few remainers pushed the notion of autonomy to a yet more individual level and were ready to ask for it, take it, and perform it under any circumstances. Such a notion of autonomy well corresponds to the performed nature of the journalistic semiprofession that is defined by journalistic practices (Couldry 2004;Ryfe 2009). Rebecca claimed that the crucial point is "whether I can resist these people, so that I don't let them meddle in what I write," and added, "I know which devil I write for, but the reality is a thousand times more complex." Robin even said that I very much dislike occupying myself with how someone from outside presses it there. … I don't care how the environment around me is de-normalized, I want to do my job in a good, careful way.
Correspondingly, the two remainers who faced attempts to influence their work (e.g., a phone call from people close to Babiš or the Mafra managers) reacted with open, seemingly naïve insistence on truthful reporting and independence, i.e., active advocacy of journalistic principles (see Mishra and Spreitzer 1998), thus confronting the "influencers" with a merciless mirror 5 . This assertion of individual responsibility for the performance of good journalism also reflected itself in their attitude to Přibil (see above), the MFD journalist who was apparently ready to act on the PM's order. Rose, for example, simply saw him as a man who unbelievably compromised basic journalistic values and did not withstand the political pressure, which is there, in any media outlet -whether we like it or not.
On the contrary, the leavers understood the Přibil case as evidence of a systemic flaw, the flaw being the fact that Mafra is owned by the PM. To them, the level of individual professional autonomy and its performance (Örnebring et al. 2016) was less relevant; they understood autonomy rather as a value, a professional ideal that shaped the construction of a professional self: The only thing that works in this profession is that you build up a name … and it takes years and then you need to keep it. … And this was the problem with Babiš. (Laura) Therefore, they understood walking out as a "gesture" (Lisbeth) of civic courage, adherence to principles and ideals. By leaving Mafra, they wanted to improve their professional self-image (Lee; see above), to give an example to their children (Lisbeth), and, above all, to promote the notion of journalism as a set of certain ideals and principles. Correspondingly, what they lacked under the new Mafra ownership-because they found it more relevant than individual autonomy-was organizational autonomy: the independence of the organization from external pressures (see Örnebring et al. 2016).

Discussion and Conclusion
It would be easy to dismiss journalism in today's Mafra as "postjournalism" or "pseudo-journalism" (Kovach and Rosenstiel 2001) because the organization as a whole clearly fails to maintain independence from Babiš, yet cannot avoid reporting on his activities. In this paper, we suggested an alternative reading of the situation based on our findings and analysis grounded in the context of the oligarchized media market.
We asked what differentiated the experiences of journalists who exited the organization after the ownership change from those who stayed put, and how the two groups understood professional journalistic autonomy.
We found that the leavers were neither subjected to influence from the owner more than the remainers nor were the remainers isolated from it. The two groups' experiences did not differ: both the groups knew that the owner (directly) or his people (indirectly) at times tried to interfere in the newswork. We interviewed journalists who had never faced any interference themselves (yet some of them left) as well as others whose work had been under pressure from the management (yet some of them remained and insisted on individual autonomy). Interviewees from both groups stressed the indirect importance of the shift in organizational culture (see e.g., Breed, 1955).
Leaving and staying seemed to be two different reactions to the trends in CEE of (1) increasing concentration of media ownership (2) in the hands of local media tycoons with other business and political interests, (3) resulting in a lack of media market plurality (Němcová Tejkalová et al. 2015;Šimunjak 2020;Stetka 2012;Stetka and Hajek 2020;Waschková Císařová and Metyková 2015). The remainers acted upon the fact that other media outlets were owned by other local businessmen. They believed that political and economic factors were natural aspects of newswork (Hanitzsch and Mellado 2011) and that "Every newspaper has a policy, admitted or not" (Breed 1955: 327), meaning that they would not have significantly better working conditions and professional autonomy elsewhere. Therefore, they kept trying to practice their notion of good, autonomous journalism within the milieu of a copybook example of power concentration. When exposed to the owner's indirect influence, they reacted with active advocacy of autonomy and independence (Mishra and Spreitzer 1998). By comparison, the leavers believed that, first, not all businessmen were the same (or: some were not the PM), and second, there were still a few relatively independent media organizations, including the public service media.
This important distinction can be understood in terms of different notions of autonomy. Both groups of journalists made an implicit distinction between individual professional autonomy, which in practice they usually enjoyed, and organizational autonomy in Mafra, which was not, and could not be, achieved under the new owner. The remainers stressed individual, practically construed autonomy and were ready to stand up for it. Like Gans, who believes that "Press freedom … is total only for the journalists themselves" (Gans 2004: 269), they saw individual autonomy as the only autonomy that they could realistically insist upon. The leavers, in contrast, valued a more general and abstract notion of autonomy as a principle that distinguishes 'good' journalism from 'bad' journalism (see Ryfe 2009) and their walkouts were gestures of protecting it.
We have already interpreted this difference in terms of individual versus organizational autonomy (Örnebring et al. 2016): the independence of an individual from internal newsroom pressures versus the structural independence of the organization from certain political and economic interests. However, the groups are also distinguished by how they constructed themselves as agents of autonomy, or subjects of a structure (Giddens 1984). While for the leavers autonomy either existed or not (and they could enjoy it or not), the remainers were ready to keep actively constructing spaces for autonomous tactics even in a milieu where organizational autonomy was unthinkable (Hughes et al. 2017;Usher 2013). To borrow Nico Carpentier and Marit Trioen's distinction between objectivity-as-a-value and objectivity-as-a-practice-describing the gap between journalistic professional ideology and journalistic practices (Carpentier and Trioen 2010)-the remainers embraced autonomy-as-a-practice, while the leavers promoted autonomy-as-a-value (cf. also Hanitzsch and Vos 2017).
Interestingly, while journalists in posttransitional countries often play a pro-active role in advancing democratic transformation (Jakubowicz 1998;Voltmer and Wasserman 2014), some of the remainers condemned any kind of journalistic "activism." They believed that by staying with Mafra and practicing journalism as autonomously as possible under the organizational circumstances, they avoided assuming the partisan, "counterpower" stance (Gross 2004) allegedly widely performed among nonMafra journalists. We can see this as a desperate attempt to fit into the boundaries of journalism or as a step towards the ideal of absolute neutrality.
The findings, therefore, support Sjøvaag's notion of autonomy as contextdependent, situational, moving, continuously adjusted, and, above all, performed (Sjøvaag 2013). As Lauk andHarro-Loit (2017: 1968) write, Journalistic autonomy depends on the moral awareness of the members of the profession and on the ability to reflect critically on everyday practices. It requires moral judgment in cases where professional autonomy is under threat. This is exactly how the remainers constructed themselves in the interviews: as everyday morally aware protagonists. Their approach stands in stark contrast to previous research suggesting that individual journalists often give up personal autonomy and are instead loyal to management (Lauk and Harro-Loit 2017). However, even the leavers critically reflected on their individual autonomy in light of the conditions under which they worked and for whom they worked (Lauk andHarro-Loit 2017: 1961): they acted based on the traditional understanding of autonomy as a positive and negative right and value; they brought in the context by distinguishing between the relative "evils" linked to particular oligarchs and were ready to adjust their judgements based on the oligarchs' ever-changing political and economic whims and orientations. This concurs with Usher's (2013) finding that agency is much more present in newswork than previously accounted for in much of the relevant literature-even when structure constraints action-and journalists are purposive actors. Furthermore, Usher shows how the agency can shape structure when new technology is introduced; the question is whether collective agency-the performance of autonomy-could also shape structure after ownership change.
We believe that the question of whether and how it is possible to perform journalistic autonomy even in illiberalizing contexts-and the hope that there is such a thing as autonomy in these contexts-should be further critically thought through and empirically studied before today's full democracies need the answer.