Staging Expertise in Times of COVID-19: An Analysis of the Science-Policy-Society Interface in the Dutch “Intelligent Lockdown”

The corona crisis of 2020 took many by surprise. Quite suddenly, politicians had to make drastic decisions to guarantee public health, affecting basic civil liberties. In justifying their decisions, politicians internationally reverted back to a direct staging of experts to legitimize their proposals for what internationally became known as the “lockdown”. In this article we analyze the performance of the Dutch government that, early on, labeled its approach to COVID-19 as an “intelligent lockdown”. Our analysis examines the dramaturgy of expertise during this period. We selected two interrelated “stages”: the official press conferences, fully controlled by the government, and the responses on Twitter, as focal channel for critique from the general public, but also from opposition parties and (alleged) experts. 26 press conferences of the Dutch Prime Minister were analyzed and a search for the most popular posts on Twitter referring to the press conference(s) was carried out covering the period between March 6th and May 29th, 2020. The results show that the technocratic framing of expertise remained stable during the sampling period, regarding the undisputed status of expertise as the clear-cut basis for decision-making in uncertain times. Framing on Twitter challenged the omnipotence of the experts advising the government in various ways, namely, by referring to dissenting opinions of other experts, by questioning the underlying motives of experts’ advice or by pointing out that the policies were clearly contrary to everyday experience. We argue that it is not so much the facts themselves that are at stake here but hidden moralities, which include the government’s alleged complacency while asking citizens to blindly trust, its unpredictable behavior in the light of the promised straight line between scientific evidence and policy making, and its motivated behavior while claiming that the facts speak for themselves.


INTRODUCTION
Effective management of the COVID-19 crisis required strong and lasting collaboration between policy makers, scientific experts and citizens. From the outset the Dutch approach was aimed at making citizens collaborate. This collaboration depended upon citizens' trust in expert advice and their judgment of the situation. However, this trust in expert advice could by no means be assumed.
On the contrary, persistent debates rooted in what experts consider solid science, such as climate change or vaccinations, have been showing for some time now that the authority of experts is not a matter of simple, let alone, unconditional acceptance (Hajer, 2009;Oreskes and Conway, 2010). Experts increasingly find themselves in a situation where they must publicly earn their own credibility (Hilgartner, 2000;Jasanoff and Simmett, 2017;Turnhout et al., 2019).
Over recent years, the contestation of expertise has shifted from disagreement with experts to a situation in which experts and expert organisations are regularly confronted with cynicism, where people's distrust extends to the intentions and motives of those organisations. Responses range from a persistent "Why can't we just ask a question?" to fierce, personal attacks on social media and vehement protests on the streets. It begs the question of what drives and gives substance to these disputes, a question that has become all the more relevant in the context of the COVID-19 pandemic, in which democratic decision-making has become unprecedentedly dependent on the advice of scientific experts (Akerman et al., 2020).
In the Netherlands, scientific expertise was already heavily and publicly contested before the COVID-19 pandemic. For example, in 2019, a judge ruled that the nitrogen pollution caused by emissions from agriculture was not in line with the law and should be drastically reduced. Farmers' organisations contested the objectivity of the National Institute for Public Health and the Environment (RIVM) in modeling the influence of nitrogen emissions on nature. Some farmers suggested a conspiracy against the agricultural sector. As it happened, it was the very same institute, the RIVM, which in early March 2020 called upon the government to take drastic measures to prevent a collapse of the healthcare system. The combined effects of international holidaying and Carnival presented such an increase in COVID-19 infection rates that the meticulous "track and trace" system to control diseases and epidemics no longer was a viable option. Consequentially, the Prime Minister directly addressed the nation on March 16th, the first time after 1973, to announce the "intelligent lockdown". Compared to the extensive restrictions of other European countries in the same period, there was no curfew during the Dutch "intelligent lockdown" and even shops could keep their doors open under certain conditions. Simultaneously, there was a heavy emphasis on the responsibility of individuals to adhere to basic guidelines such as social distancing or self-isolation in case of COVID-like symptoms.
For a society whose functioning and well-being depend almost routinely on the effective and legitimate use of science, and to which this is even more true in times of crisis, new avenues need to be explored to re-value the position of scientific expertise in policymaking. Truth claims in the Western public sphere have been treated as sufficient and robust only when associated issues of public value and purpose were addressed in tandem, as Jasanoff and Simmett (2017) point out. It is not so much scientific facts in themselves that may offend people, but the (denial of) values associated with these facts. This means that looking at disputed expertise includes studying the role of morality. In this article, we therefore examine the institutional performance of scientific expertise in conjunction to value-creation, and the public appreciation thereof.
The Dutch COVID case allows us to study how a polity with very well established and generally recognized knowledge institutions (Wardekker et al., 2008;Huitema and Turnhout 2009;Van Asselt et al., 2014) responds to the need of a new enactment of scientific authority. The established but publicly little-known connections of these institutes with policy makers proved to be an advantage in establishing a repertoire of sciencepolicy interaction, in the sense that knowledge specialists and policy makers could easily find each other. However, it also turned out to be a handicap, as the close relationships made them vulnerable to allegations of interests. An important and early part of this science-policy interaction in the context of COVID-19 will be analyzed in this article.

FRAMING EXPERTISE
For our analysis of the debates, we rely on a framing approach. In the often-cited definition of Entman (1993, p. 52), framing entails "select [ing] some aspects of a perceived reality and mak [ing] them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described". In empirical research, the actual conceptualization of framing differs considerably across studies. In our efforts to understand how the COVID-19 debate evolved between different actors on different platforms, we rely on an interactional understanding of framing, wherein framing is regarded as a dynamic process of co-creation between participants and where meaning is therefore negotiated (Edwards, 1997;Dewulf and Bouwen, 2012). We describe this process using two key principles in the framing literature. First, we consider diagnostic and prognostic elements to be crucial aspects (Snow and Benford, 1988; see also; Kroon et al., 2018). Diagnostic framing relates to problem definition and answers the question of what the problem is and who is responsible for it. Prognostic framing provides potential solutions and also indicates who can alleviate the problem. Second, the actual manifestation of frames in texts is through so-called 'packages' (Gamson and Modigliani, 1989, see for a recent application; Wichgers et al., 2020). At the core of such a package is the frame, that provides meaning to issues or events, and comes with a range of manifest framing "devices", such as metaphors, word choices, descriptions and arguments, that are related to the frame, and its diagnostic and prognostic components. In some instances, these components are explicit and manifest, in other instances, they have to be inferred from the larger context (Pan and Kosicki, 1993). The challenge is thus to identify those framing packages and analyze their contribution to the ongoing process of defining issues, ascribing responsibilities, claiming rights and building identities (cf. Dewulf et al., 2009).
Even though we regard the framing process as essentially coconstructed, ample research has shown that the larger context in which public debates take place seriously constrain the range of possibilities available to those involved (Vliegenthart and Van Zoonen, 2011). Most notably, it is those with political power that set the scene and indirectly the boundaries of the range of opportunities others have to express their views without completely "falling on deaf ears" (Snow and Corrigall-Brown, 2005). In the specific context of our study, framing boundaries were largely set during press conferences. They formed the stage where the official government measures to combat the pandemic were announced to the Dutch citizenry. These events were broadcasted live on the major public and private television channels, on multiple online news sites and on YouTube, and were perpetually referenced in newspapers and television news. The press conferences alone consistently attracted many millions of viewers, with the four most viewed press conferences in the sampling period reaching an audience of more than 7.5 million citizens (SKO, 2021, January 7). For counter voices to receive more than just marginal attention, they will usually have to find ways to align their challenging frames with the framings of those who wield political power (Snow et al., 1986). We consequently expect that government actors 1) are leading the debate, thus providing a central interpretation of the problem definition, diagnosis and prognosis; that this 2) constrains the opportunities of other actors to bring forward a radically different framing; and that 3) actors will present their own (challenging) frames that extend or transform the boundaries of the dominant, governmental framing (Snow et al., 1986).

HIDDEN MORALITIES
On the surface, disputes regarding COVID-19 policies revolved around questions that would typically fall into the jurisdiction of science, like "Can a relevant level of herd immunity be achieved to mitigate the spread of COVID19?" or "To what extent do face masks prevent the spread of the virus in public spaces?" However, experience shows that such conflicts cannot be resolved simply by providing more facts (Shapin, 2007). When facts are controversial and there is a lot at stake, correcting misinformation will not so much allay existing concerns but reinforce them (cf. Nyhan and Reifler 2015). Moreover, in these kinds of controversial situations, citizens refer to facts and figures just as frantically as experts do (Te Molder, 2012;Te Molder, 2014;Versteeg et al., 2018). Both phenomena suggest that the cause of disagreement does not lie in science as such; it transcends science's (fuzzy) boundaries. When knowledge is at stake, so are the values associated with it (Jasanoff, 2004;Jasanoff and Simmett, 2017;Durnová, 2019).
In this study, we therefore pay attention to how both government and civil society actors attend to these values when making sense of corona policies and the role of scientific expertise therein. We use the term "hidden moralities" to emphasize that we are analyzing practices that normally remain under the radar. Swierstra et al. (2009) point out that morality exists in daily life of practical routines that are difficult to articulate whereas ethics, on the other hand, is marked by explicitness, reflection and controversy. Moralities are not easy to distinguish because they present themselves in close harmony with the facts people use. They comprise conceptions of what constitutes a good relationship, for example between governments and their citizens, or what entails a good life more generally (ibid.). Vaccine-hesitant parents tend to use science to make their point, rather than directly referring to the idea that a good parent should not blindly trust (Te Molder, 2012;Reich, 2016). Climate sceptics suggest that abatement measures are ineffective rather than publicly unpacking notions of good governance.
We agree with Jasanoff and Simmett (2017, p. 764) that "we must remember to ask, and insist on good answers to, questions about what underpins both sets of authority claims [i.e., scientific and political, our addition and emphasis] in the first place" and pay attention to issues such as "Who made the claim?", and "In answer to whose questions or purposes?". In many cases, however, the answers to these questions are difficult to provide because facts and values are so intimately linked. The "mixing" of facts and values is, for a large part, a tacit and practical achievement, that can only be understood when viewed in the context of everyday life. Before knowledge-political repertoires can be disentangled and considered, at least to some extent, we must first surface hidden moralities as they are attended to by participants themselves, in real-life interaction. To this end, we propose a combination of a framing and a discursive analysis.

DATA AND METHODS
To investigate when and how the Dutch government framed experts and expertise when formulating problems and proposing solutions, publicly available transcripts of the Prime Minister's press conferences have been analyzed. These events summarized the newest developments discussed during the prior crisis-or Ministers gathering and provided journalists the opportunity to ask follow-up questions. Whereas the Prime Minister was part of all press conferences, some featured additional key figures of the crisis, like the Minister of Public Health or the chair of the Outbreak Management Team. During the sampling period, starting from the first confirmed corona patient in the Netherlands on February 27th until May 31st, 26 press conferences were organized, including the ones that take place each Friday after the meeting of the government council. We collected them from the official site of the Dutch government. 1 The respective transcripts were split between two of the authors and qualitatively analyzed for all references to individual, institutional or categorical expertise. Here, we were guided by questions such as what was considered to be the problem, what the underlying causes for the problem were and what was presented as the potential solution. For this, we relied on an inductive approach. Core questions that were answered are how the larger pandemic was presented as a problem (diagnosis), as well as whether and in what ways science and scientists could offer relief for the consequences of the pandemic (prognosis). We looked for manifest metaphors, phrases and arguments as indicators for underlying diagnostic and prognostic framing elements. Only those sections that dealt with COVID-19 were analyzed. Cross analyses of four randomly chosen transcripts indicated robust agreement of expert references between analysts. They selected the same relevant excerpts and also provided highly similar interpretations of the framing of expertise. This approach resulted in a collection of around 100 paragraphs, each of which contained at least one reference to expertise, science or individual experts. These paragraphs provided the basis for describing dominant framings of expertise.
Second, a discursive analysis was performed to scrutinize the uptake of these governmental messages on Twitter. We were interested in what aspects recipients made relevant from earlier messages and how they thereby treated the utterance, i.e., as what kind of action, for example, an accusation, a compliment or an offer (Potter, 2012;Te Molder, 2012). This step allowed us to analyze how participants themselves attended to (hidden) moralities, similar to how, for example, participants in anti-vaccination forums suggest that "a good parent" should refrain from blind faith in government and science, by accusing parents of naivety when they fail to check the facts before accepting them (Versteeg, 2018).
For this part of the analysis, we used Coosto, which provides an online archive of Dutch language Tweets. As one of the biggest, freely accessible social media platforms, Twitter offers a forum to all citizens who aim to join the public discussion about the government's corona policies. During times of crisis in general and outbreaks of infectious disease in particular (for an overview, see Tang et al., 2018), scholars have increasingly turned their attention to Twitter to understand how people make sense of the situation (e.g., Vos and Buckner, 2016;Stieglitz et al., 2018). Twitter might not be a perfect representation of public opinion or views (McGregor et al., 2017), in that it analyses public voices in a restricted and somewhat magnified form. On the other hand, it offers insight into what normally may remain marginalized voices. All in all, it offered us an accessible and suitable forum to analyze initial responses to the content of press conferences. The use of hashtags related to the press conferences and announced measures provided us with ample opportunity to select relevant Tweets.
We developed a search string to identify these relevant Tweets-i.e., those that referred to 1) COVID-19, 2) the press conferences and 3) expertise (either in the form of reference to the main advisory bodies of the government, or to science or knowledge generally). 2 This search revealed 10 days in the sampling period wherein the number of Tweets relating to the search query was peaking, ranging between 31 and 266 Tweets per day. From each of the days, the twenty most commented posts (including their respective comments) were sampled. Selecting the most commented Tweets, rather than the software's option to sort by "influence", effectively excluded Tweets from the dataset that gained lots of views by virtue of their large followership only. This approach ensured not disadvantaging Twitter users with a small to medium followership, compared to large institutions or prominent individuals on the platform. Retweets were excluded from the sample at the last step of selection.

RESULTS
We present our results in pairs of three. For each pair, we first describe one aspect of the expertise framing in press conferences. Subsequently, we show how people on Twitter respond to that framing and demonstrate how this uptake points toward a moral concern that remains (largely) unaddressed but is of crucial relevance to the development of the COVID-19 debate. An overview of the findings can be found in Table 1.

Press Conferences: Trust the Virologists
References to expertise were most frequently used to emphasize that decision making is based on expert advice. From the very beginning, it was established that the Dutch Institute of Public Health and the Environment (RIVM) is the government's primary scientific advisory body. From March 12th onwards, an Outbreak Management Team (OMT) was assembled to provide scientific input for governmental decision making and coexisted as initialism for an expert body next to the RIVM. However, it was publicly known that most OMT members are either directly employed by the RIVM or are strongly affiliated with it and therefore it made little difference whether government officials referred to one or the other. On various occasions during the press conferences, the Prime Minister suggested that these are the experts that will provide the intelligence and know-how for outbreak management policies (March 6th: . . .that this really requires specific expertise from the GGDs 3 , from the RIVM. So, they are very important. They are also, world-wide, among some of the best experts in the world in this field. Fortunately, we have them in the Netherlands. And it is very important to build on their advice). When a high-profile doctor publicly announced that hospital capacity allowed for a loosening of outbreak measures on April 24th, the Prime Minister was confronted with the question of why he did not follow up on that insight. In response, he pointed out that the advice of around 40 OMT members weighed heavier than that of a single expert (April 24th: We rely on the advice, of a group of forty experts really (...) we cannot make our decisions based on one physician).
Next to this institutional demarcation of expertise, extensive boundary work was performed on different categories of experts: During the sampling period, virologists were repeatedly put forward as the ones who possess relevant knowledge (March 9th: I'm not a virologist, neither are you, I believe. This is really specialist work; And to be honest, the Netherlands has made a very big mistake of making me Prime Minister, because I am not a virologist. If only I were one, I could have thought of it all myself). Others, like legal experts and historians, should not be making decisions because the unique circumstances require a very particular expertise (March 12th: But the advice of experts is important because this is a highly specialized problem, and you 2 The following search string was used (covid* OR corona OR virus) AND (ggd OR omt OR "outbreak management team" OR expert* OR deskundig* OR wetenschap* OR adviseur* OR advies OR rivm) AND persconferentie. have to be careful that a lawyer or a historian will not ultimately make all decisions). Taken together, this governmental framing followed the reasoning of three connected steps: First, it is up to highly specialized individuals to define problems and propose solutions related to the COVID-19 pandemic. Second, the epitome of highly specialized individuals is to be found in the OMT and nowhere else. Third, these experts should receive the trust of citizens (March 13th: Then I think the Netherlands has reason to trust in the experts who advise us).
They are thus presented not only as those who have the right to propose measures, but those measures also logically flow from the experts' unique expertise and therefore deserve "blind" faith.
Twitter: Why Would we Trust Blindly? Portraying the Government as a Complacent Actor This governmental sentiment gained a lot of attention and created a considerable amount of pushback on Twitter. Typical reactions to it are depicted in Tweets 1, 2 and 3 below. Their first common element is that they all challenge the scientific foundation on which the government takes action, thereby achieving quite the opposite of the governmental plea "leave it to the experts". It demonstrates that citizens have a stake in the scientific advice that translates into policies which in turn directly affect their lives. Secondly, by keeping a close eye on the experts and reserving the right to criticize, especially as nonexperts, Tweeps 4 make it clear that they will not trust experts blindly. People who do trust without questioning, are being compared to sheep that simply follow wherever the herd may go (Tweet 3). Thirdly, the Tweets suggest a strong contrast between advice from the Dutch OMT and views of other expert entities. These other entities include the scientific community and publications (Tweets one and 2), the alleged superior approaches of other countries (and their expert advisors) (Tweet 4) and individual scientists (Tweet 5). The plentiful references to expertise outside of the OMT treats the government's exclusive focus on OMT advice as an unwillingness to broaden their horizon and learn from others. In more extreme cases, the disregard for other experts' opinions is interpreted as self-righteous behavior (Tweet 5), which is suggested to further undermine the government's trustworthiness. This type of counter framing does not call into question the authority of scientific expertise as such and therefore does not provide a fundamentally different interpretation in terms of diagnosis and prognosis. However, it does dispute the inevitability of the specific institutions and experts the Dutch government relies on. 4) 6,022 followers 5.4 influence 15 comments #Rutte and the RIVM are doing their utmost to guide the Netherlands through the #Corona crisis. But dear experts, do you monitor the Scandinavian approach? Sweden (10 million inhabitants), for example, has 3 times fewer infections and 7 times fewer deaths! #pressconference #coronanederland. 5) 11,239 followers 130.6 influence 801 comments If Minister de Jonge hears this, he jumps out of his skin. He was already angry about it during the press conference. A professor of medical microbiology with his own initiative, indeed, things shouldn't get any crazier in this country. #coronavirusNetherlands. There is no logic to these policies. Turning government into a whimsical actor When stated that facts speak for themselves, we suspect something else: identifying the government as an interestdriven actor Topics Other countries, other experts Schools; face masks Schools; face masks; herd immunity 4 a person who uses the Twitter online message service to send and receive tweets. 5 "followers" indicate how many other people follow the Tweet's author; "influence" is a score that indicates how likely it is that Tweets by that particular author will be seen by other people; "comments" display the number of responses that particular Tweet has generated. When the first corona patients were confirmed in the Netherlands, the government took the stance that the situation was under control. This claim was supported by reference to experts' levelheadedness in the face of a potential public health crisis (March 6th: And one is always impressed by the calm and expertise and the focus and the enormous dedication in which all this happens). The growing number of infections made clear that the tracking and tracing of individual cases would soon become impossible. In the light of this development, a lack of knowledge was first problematized on March 12th (March 12th: The fact is that in a crisis like this you have to make 100% of the decisions with 50% of the knowledge). From this point onwards, uncertainty remained a part of the expert framing, but it evolved over time.
During the first half of the sampling period, uncertainty was portrayed as a general nuisance of life rather than something that obfuscates expert advice. For instance, the WHO's declaration of a coronavirus pandemic has been instrumentalized to point out that foreign introduction of the virus can have unpredictable consequences for an open country such as the Netherlands (March 12th: we have cases all over the world and that of course has consequences for an open country like the Netherlands). Consequently, this particular framing of uncertainty led to an almost unconditional reliance on experts, as they were portrayed as the only ones capable of addressing and dealing with uncertainty (March 12th: You can never be 100 per cent sure whether you are doing exactly the right things. But we do things based on the latest scientific insights).
At this stage, uncertainty was portrayed as external to expert advice and therefore unaffected by it. This sentiment was essential in creating one of the biggest controversies during the early pandemic, the question of whether or not to close (elementary) schools. On March 13th, the Prime Minister announced that schools were meant to remain open and that this decision was motivated by medical considerations (March 13th: Medical experts advised us on that basis not to close the schools). This announcement caused protests among school representatives and society at large and ultimately changed the government's stance on this position. They made clear that in this case, the will of the people had to prevail, even if that meant that valid advice could not be followed (March 20th: And I think the situation regarding education is that society corrected us and said, we don't agree with you. Scientifically it may be correct, but it feels different. We close anyway). Note how, while it was the scientific reasoning that people questioned, the Prime Minister made a dedicated distinction between the scientific validity of keeping schools open and the emotional will of the people to want them closed.
The framing of uncertainty changed in light of the emerging disagreement between experts concerning the added value of using face masks in public spaces. On April 17th and May 1st, the Prime Minister referred to the uncertainty of expert advice, when questioned about increasing the use of face masks as a preventive measure (May 1st: the problem is, the face masks, that has already been explained of course, is that it, it's not black or white; we just talked about the face masks, then you see that they take a little more time, which is fine, because it is not all black or white or plus or minus, there are trade-offs). In defending the delayed decision on the matter, the Prime Minister explained that experts are still in doubt about the added value of face masks, conceding uncertainty regarding the advice. It was further argued that face masks are no alternative to what experts consider the safest behavior: keeping one and a half meters distance to others (April 17th: It is not an alternative to the meter and a half. The meter and a half, all the experts tell us, is really the safest. That is really the safest). While the government's stance on face masks remained unchanged by the second week of May, the certainty about the uncertainty surrounding this issue was in itself described as a basis for inactivity (May 8th: no, of course, in life there are rarely 100 per cent hard facts, but you have to base yourself on what you know, and even if you don't know something, that's a fact too).
During our sampling period, the scientific status of certainty was constantly negotiated as to fit the ends for which this (un)certainty was deployed, consciously or not. Scientific evidence was presented throughout as the clear-cut and untouchable basis for decision-making, even if the evidence itself was considered not certain or found inconclusive by citizens. The many acknowledgments of uncertainty were used to suggest that government decisionmaking was carried out with due care, by transcending black and white thinking, or by presenting it as a bad excuse for not acting. Paradoxically, uncertainty was also used for justifying inaction in the case of face masks, and to encourage more action in the case of social distancing, as this was treated as the most "certain" of all other possible measures.
Twitter: There Is No Logic to These Policies.

Turning the Government Into a Whimsical Actor
The notion of scientific evidence as the indisputable and clearly defined basis for policy making, is challenged in a variety of Tweets during the whole sampling period. On Twitter, expert advice and resulting policies were continually re-evaluated in the light of prior decisions and the ambiguous circumstances of everyday life, functioning as a reminder that expert advice can only be meaningful if it accounts for a complex social reality, and if it is seen as consistently applied within that reality. If the measures stemmed directly from the evidence available, why then apply (strict) measures in one area and not the other? For instance, when the government declared the situation highly unpredictable due to foreign introductions of the virus, the question arose as to why flights from high-risk countries had not yet been canceled when strict measures were in place in similar areas (Tweet 6). 6) 16,820 followers 121.6 influence 113 comments Why are flights from seriously infected countries such as Iran and Italy still allowed to land in NL? Now that the RIVM Frontiers in Communication | www.frontiersin.org July 2021 | Volume 6 | Article 668862 6 mentions the introduction of viruses from other countries as one of the reasons for the new measures? #pressconference #coronavirusNL.

7) 965 followers 17.3 influence 42 comments
Soooo ... #RIVM says that the chance that children will transmit the virus to adults is almost nil, but the question whether grandchildren can now visit grandpa and grandma was answered in the negative because of the risk of infection!!! Pretty strange #pressconference. 8) 15,486 followers 134.0 influence 134 comments #pressconference So #Rutte and Co. follow (hide behind) the advice of Jaap van Dissel of #RIVM and say that it is safe for teachers and parents to open the schools, but the same children, who do not transmit the virus, are still absolutely not allowed to go to grandpa and grandma ...
In a similar case, the Prime Minister announced that experts advised that closing the schools would not be necessary because children would not play a major role in the transmission of the coronavirus. This decision caused disbelief among parents because grandparents were simultaneously discouraged from seeing their grandchildren. Tweeps pointed out how these measures appeared to be in clear conflict with each other, hinting there was probably something wrong with the evidence and reasoning underlying the measures (Tweets 7 and 8). This sparked a social debate that quickly led to the decision to close schools after all, not because the science was wrong but because of the contrast with how people "felt" about this measure (see governmental framing above). Similar reactions can be found regarding the government's tenacity that one and a half meters distance in public spaces is enough to mitigate the spread of the disease. People countered that in many situations, physical distancing rules are not or cannot be adhered to (Tweets 9 and 10). These examples show that Tweeps reject the suggestion of a logical line between the scientific evidence on the one hand, and COVID19-policies on the other. If the line is so direct as claimed, why does the same evidence yield strict measures in one case, which are then declared inapplicable in a different but comparable case? They also signal a sharp contrast between the claimed unambiguity of the measure and the lack of government understanding of the messy day-to-day practice. By calling out these inconsistencies, people hold policymakers accountable for their seemingly incompatible decisions. Substantiation for this claim can be found in Tweets 8 and 11, in which politicians are being accused of hiding behind experts and their advice. These accusations orient to the norm that it is politicians, not experts, who ultimately make the decisions and who should be held accountable. Accountability in this context refers to the ability of policymakers to engage in deliberation with citizens, especially in situations that are marked by a high level of (also social) uncertainty and multiple paths forward. Since uncertainty was never acknowledged in this broader way, Tweeps treated the governmental references to uncertainty as lip service, allowing the government to hide behind expert advice. 11) 2,353 followers 31.6 influence 17 comments Hugo de Jonge emphasizes that the cabinet is hiding behind the advice of experts. #corona #pressconference #rivm.

Press Conferences: While Embedded in Politics, Fact Still Speak for Themselves
On April 7th, a reporter asked whether advice from the OMT is sacred, upon which the Prime Minister responded with a sole, decisive "yes". The question did not come out of the blue but was a result of the government's overt dependence on expert advice, leaving commentators wonder about who truly is in charge during the crisis. The primacy of science was strengthened on various other occasions during press conferences (March 12th: ultimately it is very crucial that expert advice is at the basis of these measures). In a particularly remarkable example, the Prime Minister explicated that lengthy discussion about policy alternatives are rendered unnecessary when one can also rely on expert advice (March 9th: It is very wise to follow that advice. And then I understand that some people say: yes, but shouldn't you go further or this or that. And it applies every time: no, we build upon that advice). That this was more than just political talk became clear when primary schools were made the first institutions to reopen after the partial lockdown. The rationale for this decision was not that the will of the people had changed, but that new evidence reaffirmed that children do not play a major role in the spread of coronavirus (April 21st: We now know a lot more of course. The RIVM is now also conducting research in the Netherlands itself. That was also one of the commitments we made after that weekend when the schools closed). Thus, where previously "the will of the people" had been prioritized above expert advice, a month's worth of research turned the tables again.
The government noticed early on that this mode of decision-making, wherein a small group of experts appear to dictate the course of action, is a problematic arrangement and emphasized that all advice was evaluated in the light of political considerations (March 12th: where of course we build upon the advice of experts, that we do not adopt Frontiers in Communication | www.frontiersin.org July 2021 | Volume 6 | Article 668862 7 blindly, as is sometimes assumed or that we don't have our own opinion). However, even the Prime Minister had to concede that OMT advice is usually complied with (April 7th: Of course, we always have a political administrative discussion about this. But up to now we have actually always adopted the OMT advice afterward). However, this political context was only identified in general terms, without demonstrating and publicly analyzing the boundary work between politics and evidence in practice, and in actual detail.
Accordingly, virtually no deliberation about how specific scientific insights may rank one policy alternative over the other, was staged in the public sphere. Instead, it was suggested that facts speak for themselves and, as experts deliver these facts, it would be foolish not to follow them.
Twitter: When Stated That Facts Speak for Themselves, we Suspect Something Else: Identifying the Government as an Interest-Driven Actor The strong focus on expert advice combined with the dedication to quickly reopen schools gave Tweeps reason to believe that the government was pursuing a hidden agenda. Specifically, parents were concerned that what was truly behind this decision, was a covert attempt to strive for herd immunity (Tweets 12 and 13), a concept that had caused lots of commotion in mid-March already.
12) 13,869 followers 336.1 influence 159 comments #pressconference About that face masks affair: could it be that Rutte and @rivm still aim for #herdimmunity despite all the warnings? I get suspicious of that twisting and lack of action. 13) 408 followers 25.7 influence 3,379 comments @[Geert Wilders, opposition party leader] #herdimmunity still the plan. Now via children at school and day-care, who infect the parents, who "hopefully" also have mild complaints, says member OMT at #jinek at odds with what Van Dissel said in press conference! #closeschools.
[Comment to post above] WTF! So #herd immunity after all, even via children in school or day-care, "hopefully the parents will have mild complaints" she literally says. Our kids are guinea pigs! @rivm @[Hugo de Jonge, Minister of health] @[Mark Rutte, Prime Minister] explain this! Next to schools, the second big controversy revolved around the use of face masks in public spaces. While the reason put forward by the RIVM and the government was that the masks' added value was very questionable, Tweeps suspected that in reality, there was a shortage of facemasks and that they had to be spared for healthcare workers (Tweets 14 and 15). The fact that policymakers could simply claim ignorance and refer to their status as non-experts whenever faced with uncomfortable questions (Tweet 15), was seen as another sign of them hiding behind experts (Tweet 11). The novelty of the current framing when compared to the second framing, is that policymakers and experts are now portrayed as conspiring together in the pursuit of dubious goals. Thus, the heavy reliance on expert advice is treated as a way to bypass public discussion about policy alternatives (Tweet 16). To many Tweeps, this behavior only made sense if the government had an interest in undermining dissent in order to follow their hidden agenda. In their response, they show that they are neither naive nor stupid, by uncovering concealed motives the government or OMT may have. 14) 3,898 followers 202.9 influence 487 comments In the press conference, De Jonge finally admits it after followup questions about face masks for Dutch citizens. "The experts also advise with scarcity in mind of course". In plain Dutch "we do not recommend masks, otherwise we admit that we cannot arrange that either". 15) 1,119 followers 93.1 influence 366 comments Question to Hugo de Jonge during the #pressconference: "Belgium and Germany strongly advise to wear face masks, why not the Netherlands?" De Jonge: "I don't know, I follow the experts, and healthcare professionals really need them." We now know that they are simply not there. 16) 27,059 followers 224.7 influence 1,589 comments Cabinet refuses to disclose corona crisis documents after WOB 6 requests, until they become irrelevant https://t.co/ lfvUJbOSaX Thus a Minister with a handful of experts will run the country without sharing the information .. Scandalous! #pressconference #coronameasures https://t.co/bGIQv8v8vB.
[Comment to post above] Corona policy has major consequences for democracy, fundamental rights, the economy and the social life of people. There was never any significant discussion about alternatives .. #Coronavirusnl Sign for a parliamentary inquiry on corona policy! https://t.co/bgHgZ670ni. Again, the overall framing on Twitter did not deny the importance of expertise in decision making as such but criticized the government for withholding information from the public, thereby failing to be honest and fair in their decision-making processes.

DISCUSSION
The COVID crisis took the Dutch government by surprise. By implication, the initial response had the character of an 6 Refers to the Open Government Act (Wet Openbaarheid van Bestuur: WOB).
Frontiers in Communication | www.frontiersin.org July 2021 | Volume 6 | Article 668862 8 "impromptus" performance. Yet any choice in terms of staging yourself as government is a particular "enactment" of authority and thus comes with consequences. The performance as found in the first phase of the pandemic fits the classical "science-forpolicy" formula (cf. Van Dijck and Alinejad, 2020). It goes back to the formulation of Aaron Wildavsky (1979) who introduced it as the "speaking truth to power" format. According to this positivist approach, the legitimacy of political choice is backed up by expert advice, in which experts come to policy suggestions based on the best available scientific advice. Typically, the expert here draws on classical sources of knowledge, based on Humean general laws. This type of knowledge, found in the scientific literature, has a particular standing as it is seen as codified by the practices of science such as peer-review. It brings out universal laws that are argued to be culturally neutral and objective (cf. Hajer et al., 2009). Furthermore, essential in technocratic governance such as this is that facts, data and information are expected to resolve possible conflicts (Fischer, 1990: 28), and that they are visibly and clearly distinguishable from values or politics, also with a view to their public accountability.
However, the resolving of COVID-related conflicts and building of strong truth regimes (Jasanoff and Simmett, 2017) cannot be achieved by addressing facts only. Or as political scientist (Fischer, 2019, p. 135) puts it: "fact-checking or the discovery of more and better facts will not put this controversy to rest." What is needed is the addressing of moralities and the institutional arrangements that make those facts relevant and possible. While many scholars would agree, few approaches actually reveal these moralities, and virtually no approaches can surface them unless they are explicitly available. We hope to have shown that our analysis was able to do so.
We demonstrated that in their responses to government policies, Dutch Tweeps not only -or not so much-disputed the scientific evidence brought up, but especially their moral relationships with government, politics and experts, including the government's alleged complacency while asking citizens to blindly trust, their unpredictable behavior in the light of the promised straight line between scientific evidence and policy making, and the invested behavior they showed, while claiming that "facts speak for themselves".
Interestingly, "uncertain" knowledge was drawn upon by the Dutch government in their press conferences, both to account for action (we cannot but act, even in the light of uncertain knowledge), and to account for inaction (we do not recommend face masks in the face of inconclusive evidence). Perhaps the most striking result was that while Tweeps accused the government of inconsistent policies, by indirectly contrasting these policies with the straight line between evidence and policy making that the government had promised, none of the responses called into question the underlying technocratic model itself. That is, there was no disputing of the value and position of scientific expertise itself for the making of these policies, other than referrals to alternative scientific positions than the cherished virologists and epidemiologists represented.
Note that this is the case and seems characteristic for at least the early period of the COVID crisis that we studied. Our analysis is confined to the Dutch situation in the Spring of 2020 and cannot be automatically extrapolated to other countries or periods. As Jasanoff (2005) demonstrates, nation-states may vary in their "civic epistemologies", that is, their own cultural ways of public knowledge-making and resolving disputes around data and evidence. Nor can it be ruled out that the Dutch government deployed other knowledge-political repertoires in the later stages of the crisis, which subsequently provoked different reactions. However, first impressions of changing relationships between science and politics can easily be deceiving. As we have shown, in this first period of the pandemic, the underlying technocratic model remained unchanged, even when there was talk of opening up the debate to non-experts and citizens (in what Van Dijck and Alinejad (2020) call the "smart exit phase").
An important element of the dramaturgy of the "intelligent lockdown" was the coupling of the political leader (Rutte) to the prime expert (Van Dissel). While this makes sense for a media savvy politician like Prime Minister Rutte, it creates a tension with the traditional "science-for-policy" format in which scientific expertise is often seen as an institutional resource. Yet the press conference, and the personal communication of the Prime Minister in the first weeks created an explicit personalization of expertise. In his first speech he named "Jaap van Dissel" as the leading expert. Staging expertise in a person inadvertently loosened up Jaap van Dissel as a person from his position as one of the directors of the National Institute for Public Health and the Environment (RIVM). Following the mediaformat, it was the person of Van Dissel who was taken on in the social media critique, not the institutional powerhouse of the RIVM he represented. In the end this conscious and insistent "staging" of the Prime Minister and the director of the RIVM as a "pair", created a personalization of the institutional expertise of RIVM. In that case the authority comes to be dependent on a mix of the institutional reputation and the personal style of its spokesperson. As Dick Pels observed, a long time ago: "Increasingly, positions of political power are dependent upon public trust, belief and confidence (and upon those who are able to manipulate these volatile variables), and hence upon a recognizable political style that weaves together matter and manner, principle and presentation, in an attractively coherent and credible political performance" (Pels, 2003, p. 57).
So, while putting the expert Van Dissel on stage may have helped the politician Rutte in supporting his choice with a scientific basis following the "science for policy" format, the choice to communicate via press conferences may have created a feeling of an expert that is aloof and distant. This is the more plausible as Van Dissel himself constantly used the "science for policy" format with insistence on finding his scientific authority in referring to the peer-reviewed literature, as for instance in the cases of school openings or closures or the wearing of face masks or not. If Van Dissel would have employed a different, dialogical model, he would have shown himself to be open to the knowledge of teachers or people operating in contexts in which face masks may seemingly have had an effect.
The weekly press conferences functioned as a clear orientation point for the critique on social media, most notably Twitter. In that context one cannot assume a shared knowledge of the way in Frontiers in Communication | www.frontiersin.org July 2021 | Volume 6 | Article 668862 9 which the science-policy interface is organized. Hence, that what is uttered in the context of the press conference becomes the lead for interpreting how policy making is organized. Here the complex layering of organisations -each with its own acronym (RIVM, OMT, Van Dissel as chair of the OMT-as a council of specialists spread over various organisations, Van Dissel as director of RIVM), reinforces the suggestion of a hidden technocracy.
There was always a complexity in the performance of Jaap van Dissel as the sole spokesperson of the science. On the one hand the Prime Minister could "simply" refer to him to speak for the science, on the other hand Van Dissel had to share what he and his council (the OMT) had agreed upon. It is interesting to note that the OMT had decided that their internal deliberations and decision-making moments were not transcribed and that the meeting were not supported by minutes that were publicly accessible. While we can only guess, this may have been because of the alleged undermining of the authority of the science should people be able to "listen in" on the arguments among scientist experts. Yet, there is reason to doubt whether that effect would indeed have taken place. The critique we now found on Twitter, suggesting that the Dutch government was operating in a sphere of secrecy and elitism, could have been effectively combated in this way.
Unwittingly, the Dutch staging of science in terms of one key person and keeping the broader deliberation in the OMT from the public eye, opened the gate to suggest that secrecy was necessary, because something had to be kept hidden. The fact that it remained undisclosed on the basis of what science measures such as halting public transport or restrictions on outdoor gatherings were taken, fed the sense that it was not science but politics that was dominant.
The dramaturgy of the press conference was that of the classic "speaking truth to power", with an expert staged as advisor to politics. In this classic model (Wildavsky, 1979), it is the politician that is to do the weighing of pros and cons based on hearing the evidence. But what is remarkable in the Dutch case is that the weighing of evidence, of pros and cons, was suggested to be outside the event of the press conference. The OMT was a primarily medical council and to the extent that social scientists were involved it was to come to an assessment of the effects of policy measures on fighting COVID. Yet the deeper choices, whether the measures were proportional vis-à-vis other (side)effects, e.g., on personal freedom or economic and social well-being, were not part of the publicly presented weighing of the evidence. That these are political rather than scientific considerations, has been repeatedly put forward by OMT members.
Interesting about the dramaturgy of the Dutch "intelligent" lockdown is that we have seen a proto-professionalization as the critics imply that if they had seen any convincing scientific evidence, they might have been persuaded. Moreover, we see how antagonists criticize government for rigid measures following scientific advice, but use scientific "facts" in their argumentation, hence take up the language that they see belonging to the performance of which they see themselves a part. This proto-professionalization also comes out in the insistent epidemiological and virological frames with key words like "flattening the curve" "patient zero", and the "reproduction number R0". It pushes out "common concerns", just like the insistence of the OMT to keep schools open, as there was no evidence in the peer-reviewed scientific literature that school children contributed to the pandemic.
All in all, it was not politics and values that threatened the robustness of (the scientific evidence that underpinned) corona policymaking and measures. Scientific expertise is treated as credible and institutionally legitimate, precisely because the underlying values are openly and appropriately addressed. Building a robust factual basis for decision-making in modern society therefore requires a fundamental shift from a "Let-meexplain-it-once-more" repertoire and classical "science-forpolicy" dramaturgy, to a sustained dialogue in which the interplay between morality and science can be openly examined, questioned and discussed. The notion that such an approach might lead to the marginalization, or even dissolution of "substantive expertise", is misguided, as we cannot "wish away" that moralities together with facts decide whether society accepts policy truths as truths.
The boundary between science and values is diffuse but this does not absolve us of the responsibility to discuss these boundaries and open them up to dialogue. Other than "debate", which is aimed at stating differences, dialogue surfaces and critically reflects on underlying moralities (Scharmer, 2016;Van Burgsteden et al., 2021) by inquiry and "thinking together". It is good to realize that the science/politics border is negotiated in all kinds of ways, by all kinds of actors, and at all times. Negotiations range from determining the status of the overall scientific evidence on a specific theme ("contradictory" or "inadequate" in the case of the face masks), to stating that with 50% of the knowledge, we should make 100% of the decisions. Transparency in this area, including the struggle with the demarcation of facts and values, is an essential pillar for a reliable government policy, and a reliable role of advisory councils and knowledge institutions in this regard.
Citizens must be able to participate actively in this process, but we must be careful not to blackmail them with this participation (an active contribution is mandatory, otherwise you lose your right to participate). Or as the Dutch writer Remco Campert once said: "I myself was approached on the street by someone who asked me if I was a citizen. I just denied it to get rid of him." The desired participation options would be sustainable but also flexible and diverse in form. In addition to stakeholder participation and citizen consultation, we should also consider increasing the learning capacity of the government itself, so that it can quickly enter into a dialogue with changing coalitions of citizens, at moments that are difficult or impossible to anticipate.
The shift from debate, which creates winners and losers on the basis of the exchange (and presupposition) of points of view, to dialogue, which brings hidden assumptions to the surface to make them negotiable, is essential here. Above all, it requires the courage "to talk morality" in a world dominated by "fact-talk", and the commitment to constantly take seriously those who criticize a position as expressed by the powerful. So, authority is not derived from the institutional origin of the argument (e.g., the OMT), but is to be constantly and repeatedly enacted. It is also an approach that does not depend on a staging of authority but derives its authority much more from its open response to critics. The openness of this model is expressed in its willingness to take into account other non-scientific "ways of knowing" (Schneider and Ingram, 2007). Those situated understandings can often help explain why a universal knowledge claim does not work in a particular situation. Allowing those other sources of knowledge in may feel risky as it effectively means surrendering the sole authority of knowledge. Yet, if done well, this dialogue model can be very effective in building good, sustainable and truly democratic relationships.

DATA AVAILABILITY STATEMENT
The anonymized, raw data supporting the conclusion of this article can be made available upon request.

ETHICS STATEMENT
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.
Written informed consent was not obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.