Trustworthiness in the Swedish Strategies for Covid-19 in Recorded Press Conferences from the Public Health Agency of Sweden

Abstract

This article investigates structural patterns of eight recorded press conferences on Covid-19, and how the Public Health Agency of Sweden (PHA) builds trust in its information. The analysis draws on Wodak’s model, using the theoretical and methodological concepts discourse, texts, genres, situatedness, and discourse strand. The press conferences are regularly scheduled, always with the same structure. PHA uses a range of multimodal features with which they interact orally. Tension often arises in the dialogue between PHA and journalists wanting to know about death numbers, the spread of infection, and how citizens should act to stay healthy. Two discourse strands are identified: death numbers and face masks. PHA handles these discourse strands in slightly different ways. Journalists ask about the relatively high Swedish death numbers, which PHA does not deny, but is reluctant to talk about. PHA does not want to acknowledge the usefulness of face masks, saying that other measures are more effective. Both discourse strands pose potential threats to PHA’s trustworthiness. Journalists question why PHA’s death statistics show lower figures than those of other authorities, and they ask about the scientific basis for its face mask strategy, pointing to countries with different approaches.

1 Introduction and Background

When Covid-19 reached Sweden at the beginning of 2020, it fell to the Public Health Agency of Sweden (PHA)¹ to inform citizens how they should act during the pandemic to suppress the spread. The PHA’s strategy has been to provide such information via press conferences, assisted by representatives from The Swedish Civil Contingencies Agency (CCA)² and The National Board of Health and Welfare (NHW),³ who were also involved in handling the consequences of the pandemic. All press conferences were broadcast and available on YouTube. The material for this article derives from the online press conferences only (PHA Folkhälsomyndigheten 2021a). During parts of the pandemic, when the spread of infection was low, some journalists were on site, while others participated online. When the spread of infection was high, no journalists were allowed on site.

Already at an early stage, it became clear that the Swedish pandemic strategy differed from other countries (Rocklöv et al., 2020). Moreover, Sweden had high death rates, especially in elderly care home. Facts show that there were far more deaths in Covid-19 in Sweden than in the other Scandinavian countries in total (Bauhn, 2020, 1–2). For both of these reasons, we took an interest in analysing the PHA’s communication, assuming they would face challenges in trying to convey their messages and establish trust from the public, who at the press conferences were represented by the journalists.

Studies show that Swedes’ trust in authorities is unusually high in an international comparison. Polling from early 2021 shows that 50–57% of the Swedish public have quite high or very high trust in the PHA.⁴ During the pandemic, though, trust in authorities has been high in other places too. According to a recent survey covering seven European countries, 86.3% of the respondents reported being familiar with the WHO recommendations about Covid-19, while 59.8% trusted their information, and 15.5% reported distrust in the information from the WHO (Varghese et al., 2021, 5–7).

The PHA is an authority receiving its mandate from the government. It cannot issue restrictions to prevent the spread of Covid-19. In most other European countries, government has the power to intervene directly in the ongoing work of the authorities. In Sweden, the government must never control how the authority uses the laws, referred to as a ban on ministerial rule. In other words, the PHA is expected to provide citizens with general advice and recommendations, but it is up to the government to decide on how their information should be interpreted, and only the government can issue restrictions in any form.

Through the pandemic the PHA has encouraged personal responsibility, summed up in the motto “Together we can slow down the infection” (PHA, website 15 September 2021). The recommendations have changed somewhat during the pandemic, but the overall message is to keep your distance and stay at home when ill, watch out for symptoms of Covid-19, and wash your hands frequently. In this, the PHA’s strategy resembles those in many other comparable countries, such as Germany, Italy, and Spain, since the governments in these countries also highlighted individual responsibility and individual sacrifice as the solution to the ongoing health crisis (Sjölander-Lindqvist et al., 2020, 9–12; Wodak, 2021). But the PHA has not always been compliant with WHO recommendations, for example about the usefulness of face masks, which may have made it more difficult for them to gain public trust.

2 Trust and Trustworthiness

There is a lot of research on trust, trustworthiness, and credibility in various fields of research. In this study, we mainly rely on the sociologist Niklas Luhmann (1979, 1988), but also on the philosopher Per Bauhn (2020), and the economics researcher Lovisa Näslund (2018).

Trust relates to trustworthiness, and sometimes these concepts are used as synonyms, but according to the sociologist Niklas Luhmann (1979, 1988), it makes sense to distinguish between trustworthiness (Ger. Vertrauen, Sw. förtroende) and trust (Ger. Zuversicht, Sw., No. tillit, Da. tillid). Trustworthiness is a choice, implying that you have to ignore your vulnerability to assess someone as trustworthy. Trust, on the other hand is not a choice; it is a generalized attitude towards the outside world, that is, individuals’ general tendency to trust their surroundings. The alternative is to be insecure and afraid.

Trust and trustworthiness will therefore also work in different ways. Trust is not based on an estimate of a person’s character. Trustworthiness is interpersonal, while trust lies as a basic perception of the world in the individual (see Bauhn, 2020, 1–2). However, there are, of course, links between trust and trustworthiness. A person with high trust will probably also be more inclined to gain trustworthiness in others, and vice versa (Näslund, 2018, 7).

Trustworthiness thus helps us to deal with uncertainty. If there is no vulnerability, no trustworthiness is needed either (ibid., 2). We tend to gain trustworthiness from those who believe they can trust us, and vice versa: Those who are viewed with distrust will respond with distrust (ibid., 4).

There have been constant changes in the state of the pandemic and the state of knowledge about Covid-19. No matter the situation, the PHA is expected to provide useful information, advice, and recommendations. All in all, we see our analyses of the dialogue between the PHA and the journalists, who are the public’s representatives, as a kind of measurement of public trust in the PHA.

3 Aim and Research Questions

In this study, we investigate the multimodal world of the PHA’s online press conferences, how it establishes public trust in its information, advice, and recommendations, and how its trustworthiness is negotiated in communication with journalists. The study has a special focus on two topics: death numbers and face masks. We ask the following questions:

• 1. What are the structural patterns of the Covid-19 press conferences compared to political press conferences?

• 2. What kind of information does the PHA want to communicate?

• 3. What kind of information are the journalists looking for?

• 4. How are multimodal meaning-making resources used in the video clips?

4 Material, Methods, and Theory

This study is based on eight of the PHA’s press conferences on Covid-19, all of which were broadcast live and are available on YouTube. The sample is intended to show variation over time, to make it possible to follow the PHA’s communication from the beginning of the outbreak and as far ahead as possible within the period of this study. Table 4.1 shows the press conferences included in the material and basic data on the pandemic situation in Sweden presented by the PHA at each press conference. The duration of each press conference can also be read from the table; the shortest lasted for 33 minutes, the longest for 61 minutes.

Table 4.1 Analysed press conferences on Covid-19: date, duration, accumulated number of deaths, total cases, and cases in ICU or hospital care

We have chosen to study recorded press conferences from different seasons, as the Covid-19 virus spreads more easily in cold weather, leading to higher death rates in autumn and winter than in spring and summer (PHA, 2021b, 18–21). In addition, questions about public behaviour and protective measures of various kinds can be assumed to change in character and significance depending on whether the country is facing summer holidays, return to work and schools, or family holidays such as Christmas and Easter.

Table 4.1 shows that there was a sharp increase in both cases and death numbers between 16 March and 27 April 2020. The increase was relatively high throughout the period for both infections and death numbers, but it shows a clear dampening some time into the new year of 2021. This is certainly due to the vaccinations starting in December 2020 in elderly homes and accelerating during the winter and spring of 2021.⁵

The numbers of patients admitted to ICU or other hospital wards vary over time (Table 4.1). On 16 March 2020, no admissions were reported, but the number increased drastically in the spring of 2020.⁶ The risk of congestion in hospitals and health care facilities has been imminent from time to time but dropped radically during the later parts of the pandemic thanks to extensive vaccination.⁷

As a method, we use Wodak’s (2021) categories for speech, text, and image in a critical discourse perspective, with multimodal and communicative aspects included. Her article is also about crisis communication concerning Covid-19. Her claim is that “[m]ost governments employed specific modes of crisis-communication vis-à-vis the COVID-19 pandemic, depending on the respective socio-political context and historical tradition” (ibid., 1). The basis of Wodak’s model is discourse, frame, and legitimation. She outlines the method in the following eight steps in her model:

• 1. Activation and consultation of preceding theoretical knowledge.

• 2. Systematic collection of data and context information.

• 3. Selection and preparation of data for the specific analyses.

• 4. Specification of the research question/s and formulation of assumptions.

• 5. Qualitative pilot analysis, including a context analysis, macro-analysis, and micro-analysis.

• 6. Detailed case studies.

• 7. Formulation of critique.

• 8. Practical application of analytical results. (Wodak, 2021, 6)

The present study is an in-depth analysis informed by the research questions. As an analytical tool, we use Wodak’s methodological steps presented above but exclude numbers five and eight, as we have not conducted a pilot analysis (5), and we cannot suggest practical applications from our results (8). We believe that the six other steps provide a good methodology for the in-depth analysis focusing on genre traits, multimodal meaning-making, and information given by the PHA versus information asked for by the journalists. This approach allows us to discuss how the PHA works to establish trustworthiness, what kind of meaning-making resources are used, and how the topics are framed in the communication between the PHA and the journalists.

As theoretical concepts, we use discourse, genres, situatedness, and discourse strands. Wodak views discourse as “a set of context-dependent semiotic practices” which are “socially constituted and socially constitutive, related to a macro-topic, characterised by a pluri-perspective” and linked to argumentation (2021, 6). In the present study, the discourse is crisis communication via the PHA, communicated through various media, but we concentrate this study on recorded press conferences. Wodak points out that “the key analytical categories of thematic analyses are discourse topics, which, conceptually, summarize the text, and specify its most important information” (2021, 6; see also van Dijk, 1991, 113). Thus, discourse topics provide relevant frames and interpretation frameworks.

The term discourse strand is understood as topical threads within discourses, and they are distinguished by topical continuity and boundedness (Rheindorf, 2019, 210–211; Wodak, 2021, 4–6). Specific discourse strands can be identified through subsets of data within a corpus, representing the discourse they are part of. In many instances, an initiating event or events can trigger debates that feed into such strands, in Wodak’s study, for example, the official announcement about lockdown came in March 2020, and most European countries decided to quickly deal with the crisis (2021, 4).

Texts (audio, spoken, visual, or written) are realized in specific genres and viewed in terms of their situatedness (Reisigl & Wodak, 2001, 40 ff.). The concept of genre is widely discussed, which makes it difficult to summarize the genre-theoretical approaches from different disciplines. We follow Swales’s angle, advocating that genre is associated with prototype performances, and that genre is characterized by repetition, as well as variation and change (see Swales, 1990, 44f). Consequently, a delimitation of the genre concept could include the following points: A genre connects texts to recurring social processes where interaction through text is central. It also includes prototype notions of design; it is normally named, and linguistically and socially codified (see Ledin, 1996, 26–30).

In this study, genre applies to recorded press conferences on YouTube. According to Ekström and Eriksson, a press conference “represents an institutionalised arena for political actions and journalistic interrogations” (2018, 352). Press conferences possess a ceremonial character, and “serve as symbolic representation of public political accountability, but they are also in a context in which political discourses are articulated, (re)enacted and negotiated” (Ekström & Eriksson, 2018, 352). Typically, the authority over the conversation belongs to the authorities or politicians who establish the norms for participation, but through their questions, journalists can partly control the press conferences (Eriksson et al., 2013, 121–122; Ekström & Eriksson, 2018, 342–343).

We assume that meaning-making is situated. Consequently, the development of the pandemic affects how meaning-making in the press conferences is processed through different semiotic resources, i.e. artefacts, actions, and materials (see Van Leeuwen, 2005, 285). Further, we assume that all meaning-making is multimodal (Kress & van Leeuwen, 2020). Typical modalities of importance for this study are speech, texts, and image.

It should be stressed that we limit the multimodal integration in the analysis to frames of verbal, visual, and auditory modes in the live broadcasts now available on YouTube. In addition, the communicative and multimodal situation for broadcasts and the press conferences on site differ in that the live events are compressed into a one-screen window of audio-visual communication in the YouTube films.

5 Analysis and Results

Below we first report genre-specific features in the Covid-19 press conferences, second the dialogue in the themes: Death numbers and face masks with the integration of multimodal tools in the recorded press conferences, and how the tools are framed and situated through different multimodal resources in the YouTube recordings. The PHA uses those tools to establish trust and negotiate the credibility of the information.

5.1 Genre-Specific Features in Covid-19 Press Conferences

At the press conferences generally, the PHA practises its most important safety measure in a very visible way: When the first press conferences were broadcast, the journalists were sitting remarkably close together in the room. Later on in the pandemic, they were sitting further apart, and when the infection was spreading dramatically, everyone participated by link, including the PHA and other officials at the press conference. It is possible that this action had a broader purpose, to create trustworthiness.

All press conferences on Covid-19 have the following structure: First a brief monologue by the PHA followed by a longer question session with journalists. A press secretary from the PHA moderates the press conferences. As at other press conferences, a hierarchy has developed that allows the news media considered most influential to ask their questions first, such as TT News Agency, Swedish Radio, Ekot (The Echo, a news programme on public service radio), and Swedish Television (SVT), and the big national and international newspapers. Then the media can speak in descending order of importance.

The PHA press conferences differ from political press conferences in several respects. The main differences from political press conferences at the macro level are that press conferences on Covid-19 recur on a daily basis and later on twice a week throughout the whole pandemic. At press conferences in general, it is usually only accredited journalists who are invited, but at the conferences about Covid-19 there are quite often medical experts asking questions, such as Lena Einhorn, a doctor and virologist, and Benjamin Kalischer Wellander who is a doctor. These two, and several other doctors, are critical of the PHA’s Covid-19 strategy.

There are many activities going on during the Covid-19 press conferences. Different actors from three different public authorities (PHA, NHW, CCA) with different tasks are present and sometimes invited to answer questions or present information. The PHA also uses a wide range of different visual resources. The multimodal elements mostly concern the beginning of the press conferences. In the information part of the press conference, the PHA speaks with the support of PowerPoint presentations. In addition, sign interpreters are embedded in the video, translating the conference into sign language. Moreover, oral information is continuously transferred into a subtitle strip at the bottom of the video.

5.2 The Topic of Death Numbers

On 16 March 2020, the first Swedish deaths from Covid-19 had occurred. Six people had died, but the PHA added that “there are more [deaths] on the way”, because of a certain lag in the reporting. The PHA wants to give the impression of being in control in order to maintain their trustworthiness. Its discursive strategy in this situation seems to be to instil calm and at the same time prepare the public for a potentially more serious situation, which is parallel to Wodak’s results (see Wodak, 2021, 8, 14–15). When the Swedish public service television (SVT) asks about how to assess public health in relation to the death toll, the answer is somewhat reassuring: The deaths are basically limited to Stockholm, and in addition there is an excess mortality of “about 1,500 from the flu alone” every year. Still, they conclude by emphasizing the importance of “working to deal with a much worse situation”.

SVT then follows up with a question about how many deaths it will take for the Covid-19 virus to be “classified as something out of the ordinary” and receives the following answer:

We do not know how relevant it is to calculate it…. We do our best to reduce the risk of death as much as possible. And it is very much about trying to protect the elderly from infection. And now we have time to do it. In Italy, they did not see this until it was a bit late to do so, because they already had the infection in health care and nursing homes. We will work very hard together to try not to end up in that situation. (16 March 2020)⁸

This quotation reveals a certain reluctance to talk about death numbers and calculations. Instead, the PHA turns to its overall strategy, i.e. to slow down the spread and protect the elderly, points which are always clearly emphasized in words and images in the introduction to each press conference. The PHA seems to invalidate the journalist’s question by pointing out that it may not be relevant to count. This creates a tension in the dialogue between them, with regard to the specific question as well as to the general strategy. In fact, it is one of the PHA’s main tasks to collect data and make calculations about death numbers. Accordingly, this could be an instance where their trustworthiness is undermined.

The question about death statistics triggers debate and creates a clearly defined discourse strand, recurring on several occasions and “distinguished by topical continuity” (Wodak, 2021, 6). Creating tension in the communication between the PHA and the journalists, this discourse strand is a possible threat to the PHA’s trustworthiness, because it can undermine the general trust and the public trust in the PHA in the long run.

In the spring of 2020, the death rate rose dramatically in Sweden. At the press conference on 27 April, the PHA’s representatives initially states that despite the high numbers now reported, there is also a certain lag in the statistics, so the number is probably even higher. On this occasion they are supplemented by the NHW, who present their cause of death statistics, indicating a clear Covid-19 effect. Thus, their statistics differ from those of the PHA. The statistical discrepancy raises questions from journalists. TT News Agency asks why the NHW reports 10% more deaths than the PHA. The answer is that the two authorities use different measurement methods:

But the big difference is that we also have cases that have not been laboratory-tested but where the doctor in charge has documented that it is a Covid-19 patient, even though the person has not been laboratory-tested. (NHW, 27 April 2020)⁹

Unlike the PHA, the NHW thus include deceased persons who have not been laboratory-tested. The journalist at TT then turns to the PHA to find out if their statistics will always show too low numbers. The PHA admits that there will be a certain discrepancy but argues that consistency in the measurements is a crucial factor for the credibility of the statistics. However, TT calls for further clarity regarding the relationship between the two authorities’ statistics and turns directly to the PHA: “So, you do not consider NHW’s figures, so to speak, in this situation then?” The answer is brief: “We take it to heart, but we have not added them [the numbers], no”. This vague answer pointing in two directions, could potentially damage the PHA’s trustworthiness.

TV4 then asks why they did not test all the deceased for Covid-19, when Sweden has now received international criticism for the high spread and high death rates:

Yes, it is good that people look at us, see how we do and compare and so on, but we believe in this strategy, and we work from it.… And so far,… it [the death rate] is no longer increasing in the same way as it did before, so that now I think we’ll stick to this and help keep the curve where it is. (16 March 2020)¹⁰

Here, the PHA spokesperson says that they believe in adhering to their overall strategy, which, above all, stresses everybody’s personal responsibility, a fundamental element of the pandemic strategy also highlighted by Wodak (2021, 15). The PHA follows its discursive strategy to instil calm by pointing to the relative stagnation in mortality, but at the same time they introduce a discursive strand that will create tension in the dialogue with the journalists, namely comparisons between Sweden and other countries (see ibid., 14–15). As a result, the South Coast, a Swedish news site based in Marbella, wonders how it can be that Andalusia with 8.5 million inhabitants has only half as many dead as Sweden, even though the pandemic outbreak occurred almost simultaneously in Sweden and Andalusia. In this case, the PHA does not really answer the question:

I cannot comment on that, but it is clear that an extremely strict lockdown also has other effects, you balance different things.… The disadvantage of a hard lockdown is that there is not so much spread and then you are left with the problem, so to speak, when you ease the restrictions. (27 April 2020)¹¹

The PHA here reinforces trustworthiness in the Swedish pandemic strategy by focusing on possible future drawbacks with hard lockdowns and suggests that the present high numbers are due to the mild restrictions applied in Sweden (see ibid., 14–15). During the summer of 2020, death numbers fell significantly. In the introduction to the press conference on 1 September, the PHA emphasized that there was now zero excess mortality. Consequently, mortality stayed at a normal level for the season.¹²

In the late autumn of 2020, however, the numbers started increasing again, and on 17 December, the PHA warned that they would “most probably continue to increase”. Relating to their overall strategy, they stressed the importance of stopping the spread.¹³ The newspaper Expressen followed up with a very direct question: “Do you think we have failed?” to which they received a somewhat evasive answer:

Considering how many have died in Sweden, it is deeply regrettable, and we will in the end have to try to understand what we can do better to prevent it from ever happening again. Whether one should call it a failure or not, I think we should leave it to those in the future who will investigate these issues and decide about that. (17 December 2020)¹⁴

At this point, the PHA’s trustworthiness faces a threat through the words failed and failure connected to the Swedish pandemic strategy. Clearly, the PHA does not want this issue to become a discursive strand. Its communicative strategy is to neutralize the issue by postponing it to the future; it is not yet time to evaluate the pandemic strategy, nor is it an issue for present-day researchers.¹⁵ While the PHA here avoids touching on the Swedish pandemic strategy and possible factors behind the high death numbers, they admit the need to learn from experience for future pandemics, which might constitute a certain concession to the journalists.

By 14 January 2021, more than 10,000 have died from Covid-19. Even now, the PHA admits, the statistics are incomplete to some degree. They also expect that the increase will continue, as there is still an extensive spread in the country. This statement provokes questions from the journalists, about the Swedish pandemic strategy in relation to mortality and about statistical methods. Both issues have been on the table during previous press conferences.

The questions from the journalists concern whether concrete measures are right or wrong and ask about causes and effects, while the PHA rather talks about overall strategy. The newspaper Expressen wants to know whether the high number is in any way related to the Swedish pandemic strategy. Again, the PHA seems slightly reluctant to answer the actual question, as the spokesperson argues that deaths from Covid-19 must be regarded from a holistic health perspective. This could be an attempt at a discursive shift concerning death numbers in order to reinforce trustworthiness for the information from the PHA, which can lead to improved trust in the PHA in the long run (see Wodak, 2021, 6):

But the corona strategy itself and our handling of it must be seen from more perspectives than just the deaths. You must look at the whole health in society and the consequences of the measures in relation to it. (14 January 2021)¹⁶

The strategy to establish a discursive shift does not seem to be successful, since a journalist from Vetenskapsradion (Science Radio) immediately asks whether deaths are now measured according to different principles than before, and whether it is now possible to make comparisons between countries. The PHA denies that anything has changed but admits that there is still a difference between the NHW and the PHA figures. On the other hand, Vetenskapsradion does not receive an answer to its question about international comparisons.

The statistical method is later revisited in a question from the online newspaper Bulletin, which, due to the lag in statistics and the fact that these only include deceased persons who have been tested, raises the question of a possible dark figure. The PHA answers:

We will not change our statistics in any way. But, on the other hand, when assessing the whole in terms of the effect on death rates in the country, one needs to look at different statistical sources.… To look at the excess mortality that is updated on our website, to look at the NHW’s statistics on deaths within 30 days. (14 January 2021)¹⁷

The answer here is ambiguous, as it is initially stated that the statistics should not be changed, at the same time as it is obviously considered necessary to “look at different statistical sources” to get an overall picture of the situation. The latter part of the statement could be regarded as a concession to the journalists, and thus as a step towards releasing the communicative tension that the discourse strand about death numbers has created between the PHA and journalists.

The newspaper Bulletin, however, returns to the significance of the hidden statistics for the overall picture and receives the answer that the PHA has an overall view of this because they can use excess mortality as a reference point:

It is the excess mortality that we think gives a good measure, because we had a very low flu season this year. So, it is Covid-19 that is responsible for the large mortality in these groups that make up the large groups where we have deaths. (14 January 2021)¹⁸

Previously, the PHA has dismissed international comparisons. Now, they say that it might be relevant to compare mortality from Covid-19 between countries, if excess mortality figures are calculated.

Clearly, during the period of investigation, the PHA has made a discursive shift in the matter of statistics and death numbers, but they have never explicitly commented on it. This turn of opinion could be interpreted as an attempt by the authority to partly neutralize the discourse strand of death statistics, which has often posed a threat to its trustworthiness.

5.3 The Topic of Face Masks

The Swedish strategy to protect people from the virus differ from other liberal democracies. Personal responsibility is invoked, and people are not ordered by the authorities but are asked to keep their distance. The Swedish overall strategy has a clear impact on the PHA’s attitude to face masks.

At the first two press conferences included in this study, i.e. 16 March and 27 April, face masks are a non-existent discourse strand in the press conferences. On 27 April journalists ask if more protective measures are needed, as there is information from several parts of the country that the PHA’s advice is not being followed; there is congestion in public transport, restaurants, bars, and outdoor cafes. The journalists still ask questions indicating trust in the PHA, and when asked directly by the newspaper Göteborgs-Posten if there is “any other call to how to behave” than to keep distance from others, the PHA simply answers “no”, without further explanation.

In spite of this, face masks receive increasing attention in the spring and summer of 2020, when the WHO recommends the use of face masks:

If Covid-19 is spreading in your community, stay safe by taking some simple precautions, such as physical distancing, wearing a face mask, keeping rooms well ventilated, avoiding crowds, cleaning your hands, and coughing into a bent elbow or tissue. Check local advice where you live and work. Do it all! (World Health Organization website 2021)

As the quotation shows, the WHO’s advice largely coincides with the PHA’s, but not in the case of face masks. This is probably why the issue receives increasing attention from journalists. They will later make face masks into a discourse strand triggering communicative tension between the PHA and the journalists, thereby challenging PHA’s trustworthiness, who must now find ways to legitimize their position (see Wodak, 2021, 7–8).

On 7 May and 16 June, critical questions are being asked about the PHA’s views on the use of face masks. Journalists believe that the PHA is vague in its communication on this point, others wonder if it is not time to add the use of face masks to the list of protective measures. However, the PHA is consistent in its discursive strategy, claiming that other hygiene measures are much more important than face masks. According to the PHA, face masks can even prove counterproductive in the fight against the pandemic, as they can induce a false sense of security, and thereby cause the wearer to fail to follow other more effective hygiene advice.

Throughout the period covered in this study, the PHA is consistent in its discursive strategy, insisting that face masks are not crucial in fighting the spread of infection. At best, the PHA says, face masks can be a complement, but only in certain limited situations (16 June 2020). According to the PHA, the public would have difficulty managing the face masks properly. It is difficult to determine how this position, that face masks are best left to the experts, affects the PHA’s trustworthiness and the communication with the journalists. What we can say, though, is that journalists throughout the period of our study will revisit this discourse strand, creating tension in their dialogue with the PHA (see Wodak, 2021, 5–6).

It is very unusual for the PHA to raise the issue of face masks, so in this respect the press conference on 1 September 2020 is an exception. The main message is the same as before, meaning that limited use of face masks may be considered:

The issue of face masks has now, as it were, landed in the position that there may be a role for face masks at the local level when everything else is in place.… Then there may be a signal value in introducing face masks in certain situations. (1 September 2020)¹⁹

At first sight, the quotation above indicates a discursive shift, since the PHA admits that there may be a function for face masks. Yet the language is vague, marked by modifiers and reservations (as it were; may be a role; when everything else is in place; may be a signal; in certain situations). The PHA also makes clear that the effect of face masks is symbolic rather than real. Thus, it seems that there is no actual shift in strategy, just an apparent one, maybe a concession to the citizens, and to WHO’s strategy for face masks. The PHA then concludes the sequence by saying that “we remain in the centre, what we do is to…” They leave the sentence unfinished, pointing to the picture where the advice is listed, which does not include the use of face masks. This underlines that the agency’s approach to safety measures follows a consistent line.

During question time, the newspaper Aftonbladet’s journalist wonders if the PHA regrets having previously excluded face masks from the advice on how to fight the infection. Once again, the PHA follows its discursive strategy that face masks would not have fulfilled any function earlier in the pandemic, and that they can only be assumed to be effective in specific activities for a limited time, to build trust:

So, I strongly believe that one should do it as a targeted effort for a brief period and have a particularly good value signalling that there is a problem right now. And it can provide a certain protective effect in public transport and other things but having it for extended periods of time in different places does not seem to have worked well anywhere. (1 September 2021)²⁰

The journalist’s choice of the verb regret when asking if the PHA regrets excluding face masks from the list of advice, challenges the PHA’s trustworthiness, implying that the authority has been wrong in the past and there is, in fact, something to be sorry about. In the answer, the PHA avoids the topic of regret. Instead, the PHA follows its discursive strategy, but with the admission that face masks could have a positive effect in public transport. On the other hand, the PHA legitimizes its strategy by saying that there are clear indications from other countries that widespread use of face masks does not serve its purpose (see Wodak, 2021, 7–8).

On 17 December, the question comes up again when a journalist from TT wonders if it may be time for the PHA to revise the advice on face masks. The answer is no, but indirectly, in that they immediately refer to the weak research situation in the field:

No, unfortunately there are very few studies. There was an interesting study from Germany similar to the one done in the city of Jena some years ago. But it has the same basic problems as many others. Otherwise, we have the Danish study, which is quite clear, so we do not have much data, unfortunately.… There is, sort of, no firm position on that issue. (17 December)²¹

In the quotation above, the PHA claims that the existing international studies have some fundamental problems. They do not, however, indicate what the problems are, but believe that the issue is not yet scientifically settled. Here, the journalists do not choose to challenge the PHA’s trustworthiness, which might be due to the PHA legitimizing themselves as scientists, knowing how to decide if a problem is scientifically settled (see Wodak, 2021, 7–8).

During Christmas, the spread increased in Sweden. At the press conference on 14 January 2021, a journalist from SVT asks how the PHA views this. They respond by repeating some of the ordinary advice, but at the same time call for the use of face masks in public transport during rush hours:

We think it is important to adhere to the recommendations for the infection before you get symptoms. Therefore, it is important to work from home, keep your distance and use face masks when there is congestion in public transport. And that employers consider a situation where people are close to each other indoors for a long time in small premises, and consider face masks, and there are recommendations about that on our website. (14 January)²²

The quotation above signals a discourse shift. A new approach to face masks emerges, something that has also had an impact on their website. However, it is a silent discourse shift, because the changed approach is not justified verbally, or commented on, by the PHA.

Despite the slight discourse shift with regard to face masks, they are never an item on the list of protective measures. On 17 June 2021, when the Swedish government has decided that the restrictions will end 3 months later, the journalists ask many questions about protective measures before the opening of the country. Again, the PHA does not mention the use of face masks, which apparently prompts journalists to challenge the authority’s trustworthiness. Accordingly, a TT journalist wonders what to do if it gets crowded at workplaces and in public transport, if distance to others is impossible. The answer is that people must still try to keep their distance and follow the list of recommendations. Thus, the conclusion to be drawn from the discourse strand of face masks is that the PHA on certain occasions makes slight changes in its discursive strategy, but that these changes have no impact in practical life.

5.4 Situatedness and Multimodality

Situatedness is crucial for understanding the meaning-making through semiotic resources that takes place in the PHA’s press conferences (see Van Leeuwen, 2005, 285). The communication in the press conference venue is affected by multiple factors: the development of the pandemic world-wide and in Sweden, discourses about the pandemic in other places or media, and the social and epistemological relation between the PHA and the journalists. Furthermore, the information provided in the press conferences should be accessible to everyone, no matter whether you follow them on place, broadcast on radio, television, or YouTube.

The physical room of the press conferences is of special interest in relation to the use of multimodality. There is a clear pattern through all press conferences in the PHA’s use of visual resources. All sessions start by the PHA showing and explaining a set of charts, diagrams, and statistics about the recent development of the pandemic in Sweden and internationally, advice for containing the spread of the virus, etc. The question session, on the other hand, is mainly oral without the additional visual resources, but the PHA often refers to the charts and statistics when answering journalists’ questions. Thus, statistics, diagrams, and charts are used to emphasize the verbal information and thereby to build trust in the message as a whole. The importance attached to multimodality is finally illustrated at the end of all press conferences, when the PHA shows the poster with the advice that people should follow to protect themselves and others from the virus (Fig. 4.1). The sign says:

Fig. 4.1

A screenshot of protective measures presented on 17 June 2021, with the permission of the PHA

Slow down the infection by keeping your distance.

Keep a distance from others both indoors and outdoors.

Spend time in a small circle, avoid new close contacts.

Stay at home even if you just feel a little sick.

Work from home if possible. (PHA, 17 June 2021)

Multimodality can be a double-edged sword, though. When overused, it can be confusing, since everything happens simultaneously. Hence, it is hard to know, as a listener or viewer, where to focus your attention. Figure 4.2 illustrates this situation with a lot of graphics, text, subtitling, and visual interpretation for those with impaired hearing.

Fig. 4.2

A screenshot from the PHA’s press conference on 27 April 2020, with the permission of the PHA

Of course, viewers might learn over time how to focus and to read the multimodal signs at the press conferences. Still, everything that happens on the screen entails a large cognitive load that can make it difficult to absorb the message from the PHA in the video clips.

A press conference in the material that deviates in several ways from the usual pattern for the PHA’s press conferences was the one on 17 June 2021. The introduction was unusually long, 13.5 minutes, as opposed to the usual 4–5 minutes. In addition, the PHA comes back when all other officials have spoken, which also deviates from the pattern. This press conference was just before the Midsummer weekend, which is an important Swedish family holiday. Themes for the day are that there is a continued reduction in the spread of Covid-19, but the PHA nevertheless emphasizes that citizens need to be aware of new outbreaks and that infection tracing is important. The general guidelines have been extended but adjustments are planned. People are encouraged to follow the advice: to stay at home at the slightest symptom, avoid festivities with many guests, and socialize outdoors.

On the same day, the PHA’s press conference routines or genre rules are put to the test from a somewhat unexpected angle. This time, it is not a critical question, but a journalist from a foreign news channel, in clear support of the PHA using visual resources to highlight the benefits of the Swedish pandemic strategy. Behind her, written texts such as “Award winner, A mother’s quest for more common sense in the Pandemic, and Covid, Tango and the Lagom Way” idealize the PHA. All in all, it is difficult to understand and analyse what the journalist wants to say with all the illustrations behind her, but her positive attitude to the Swedish strategy and especially Anders Tegnell is clear. She portrays Tegnell, the Swedish head epidemiologist at the PHA, as a rock star. The text with the message Tango and the Lagom Way is not completely clear, but tango is a dance that requires two, i.e. “it takes two to tango”. Possibly, it is the mutual trust between the PHA and the people in Sweden that is invoked, and that strategies in other countries are based on control and not on trustworthiness. The control strategy has in some countries led to protests in the streets.

Of special interest from a communicative point of view is perhaps the use of the word lagom, meaning “just right, not too much, not too little”, and commonly seen as the essence of Swedish mentality. But the journalist’s multimodal action during the press conference is absolutely not lagom, but a cornucopia of various images and statements to celebrate the beauty and adequacy of the Swedish Covid strategy. The word lagom therefore has a strong symbolic function, in this case not least since it relates to the expression common sense (folkvett; see Wodak, 2021, 16), which has a prominent place in the journalist’s collage. In addition, her background screen shows a picture of Anders Tegnell, the PHA’s representative at nearly all press conferences. She also requested permission to record the interview during question time. Doing this, the journalist broke the genre conventions of objectivity and distance that otherwise characterize the press conferences about Covid-19.

The PHA, in this case Anders Tegnell, makes no comment about the unorthodox background screen, but briefly refuses the journalist’s request. The moderator then refers her to question time and hence shows that rules apply to everyone without exception. Seemingly, the PHA’s trustworthiness as an authority remains unharmed thanks to these measures, since the situation is left uncommented by journalists present at the press conference, and elsewhere.

6 Discussion of the Results of the Analysis

Relating to the categories discourse, texts, genres, situatedness, and discourse strands (Wodak, 2021, 5–6), we discuss how the PHA handles the tension in the dialogue with the journalists, when trying to create and maintain trustworthiness. The discourse in this study concerns crisis communication in recorded press conferences where, in short, the Swedish pandemic strategy is presented and negotiated. Ultimately, the actual crisis is about people being afraid that they themselves, their relatives, or friends will catch Covid-19, become seriously ill and even die. We note that tension often arises in the dialogue between the PHA and the journalists. Of course, tension is an integral part of the press conference genre, but this is still a symptom that trustworthiness is at stake.

On the one hand, there are the representatives from the PHA providing facts and instructions about the pandemic, thus trying to calm and reduce anxiety in people living in Sweden. On the other hand, there are the journalists who, in their professional role, are representatives of the Swedish public, seek guidance and answers to very specific and often critical questions about protective measures that just about anyone would have asked, thereby challenging the soothing messages from the PHA.

The major difference between the PHA’s press conferences and press conferences in general, is that those on Covid-19 are regularly scheduled. In fact, they were broadcast every weekday in the beginning of the pandemic. Another salient difference is that the PHA uses a range of multimodal features, such as images and graphs, speech, text, and sign language to underline their trustworthiness, among other goals. Their oral presentations are at all times supported by a PowerPoint presentation. Oral comments explain the visual data most of the time but not always. Sometimes they collide, or the speaker may let data speak for itself by pointing to the PowerPoint saying nothing, where it could be appropriate to give an oral explanation, since it is not always clear what part of the image the PHA is referring to.

It could be said that the PHA’s recorded press conferences are like the film Groundhog Day, about a man who has to relive the same day over and over again, only with some minor difference.²³ The situatedness contributes to this impression, since the events always have the same structure and identical multimodal elements. In this respect, the press conferences are very predictable, and the attending public thus knows what to expect from them. Sometimes, though, technical problems cause difficulties in hearing what speakers say. The abundance of sound, text, image, sign language, and various people acting in the video may also cause cognitive overload and make it difficult to absorb the information, despite the fact that all these semiotic resources are justified by accessibility.

During question sessions, it becomes apparent that the journalists often want to address topics not introduced by the PHA at the beginning of the press conference, such as statistical methods for calculating the number of deaths or the usefulness of face masks. There is thus a discrepancy between information requested and information given, resulting in communicative battles between the PHA and the journalists. Sometimes it can be difficult for the general public to follow these battles, since they often concern issues requiring some degree of medical or statistical expertise.

The PHA has a communication problem, because they have to teach people new behavioural patterns to stop the pandemic and to convince people that their strategy is right. On the one hand, they are expected to provide general advice and recommendations during the pandemic. On the other hand, they are not allowed to issue any form of restrictions to prevent the spread of Covid-19. Only the government can do that. This ambivalent position, being in between the government and the public, could be difficult for the PHA to manage. In this study, there are indications of these difficulties in the discourse strand about face masks, where we noticed a discursive shift, from a negative view towards a more positive view. This discursive shift did not result in any change in the PHA’s recommended protective measures, though. We cannot ignore the possibility that the reluctance to add face masks to their otherwise relatively short and invariable list of recommendations, can be connected to their ambivalent position, and the fact that recommendations from the PHA are not even close to coercive measures.

At the start of the pandemic, there is a lack of research data on Covid-19, and the situation is constantly changing, but as the analysis shows, the longer the pandemic progresses, the more research results are presented. Clear information about death numbers and protective measures is required for the public at all times, but the information and advice given often appear like oscillation between vagueness and certainty, which can create uncertainty about how citizens should relate to the pandemic in different situations. For example, in the discourse strand about death numbers, the PHA expresses certainty in its own strategy, claiming that the situation is always under control. At the same time, they apply a vague communicative strategy when asked direct questions about deaths or statistics from the journalists, as they often avoid the actual question, often pointing to their general pandemic strategy, or sometimes even invalidating the question. In these cases, the PHA uses repetition as a rhetorical tool, saying it is important to keep the curve down, reduce the spread, protect the elderly, etc. Finally, they seem more willing to talk about deaths when numbers are low than when they are high, and this adds to the impression of vagueness in their communication. This vagueness in communication appears to open for questions about precisely the matters the PHA does not want to talk about, namely statistical methods, deviations, and international comparisons.

We observed the same communicative pattern in the discourse strand about face masks, i.e. reluctance to address the issue and vagueness in expression. For example, the PHA spokesmen never bring up face masks in their introductions to the press conferences. Referring to the weak scientific basis for a widespread use of face masks, it seems as if they try to neutralize the issue, even though they never address the shortcomings of the studies mentioned. Further, questions about whether it is time to recommend face masks in public spaces are regularly met with references to the PHA’s standard list of protective measures; repetition is a frequent rhetorical tool in these situations.

The PHA wants to spread its message, and the journalists want to take the information from the press conferences and turn it into spoken and written news. The present study revolves around issues of trust and trustworthiness. The central question is whether the PHA’s communication about the spread of the infection is trustworthy in the eyes of the journalists as representatives of the public. The answer seems to be that it depends on varying factors. Their vague rhetorical strategies, together with the reluctance to address topics where journalists seek answers, may pose threats to their trustworthiness. Their consistency in keeping to their pandemic strategy, only rarely making minimal changes in their list of protective measures, may on the contrary have signalled stability and thereby reinforced their trustworthiness.