Communicating uncertainty in times of crisis. The legitimization of EU’s response to COVID-19 pandemic through the discourse used by the President of the European Commission

The aim of this article is to investigate the crisis communication narratives and how the European authorities approached the task of communicating uncertainty during the Covid-19 pandemic. Using the critical discourse analysis approach, t he study explores how the legitimization of the EU’s response was discursively built and if there was a distinction between uncertainty information and uncertainty experience in terms of crisis communication. To answer the question: what are the best ways for institutions to communicate uncertainties to public audiences in order to benefit from legitimization and trust for their actions? We analysed the official communication of the EU, namely the discourses and press statements by the President of the European Commission. The data were processed with Atlas.ti 8 and generated the points under discussion - correlations, narratives, linguistic recurrences, dynamics of main themes, and deconstruction of uncertainty. The study’s findings support the importance of separating between uncertainty information and uncertainty experience and the continuity of message lines.


Introduction
The reputation and blame avoidance techniques of private firms have typically been the focus of crisis communication research. As a result, there is insufficient crisis communication knowledge among public sector organizations. This is problematic since public organizations are responsible for a substantial portion of the planning, communication, and management of large-scale crisis situations. Risk communication, according to the World Health Organization (WHO), is "the process by which national and local government authorities provide information to the public in an understandable, timely, transparent, and coordinated manner before, during, and after a crisis; also, it promotes effective information exchange and coordination among scientists, public health and veterinary experts during the alert phase to better assess, manage and coordinate preparedness and response activities" (WHO, 2012, p. xii).
During public health emergencies, communicating with the public is a complicated process involving many stakeholders. To successfully impact healthprotection behaviours, authorities' messaging to the general public, targeted populations, and other stakeholders must be properly prepared. Messages from authorities during an emergency occurrence, in particular, must properly express the ambiguities around scientific evidence as well as what is known about the event's impact and progression . Public crisis communication is continuously challenged by the following characteristics of a crisis: 1) a rapidly escalating public health concern that requires a drastically increasing demand for health-related information that frequently exceeds the capacity of health authorities to deliver it; 2) prior to making a recommendation, there is a need to interact with potentially at-risk populations; 3) and a fast changing environment in which information regarding the health threat and how to stop it from spreading is insufficient and changing (Reynolds, 2002;Reynolds and Seeger, 2005).
Experts frequently emphasize that one of the most significant challenges for organizations in handling crises is that they are frequently ill-structured and complicated (Claeys and Cauberghe, 2014, p. 185). This article seeks to contribute to the knowledge of how organizations, primarily government organizations or government employees, react to or seek to produce risk communication messages that legitimize the actions of institutions and have people's trust. Also, this article examines a critical use of language in society: the legitimization process. It demonstrates how language functions as a control instrument (Cho and Gower, 2006) and exhibits symbolic power in discourse and society. Using prior legitimization research (Massey, 2001;Van Dijk, 2005), this paper discusses and suggests some fundamental legitimization methods used by social actors to legitimize their actions. The legitimization tactics can be employed alone or in combination with others to justify social behaviours by: (1) emotions (particularly fear), (2) a hypothetical future, (3) rationality, (4) voices of expertise and (5) altruism (Zhao et al., 2013). This article discusses how these techniques are built and molded verbally. Furthermore, it uses instances of speeches delivered by European Commission President Ursula von der Leyen to demonstrate how these discursive structures and strategies are used to justify restrictions, lockdown, and other human rights adjournments.

Literature review and conceptual framework
While the study of crisis communication dates back to the 1950s, it was not until the mid-to-late 1990s that the discipline began to develop as a unique area of study within public relations and organizational communication, with a substantial influx of research in the recent decade. This means that, despite the fact that there are several books, edited volumes, and journal articles on the topic, the discipline is still developing. Because crises are fundamentally public events (Benoit, 1997), crisis management and crisis communication have long been linked to themes such as strategic planning (Ha and Boynton, 2014)) and problems management .
According to the systematic review of literature (Diers-Lawson, 2017), while crisis communication research is most often published in journals focusing on management, business, social science, and, of course, communication, it is also published in journals focusing on health, science, technology, and industry-specific needs. The emergence of descriptive theories like image repair theory (Benoit, 1995) and situational crisis communication theory has been emphasized in the literature around this dominating approach to crisis communication (Coombs, 2006). In order to produce strategic messages that help an organization handle its crisis challenges, crisis response requires crucial decision-making that balances the nature of the problem, the organization, and stakeholders (Coombs and Holladay, 1996). The identification of crucial crisis response objectives (Chang et al., 2015), targeted stakeholders (Sung and Hwang, 2014), important messages (Wan and Pfau, 2004), and communication platforms (De Bruycker and Walgrave, 2014;Göritz, et al., 2011;Taylor and Perry, 2005) is all part of an effective crisis response (Johansson and Härenstam, 2013;Lee, 2004).
In an investigation of 133 crises, Diers (2017) revealed a range of different strategy sets derived from the taxonomy of crisis response techniques, which provides a fair grasp of how crisis tactics might be strategically integrated. The author found a future-oriented strategy, for example, that centred on an organization's desire to look beyond the current crises to a "better future" for the company and its stakeholders. Self-improvement, excellence/renewal, and interorganizational relationships were among the method categories, and this strategy was most commonly adopted by crisis-prone organizations.
The primary approaches employed in crisis communication research have evolved throughout time, reflecting the history of the field. In the field, there are six basic broad methodology types (Dickmann et al., 2015). Conceptual studies, which include best practices, are the first. While they make up over a third of all journal articles, their earliest average publication date was in 2002. With an average publication date of 2005, the field shifted its focus on crisis communication to rhetorical analyses (Koesten and Rowland, 2004). Qualitative analyses, the smallest methodological approach, increased the scope of study and methodology rather quickly. Then, starting in 2008, quantitative analyses (including surveys and content analyses) became more widely used, with 24 percent of all research using this method. Finally, with an average publication date of 2009 and roughly 9% of all journal publications, experimental approaches began to be used more frequently in crisis communication (Pearson and Clair, 1998). This shows that, as the discipline matures, research focus and methods are still developing.
It is a field that is anchored in communication traditions but spans over a wide range of disciplines, topics of interest, research approaches, and ontological perspectives. This is demonstrated throughout the five important variables mentioned here, which argue that understanding crisis communication entails comprehending how issues and reputation management, crisis type, organizations, stakeholders, and crisis response messages intersect. From the initial research of crisis communication in the 1950s through the field's rapid growth in the 1990s and beyond, it is certain that we now have a far better grasp of the subject.
A typology based on communication purposes and orientations has been established in order to better conceptualize public crisis communication (Olsson, 2014). Public organizations might participate in four types of crisis communication, according to the typology: reputation, resilience, strategic and operational. There are two types of communication in the first dimension: operational and strategic. Operational crisis communication has historically focused on concerns such as disseminating relevant information about a crisis to those who are most exposed so that informed decisions may be made on crucial topics (Marra, 1998). Rather than organizational survival and reputational issues, operational information focuses on comprehensive information, availability, and emotional reactions (Cho and Gower, 2006). Strategic information, on the other hand, is pre-planned and geared at attaining long-term organizational objectives. It is worth noting that strategic communication can be interpreted in a variety of ways. As a result, for the purposes of this article, strategic communication is defined as a managerial function focused on ensuring organizational longevity and/or achieving positive stakeholder views (Jaques, 2009;Sellnow, 1998).
The second component is concerned with two communication goals: preserving reputation and developing resilience. The first goal is self-centred and referential, with the goal of explaining and promoting the organization's own framing and preferences (Heinze et al., 2014;Ott and Theunissen, 2015;Turk et al., 2012). The second goal is resilience, which focuses on delivering crucial information to help communities and individuals survive and recover in case of a disaster. There is a lot of literature on resilience and, as a result, there are a lot of definitions. At a broad level, resilience refers to a community's ability to 'bounce back' after a crisis, that is, the ability to adapt and change to new surroundings while maintaining longterm stability (Glik, 2007;Holmes et al., 2009). According to Gesser-Edelsburg et al.(2014), resilience is most commonly thought of as the ability to adapt to adversity. Furthermore, rather than being viewed as a result, resilience should be viewed as a process of adapting.
Together, the four dimensions produce the basis for crisis communication strategies, which will be scrutinized in the following sections by using content analysis applied to the addresses by the President of European Commission during the first year of the COVID-19 epidemic.

What means to communicate uncertainty during a crisis?
Uncertainty connected to a public health emergency incident is measured both in terms of information delivered and in terms of uncertainty experienced by the public. The uncertainty information can come directly (or indirectly) in the form of numerical probabilities ("60 percent chance"), linguistic likelihoods ("strong chance"), and absence ("no information"), or it can come indirectly in the form incomplete, inconsistent, and conflicting/contradictory information, all of which can lead to a cognitive and affective sense of doubt (Sellnow and Ulmer, 1995). Furthermore, there are other sources of doubt, such as uncertainty about personal and others' safety, event knowledge, influx of nonlocal people, and the future of the village/ town, which alter through the course of an event's stages.
Because the public is uncertain during a public health emergency due to a lack of information, it actively seeks general and specific information from traditional mass media, social media, and interpersonal network sources (family, doctors, nurses, and community leaders) to alleviate this uncertainty (van Zoonen and van der Meer, 2015). Other predictors of public uncertainty include a lack of: local coordination and cooperation; a disaster management plan that is (not) engaging with the public; effective management of the official response to an event across all mass and social media; and fast, regular, timely, reliable, and detailed information about event progression (Glik, 2007). Personality traits, such as risk perceptions, are also linked to the sense of uncertainty.
The production of messages carrying ambiguity information on public health emergency occurrences for vulnerable communities must take into account the intended audience's entire living context. Poverty can create a lot of uncertainty in people's life, and not just because of a specific hazard or incident. It should be acknowledged that issues of economic development and the environment are just as important as official messaging in eliminating doubt about an incident (Sopory et al., 2019).
The media is an important stakeholder, and if their coverage of a public health emergency event emphasizes rapidly changing, contradictory, and conflicting information, particularly when it differs from official information from authorities, public uncertainty can be increased, which can lead to a variety of negative outcomes, such as loss of faith in authorities and suggested actions; ambiguity and anxiety; decreased intentions for health-protective behaviours; and decreased attention to health-risk news (An and Gower, 2009;Sopory et al., 2019). Other stakeholders, such as frontline health/medical personnel, are similarly limited in their capacity to address the public's concerns about how to respond effectively to the event as a result of such media coverage.
When official information about the crisis is missing or contradictory/inconsistent, the decision-making related to the event becomes ambiguous. Furthermore, when authorities hurry to announce a "fact" about an emerging situation without explicitly addressing uncertainty, it might undermine important stakeholders' decision-making.

Methodology and exploratory research questions
As previously stated, the aim of this study is to investigate the crisis communication narratives and how the European authorities approached the task of communicating uncertainty during the Covid-19 pandemic. Using the critical discourse analysis approach, the study explores how the legitimization of the EU's response was discursively built and if there was a distinction between uncertainty information and uncertainty experience in terms of crisis communication.
Critical discourse analysis (CDA) is a qualitative method for describing, analysing, and explaining how discourses build, perpetuate, and legitimate social realities. CDA is based on the idea that our language choices, whether conscious or unconscious, are intentional. As Fairclough convincingly synthesizes, the studies based on CDA have a largely agreed-upon agenda. According to the author, CDA means to systematically explore "often opaque relationships of causality and determination between (a) discursive practices, events and texts, and (b) wider social and cultural structures, relations and processes; to investigate how such practices, events and texts arise out of and are ideologically shaped by relations of power and struggles over power" (Fairclough, 1995, p. 132). In this article, we seek to establish determinative or causal relationships between the legitimacy of the European Commission's response to the COVID-19 crisis situation and the way in which official justifications and explanations were discursively constructed. The need to relate wider social and cultural structures to discursive practices has led us to choose critical discourse analysis (CDA) as a research methodology.
The research steps followed for the CDA approach were: we clarified the context and we explored the production process of the speeches, then, the text was coded by taking into account the variables inferred from the research questions, the structure of the texts was examined; the determination relations were revealed and analysed together with identification of cultural references.
The official European Commission's communication takes place on several channels and there are different official communicators, but the President of the European Commission is the one who sets the tone for the institutional executive official communication and sets out, through his speeches, the central messages to be assumed by the EC. That is why this research focuses on the content analysis of the speeches by the President of the European Commission, Ursula von der Leyen. Therefore, 5 speeches delivered in the plenary sessions of the European Parliament and 17 opening statements from press conferences held by the President of the European Commission (between March 13, 2020 and March 17, 2020) were selected as a corpus for this research. We chose 1 or 2 public speeches each month, delivered at times communicatively significant (such as the European Parliament's plenary assemblies) in order to observe the dynamics and consistency of the issues addressed, at the highest level, from month to month in the EU's official response to the crisis.
We summarize below the corpus of the analysis. The speeches are available on the official website of the European Commission, under the heading Press Corner and on the official You Tube channel of the EC. For data interpretation, the study used Atlas.ti 8. This is a soft which helps the researcher manage and arrange the layers of analysis and identifies links among qualitative data acquired through the gathering of texts. For a better understanding of the relationships between the numerous ideas addressed in the research questions, the pieces of data interpreted both inductively and deductively were coded. While a deductive method focuses on validating a theory, an inductive method focuses on the production of a new theory from data, in this instance, a theoretical model of interaction between uncertainty information, uncertainty experience and the legitimization of the crisis response. There are a few approaches for organizing and interpreting qualitative data, but they can all be simplified down to the following fundamental techniques (Ryan and Bernard, 2000): linguistic method, which considers texts as objects of analysis in and of themselves, and the sociological approach, which considers text as a window into human views and experiences. The sociological dimension is the emphasis of this research. According to Ryan and Barnard (2000), there are two types of written texts: (a) words or phrases retrieved by systematic elicitation procedures, and (b) free-flowing texts such as narratives or replies to open-ended interview questions. As a result, this research study is concerned with the first part, and the method used for analyzing the data gathered from the field research (speeches transcriptions) is keywords in context (KWIC). According to Ryan and Barnard (2000), this methodology locates all instances of a given word or phrase in a document and highlights it in the context of a number of terms preceding and following it. This method is based on tagging or labeling qualitative data. The descriptive or inferential data gathered during the study is given units of meaning by the researchers. As a result, codes were assigned to varyingsized 'chunks' of text in order to link or disconnect keywords or phrases inside certain research questions.
The table below summarizes the codes used for content analysis.

Findings and discussion
The research design for this study did not entail testing theory, but rather building theory from evidence (via exploratory research). The CDA approach was used to gain insight and establish an understanding of the relationship between communicating uncertainty and the legitimization of the EU's crisis response. The study tried to establish a connection between the two and to elicit a strategy grounded in the experience of the COVID-19 crisis. Working within such a structure allowed considerably more nuanced perceptions and subjective relationships to be studied, assuming that reality is socially produced rather than objectively established (which is the CDA's perspective). As a result, this study was an opportunity to analyse, comprehend, and explain the EU's crisis communication approach and the legitimization goal of the European Commission with the help of the research questions.

RQ1: Has a clear distinction been made between uncertainty information and uncertainty experience in terms of crisis communication?
The studies focused on crisis communication in uncertain contexts distinguish between uncertainty information and uncertainty experience. Through RQ1 we wanted to see how things are in practice, at the European level of institutional communication in crisis situations. When we refer to uncertainty information, we pay attention to the communication of the fact that no information is known about a certain situation, that the data are uncertain and that, in general, the situation is not understood. When we talk about uncertainty experience, we refer in terms of communication to the way people experience / feel a certain situation, if they feel insecure about the fact that a situation is unpredictable. Depending on how the authorities approach the two types of uncertainties in public communication during crises, the effects on the audience will be different: either the authorities/institutions will seem to have control and manage the situation well even if it is unpredictable, or the actions of the authorities will seem hesitant, lacking coherence and control of the situation. As a result, the institutions will have problems of legitimacy regarding the response to the crisis situation and will thus increase the state of uncertainty experienced by the population.
The content analysis of the speeches and opening statements given by the President of the European Commission shows us that, in the practice of crisis institutional communication, uncertainty information is associated with uncertainty experience, that, basically, they are mixed. This makes it difficult to determine to what extent the uncertainty experience is augmented by uncertainty information.

Figure 2. Uncertainty information vs. Uncertainty experience
Source: Author's representation based on data processed with Atlas.ti.8 Figure 2 displays information about the connection, groundedness (G) and density (D). Groundedness refers to the number of linked quotations, while density counts the number of linked codes. The higher the G-count for a node, the more grounded it is in the date. For the "Uncertainty information" code we have G-143 and for the "Uncertainty experience" code there is G-102. There has been a higher number of references to unknown information, to the unpredictable framework of the crisis that emphasized the uncertain information dimension. The difference between the two indicators analysed in relation to the uncertainty (information vs. experience) is large (N: -41), which means that it is possible that the authorities often seemed that they did not know how to act, that they did not have a management plan for the crisis. Often formulating phrases that contain the linguistic marks of lack of information (we are still waiting, we do not know, no news, no information, no clear answer) leaves the impression that the authorities have no plan, therefore the legitimacy of the response to the crisis situation will decrease in the perception of the audience. On the other hand, the recognition of the state of uncertainty that people feel (hence, the uncertainty experience) through linguistic marks (no safe conditions, critical situation, danger, threats, unsafe) conveys the message that people's emotional state is known, their situation is also known and, as a consequence, people feel heard and seen, and the response to the crisis will benefit from more legitimacy.

RQ2: What was the strategy that has been emphasized to legitimize the actions initiated by the European Commission?
In order to ensure the legitimacy of some institutional responses, in crisis situations, various justifications will be used and certain motivations for action will be emphasized. The legitimization tactics can be employed, alone or in combination with others, to justify social behaviours by: (1) emotions (particularly fear), (2) a hypothetical future, (3) rationality, (4) voices of expertise and (5) altruism. Through RQ2, we try to identify what was the discursive strategy of legitimizing the European Commission's response during the COVID-19 crisis, and on which component this emphasis was placed.

Figure 3. The connections between the legitimization narratives
Source: Author's representation based on data processed with Atlas.ti.8 The research found that the highest density (D-3) of association is on the subcodes "rationality", "voices of expertise" and "hypothetical future". These data show that the narrative built to legitimize the European Commission's response to the COVID-19 crisis was based on rational arguments and a call for a desired future. Altruism and emotions have not been consistently linked to other rational arguments. This narrative approach to legitimacy keeps the European Commission in the sphere of institutions with a less empathic and a more technical approach, which is often invoked in studies analysing the causes of EU-scepticism and the feeling of distance between European bureaucratic institutions and EU citizens. Connecting emotions with other categories of arguments would increase the legitimacy of the crisis response towards public perception.  The analysis shows us constant connections between the operational and strategic dimensions of crisis communication. In other words, the European Commission tried to consolidate its position on the international political scene and, implicitly, its legitimacy through the discourse dominated by the technical approach of the COVID-19 crisis, trying to communicate an operational crisis management capacity superior to other international political actors. The legitimating narrative was dominated by operational communication which sought, on the one hand, to strengthen the strategic position of the EU (strategic dimension of communication), on the other hand, to further determine a capacity and communication for resilience. The extent to which this narrative is appropriate depends on reputation communication, i.e. the already consolidated reputation of the EU that can be invoked in such moments of crisis.

Conclusions and practical implications for European institutional communication
In the messages issued by authorities, the public should be informed explicitly about the uncertainty connected with public health emergency situations. We argue that disseminating scientific uncertainty to a public that is unable to conceptualize uncertainty in scientific terms can harm the public's trust in science and scientific institutions, and can lead to panic and confusion about the scope and impact of a particular event, experts and researchers generally agree on this. Focusing on the uncertainty experience, not on uncertainty information, could strengthen trust in the authorities and make uncertainty information have fewer negative effects. Therefore, a first conclusion is that crisis communication must make a distinction between the two components -uncertainty experience and uncertainty information. In the case of approaching the European Commission's crisis communication in the COVID-19 pandemic, the content analysis revealed confusion, an overlap between the two components and the lack of distinct strategies to address the uncertainty. From here, a fertile ground for misinformation, conspiracy theories and other fake news emerged.
Uncertainty information in messages provided by authorities during public health emergency events is consistently associated with desirable outcomes such as reduced uncertainty about health-protection actions, reduced reliance on misinformation, rumours, and sensationalized media stories, and improved response to future warnings for the general public. When openly addressing uncertainty, however, the risk of certain unfavourable results for particular vulnerable communities, such as a loss of trust in authorities and hesitancy regarding evacuation, must be kept in mind.
Another finding of the study supports the European Commission's emphasis on legitimacy in terms of rationality and expertise, and less on emotions and altruism. Of course, argumentation based on concrete data, well-founded argument and scientific expertise are extremely important. Often, this type of discursive legitimization is not appealing to the public. Crisis situations arouse many emotions, especially fear, and then, the legitimization of the response must take elements from the same emotional register belonging to the crisis, for people to feel understood and, therefore, consider that the response to the crisis takes into account their real situation.
There is life after every crisis and this is one reason why crisis communication must, in fact, have in mind the strategic dimension and, implicitly, the one that affects the reputation. During the crisis, reputation is a two-sided currency: on the one hand, a good reputation is used to legitimize the response to that crisis and is a very helpful capital of trust, on the other hand, the reputation of an organization/institution is an element at stake in a crisis so communication must be meant to protect the institutional reputation. The content analysis carried out in this study shows that the European Commission has paid attention, in all moments of communication, to the strategic and reputational dimensions of crisis communication.
During this research, we tried to analyse from several perspectives the way in which the legitimization of the response to the COVID-19 crisis was constructed. However, one of the important limitations of the study is that we did not measure people's perception, the way they evaluate the actions of the European Commission from March 2020 to March 2021. The latest Eurobarometer shows an increased trust in EU institutions, which is an indicator of positive assessments of crisis communication, but more empirical studies are needed to confirm this. Our research could also contribute to future quantitative and qualitative empirical studies that would analyse the impact of COVID-19 crisis management on the reputation and image of the European Commission and the European Union, at large.
We conclude by outlining the need to further study crisis communication in the case of institutions (Grunhut, 2017;Vrnakova, 2018). At the moment, most of the crisis communication literature focuses on business or non-government, but the public sector has to deal with large-scale crises (COVID-19 is the most eloquent example, but there are many other examples) with consequences for broad categories of the population, and research could reveal useful strategies, techniques and approaches for public institutions and political leaders.