Abstract
News agencies are regularly mentioned in scholarly works among the principal sources of information for legacy news organisations and titles, particularly in the 24/7 online news environment. However, little is known about how these agencies themselves source news. To fill this gap in scholarship, we present a case study of how Belgium’s national, multilingual news agency sources its science news. We first position our study within a conceptual framework combining insights from news diversity and translation studies research. Next, we operationalise a triangulation of a content analysis, a survey, and interviews with newsroom staff to shed light on the practices and policies that shape and constrain the inner workings of a national news agency in a small, multilingual media market. Our findings reveal a large, potentially worrying, dependency on information subsidies and their assumed factuality from press releases and other news agencies, largely sourced locally. We contextualise our findings within wider news agency and science journalism research.
Keywords
Introduction
The hegemony and obscurity of news agencies
Scholarship has highlighted the “key but often hidden” role (Bielsa, 2008: 349) news agencies play in providing newsrooms and citizens with factual information that is immediately publishable. Scientific studies on news agencies were popularised by a 1953 UNESCO report on their role in shaping public opinions and transmitting news (Surm, 2020). In the 1970s and 1980s, attention turned to their role within media ecosystems. This attention was frequently geared towards international, English-language news agencies such as the American Associated Press and British Reuters, considered “wholesale news producers” (Bielsa, 2008: 363; Rantanen, 2019).
Although their histories, ownership, and governance vary, all news agencies share similar threats and risks. Since news organisations are typically the main funders of news agencies, the gradual demise of the former negatively reflects on the latter. In the current digitally-oriented media market, they can be perceived as a supplier, customer or competitor. Their economic pressure is intensifying, also because they depend on government funding (Bielsa, 2008; Boyd-Barrett and Rantanen, 2000; Surm, 2020). Another similarity across the diverse body of scholarship is the confirmation that news agency copy constitutes a key information source for legacy newsrooms. While journalists themselves consider news agency content as the foundation for more in-depth reporting (Van Leuven et al., 2013), systematic studies have indicated that sizable portions of this content appear verbatim, predominantly on news websites (Boumans et al., 2018; Johnston and Forde, 2011; Lewis et al., 2008; Welbers et al., 2018). This shared consensus makes it only more remarkable that news agencies’ own sources of information have not been the topic of much scholarly scrutiny yet. This study aims to investigate the inner workings of a national news agency and its sourcing practices, using a triangulation of research methods and a theoretical framework that incorporates news diversity and journalistic translation studies, in response to Rantanen’s (2019) call for more empirical and conceptual research on news agencies' proceedings.
Scholarship generally distinguishes national and international or global news agencies (Bielsa, 2008; Rantanen, 2019; Surm, 2020), with the latter frequently operating in multiple languages and countries. The former have been defined as “government agencies run by or subsidised from the ministry of culture or information, perhaps with management participation from client media” (Boyd-Barrett, 1980: 80). This definition still applies to Belga News Agency, the national news agency in Belgium and hereafter shortened to Belga, as it is known colloquially. It too “cater[s] mostly to the activities of the central government and issues perceived to be of significance” (Al-Rawi and Iskandar, 2022: 1159), in itself a complicated task due to Belgium’s complex political structure with various governments across linguistic borders.
In our study, we focus on the sources that Dutch-language Belga journalists use to produce science news. With the current journalistic business model “under intense financial pressure to trim expenditure on specialist, investigative reporting, it is all too often the case that science news is regarded as expendable” (Allan, 2011: 773). Linked to this is the simultaneous evolution of universities’ expansions of public relations departments as science outreach portals to news media, which suffer from fewer available resources (Vogler and Schäfer, 2020). With previous case studies indicating that even quality news titles frequently quote scientific publications inaccurately (Dempster et al., 2022; Vestergård, 2011), it begs the question of how the COVID-19 pandemic has affected the importance of science reporting, particularly since during this period, journalists acknowledged having an “innately problematic” relationship with information (Perreault and Perreault, 2021: 974). The relationship between news agencies and science news has received very little academic attention thus far, strengthening our contribution to the state-of-the-art. A rare analysis of COVID-19 reporting is limited to tweets sent out by three global news agencies (Wang and Lu, 2022). The finding that Belgian news outlets regularly use secondary sources such as news agency copy in their reporting on COVID-19 vaccine studies by Nelissen and McMartin (2022) serves as the foundation for this study and the continuation of its conceptual framework.
Conceptual framework
News diversity
News agencies have been poorly theorised in existing scholarship (Rantanen, 2019), with their presence typically taken for granted by media policymakers and scholars. The most frequently used conceptual frameworks revolve around media dependency or flows, hybridity or concepts such as Bourdieu’s symbolic power (Boyd-Barrett and Rantanen, 2000; Rantanen, 2021; Tworek, 2020). We contribute to this ongoing debate by positioning news agencies and their unique position in media ecosystems within a wider framework of news diversity. This concept too has been frequently criticised for its inadequate theorisation and operationalisation, particularly because it tends to be used somewhat interchangeably with similar yet distinct concepts such as media diversity and pluralism (Iosifides, 1999; Loecherbach et al., 2020; Sjøvaag, 2016). Notwithstanding this, there have been recent attempts to devise integrated frameworks for news diversity as a multi-faceted concept (Hendrickx et al., 2022; Joris et al., 2020). We operationalise its different facets to fit the scope and added value of news agencies in contemporary news markets and understand the concept of diversity in a news context driven by democratic theory. Put succinctly, this states that a diverse range of sources is required in news reporting to offer people sufficient viewpoints on topical matters which in turn enables them to become agentic citizens able to make decisions in pluralistic, democratic societies, which includes taking part in public debate and voting in fair and free elections (Champion, 2015; George, 2013; Karppinen, 2007; Sjøvaag, 2010).
Following Hendrickx et al. (2022), we consider news diversity as an umbrella term comprising five subdimensions (ownership, brand, production, content and consumption diversity) that have the ability to (organically) influence each other. When applying this typology to the specificities of national news agencies, particularly those in smaller media markets where they are more closely tied to local cultures and languages (Bielsa, 2008; Rantanen, 2019), we already unravel a number of their key characteristics:
Ownership diversity
Rantanen (2021) discusses the diverse ownership structures of European news agencies, ranging from private shareholder companies to various types of state ownership. The author notes that “different ownership forms of national news agencies in Europe may have been partly amalgamated through their owners and clients” (p. 276).
Brand diversity
Although it has to our knowledge never been quantified as news agencies are frequently omitted from studies into news brands and their trust among societies, we hold the assumption that news agencies generally enjoy high levels of trust within newsrooms, where their content is repurposed and used as a first-hand news source on a daily basis.
Production diversity
Very little is known of the contemporary workload and overall environment of journalists working for news agencies as newsroom studies typically rather focus on legacy or, increasingly, digital newsrooms. This is unfortunate, as news agencies are often perceived as first news sources, which would mean that they are under additional pressure to produce reporting that is both factually correct and extremely fast. As an addition to the available literature, we shed light on how science news is sourced and produced within Belgium’s national news agency and how this fits within its wider company structures. This is approached from the understanding that working conditions for journalists are influenced by changes to ownership structures (Hendrickx and Ranaivoson, 2021).
Content diversity
Journalistic production logically results in content, in this case, news agency copy that is shared with subscribing newsrooms automatically through a content-sharing system. News outlets have the liberty to use this copy directly or as the foundation for their own coverage. As argued before, news agency copy is increasingly repurposed in online newsroom settings, in many cases (almost) verbatim (Boumans et al., 2018; Lewis et al., 2008; Welbers et al., 2018).
Consumption diversity
When news agency copy is published (online) without alterations, this can dramatically alter the diversity of news content and the information flow for citizens, as this means that only the viewpoints reflected in the original news agency article will reach them (Champion, 2015). This potential lack of diverse information can then reflect negatively on the democratic function of the press in society, harming not only news diversity but, ultimately, also democracy.
Peculiarly enough, the two distinct scholarly debates on diversity within a news media context and on news agencies have rarely been linked. One exception is the study of Welbers et al. (2018) who use computational text analysis to see the evolution of the influence of news agency copy on political reporting in the Dutch press between 1996 and 2013. The authors find that the Algemeen Nederlands Persbureau (ANP), a Dutch national news agency, often influenced printed news, but also that “[o]nline newspapers, however, were often influenced by the same articles, which in combination with their strong dependence on ANP copy substantially harms the diversity of their political news coverage” (p. 329). A few other scholars have also discussed how news agency copy may be detrimental to news (content) diversity (Boumans et al., 2018; Johnston and Forde, 2011), but to the best of our knowledge, this has not led to studies that investigate the sources news agencies themselves use to produce the copy that is subsequently repurposed by legacy news brands and accepted as coming from a generally authoritative, credible news source.
Journalistic translation
We strengthen our contribution to these areas of study by integrating elements of translation studies. Translation scholars have generally accepted Jakobson’s (1959) three-part classification of translation: interlingual (from one language to another), intralingual (within one language, e.g. rewording a text for a different audience), and intersemiotic (from one sign system to another, e.g. turning textual data into an infographic). Translation in all its forms is pervasive in journalism, since news often travels across languages and cultures. This is especially pertinent in international and/or multilingual news agencies, where journalists consider themselves as “cultural go-betweens”, sourcing information from one country, region, or culture, and introducing it into another (Davier, 2015: 547–548). Moreover, intralingual translation is specifically relevant in the widespread dissemination of scientific findings, since the information must be adapted from specialised, academic language to accessible writing that a lay audience can understand.
Translation is largely ignored and/or misunderstood in journalism studies (Valdeón, 2018). As Valdeón (2020) writes, “[translation] scholars have moved from an essentialist view that subscribes to the notion of fidelity to a more encompassing and flexible definition that looks at rewriting, adaptation and appropriation as realizations of translation” (p. 1655), and, he argues, journalism scholars have not followed suit, nor have journalists themselves (Bielsa and Bassnett, 2009). Far from a straightforward lexical transfer, translation is a complex process that is closely linked to a number of concepts rooted in journalism studies, with framing (e.g. Valdeón, 2015) and gatekeeping (e.g. Valdeón, 2021) as the best-described examples. As a result, Valdeón has called for closer collaboration between the – relatively recent – subfield of journalistic translation studies and journalism studies, most recently in a special issue of Journalism (Valdeón, 2022).
Valdeón (2021) distinguishes two functions of translation in journalistic production: “on a macro-level [it] functions as a filter to allow news writers/translators (and the companies they work for) to consider what should be published and what should not, while on a micro-level it serves to select the parts of the original articles that need to be adapted, added or omitted during the translational and editing processes” (p. 32). The macro level is closely linked to news sourcing: information can only be reported on if it is available in a language the reporter understands. This has immediate implications for news diversity, though news diversity research has not previously considered linguistic diversity (nor have journalistic translation scholars considered news diversity). For example, in a study of Belgian newspaper coverage, Van Doorslaer (2009) found that the countries dealt with in international news greatly differed between Belgium's Dutch and French-language regions, Flanders and Wallonia. This was linked to the reporters’ use of news agency copy: French outlets predominantly used Agence France-Presse, while Flemish reporters most often relied on Associated Press. As a result, the languages and parts of the world one sources their news from reflects a certain worldview.
At the textual, micro level, journalists typically modify information from one or more sources to accommodate the target audience, often through a combination of (re)writing and translation, a process sometimes dubbed ‘transediting’ (Stetting, 1989). This process is influenced by the personal, organisational, and societal context journalists operate in (Nelissen and McMartin, 2022). Often, alterations or even distortions occur when textual material in one language is used to produce news in another language (Bielsa and Bassnett, 2009; Scammell, 2018). This can also happen when information is translated intralingually from jargon to layman’s terms. Although not conceptualised as journalistic translation research, several scholars have found that scientific information is often hyped or misrepresented when it is communicated to the general public (Dempster et al., 2022; Sumner et al., 2014; Verstappen et al., 2022).
By looking at sourcing practices, this study focuses on the macro-level function of translation in journalism. We hope that investigating not only source types but also their languages and countries of origin will enrich current theories on news diversity and raise awareness of the role of translation in newsmaking processes. Building on van Doorslaer’s research (2009) and that of Nelissen and McMartin (2022) into the sourcing strategies of Flemish news media, we go back one step in the – often complex – science news flow to look at which kinds of sources Belga reporters use and the languages and countries they originate from.
Methodology
This study is based on a triangulation of research methods, combining a quantitative content analysis of 364 articles, 119 surveys completed by 29 reporters, twenty semi-structured interviews with journalists and two in-depth, semi-structured interviews with the Deputy Editor-in-Chief in charge of the Dutch-language department of Belga’s newsroom. All Dutch-language science news from Belga published between 1 March and 1 June 2022 was collected and analysed. Our definition of science news builds on that of Bucchi and Mazzolini (2003), who delineate it as:
“[A]rticles explicitly reporting research findings or events related to the natural sciences or to applied sciences such as engineering and medicine, articles featuring statements by scientific experts and articles including references to science or using science-based argumentation.” (p. 8)
This broad definition goes beyond the rather narrow view of science news as reports of research results, though it is narrow in its topical demarcation since it does not include the humanities. We therefore extend the definition to all academic disciplines, following, among others, Summ and Volpers (2016), Hijmans et al. (2003), and Schäfer (2012).
New Belga articles were skimmed manually every day via Belga’s platform Belgabox. They were added to the corpus if they included at least one paragraph that focused on one of the elements from the definition above. Public health updates (such as reports on the number of COVID-19 infections) were only added if they contained a scientific argumentation or referred to a study. We also decided to include articles that were seemingly based on science, such as the results of surveys conducted by commercial enterprises, since these results were presented as facts and at least some of these studies appear to be conducted in collaboration with academic researchers. As a result, we argue that such content might be understood as being evidence-based. Out of a total of 7,445 Dutch articles, we identified 364 articles that matched our criteria, a share of 5% (rounded off to the nearest integral). Using Python, the articles were added to a relational database alongside relevant metadata, including the title, the text of the first paragraph, the tags the reporter assigned to the article, the team it originated from, and, if relevant, the name of the news agency the article was sourced from. Additional data was added manually, such as the type of source text the reporter based themself on, which we categorised based on our online searches of said texts as well as the interview data (on which we elaborate below): 1. Press releases 2. Reports by press agencies 3. Press conferences and events 4. Reports from Belga’s French department 5. News articles 6. Reports 7. Social media posts
We also noted the original source of the news (typically an organisation), which we classified inductively as follows: 1. Universities and other research institutes 2. Governments, government agencies, and political parties 3. Interest groups, such as animal rights groups or labour unions 4. Companies 5. NGOs 6. News outlets and news agencies 7. Social media
To gather in-depth data about how specific articles were sourced, written, and/or translated, reporters were invited via email to participate in a survey about their science news articles shortly after publication. In the survey, they could note their availability for an interview. This yielded 121 survey responses, and subsequent phone interviews about 20 articles. It was agreed with the Deputy Editor-in-Chief that these interviews were to be kept as brief as possible. Even so, many reporters indicated in the survey that they were not available for an interview. To contextualise our findings and better understand the overall workings of the Belga newsroom, we also conducted two in-depth interviews with the Deputy Editor-in-Chief.
Finally, further information about the sources the reporters used was retrieved by asking, along with the survey invitation, which specific texts or other materials the reporters had used to write their articles. Since this only yielded limited input and Belga quite clearly mentions sources in its articles, additional online searches via different search engines, as well as on organisations’ websites, were performed to match Belga articles to their source. This mixed-methods study design culminates in our guiding research questions below. The first research question (RQ1) is based on the survey and interview data, while the second (RQ2) is based on the content analysis. RQ1 How is science news writing incorporated into the daily operations of Belga? RQ2 How does Belga source its science news? RQ2.1 Which types of information subsidies are used? RQ2.2 From geographies and in which languages does its news originate? RQ2.3 What kind of organisations does the news come from?
Findings
The history and structure of Belga
Since we take an explicitly situated perspective by studying the practices of one national news agency operating in a specific, multilingual context, and since cultural and organisational factors shape newsmaking processes (Reese and Shoemaker, 2016), we start our findings with a brief explanation on the corporate structure of Belga. Belga was founded as a French-language news agency in Brussels in 1920 and started issuing reports in Dutch in 1944. In 1970, the company reorganised to create equally sized Dutch and French-language departments producing separate news content from the same newsroom. Its customers include media outlets (which account for 63% of Belga’s income from news subscriptions, according to data provided by a senior staff member), government organisations (16%), companies (17%), and other news agencies (5%). Belga shares news through text, video, photographs or audio - an average of 147 reports per day in 2021, according to its archive. Though there are some specialised news agencies in Belgium, Belga has a de facto monopoly as a news agency catering to Belgian organisations.
Belga’s newsroom employs around 100 people and is headed by the Editor-in-Chief, who oversees parallel Dutch and French teams, with each role existing on each side. A Deputy Editor-in-Chief, supported by three news managers, is at the helm of each language group. News production is structured around four thematic ‘desks’ or beats: national, international, political, and economic news. These are managed by a so-called ‘deskeur’ and a ‘co-deskeur’, or (co-)desk managers, roles that are assigned to reporters on a rotating basis. Together with the news manager, these people decide what should be covered and assign writers to news stories. The reporters in the politics and sports desks exclusively work on said topics; the others are part of what is called the ‘newspool’ and can be assigned to each of the three remaining desks, depending on availability and interests. There is also a photography unit, a videographer, and a so-called ‘info office’ team, which manages the main email inbox and compiles Belga’s calendar of events. The structure of Belga’s newsroom is visualised in Figure 1 below. Belga’s newsroom structure.
Next to the in-house reporters, Belga works with freelancers who act as regional correspondents (three for each of the 10 Belgian provinces). They are typically the ones who go on location, for example to press conferences. Most of them have Belga as their main or only client. Finally, Belga often works with freelance photographers.
Daily operations
Though they share the same open-plan space when working from the office in Brussels, the two language teams work independently. This means that the Dutch-language team treats news from the French-speaking team the same as other incoming news: the desk manager sees it in the feed and decides whether it is relevant for Belga’s Dutch-speaking audience. If so, a reporter is assigned to translate it and the by-line remains that of the original author. This differs from translations of international news. Most of Belga’s international news is sourced from the four international agencies Belga subscribes to: AFP, Deutsche Presse-Agentur (DPA), EFE, and Bloomberg, which operate in the language of their country of origin, respectively French, German, Spanish, and English. The first three also offer news in English as well as other languages. Belga also subscribes to three national news agencies, with the Dutch ANP as the most important one as it is regularly used as a news source for Belga reporting. The Russian TASS and Japanese KYODO are available, though rarely used. The Deputy Editor-in-Chief stressed that articles based on content from other news agencies are ‘journalistic translations’, in contrast to literal translations, since information can be omitted or added depending on the new audience. As a result, the journalist-translator is credited rather than the original author, though the source agency is mentioned above the article.
The linguistic diversity of the sources used at Belga is reflected in the recruitment process. Each reporter is required be able to translate from the other main national language, as well as from English and German. This is tested when candidates are interviewed. Nevertheless, several reporters indicated that they did not speak German well but relied on machine translation and the bilingual online dictionaries provided by Belga. This practice is also common for other languages, since, for example, there is not always someone on duty who speaks Spanish. Additionally, if (international) news is important enough, one interviewee noted, it will most likely also be covered in English, French, or German, so the reporters use those sources instead.
Before diving into Belga’s science news, some general newsroom guidelines should be highlighted. First, given the high demand for topical news, the target length for a Belga article is 300 words. If a story needs more details, or if it develops over time, Belga will send out a first, short, version, called a ‘brief’, and issue updates with additional information. Several reporters noted that the norm is to finish an article in about 45 min. This is reflected in our survey: 37% (total n = 90) of reporters said they spent less than 30 min on an article, 43% spent 30 min to an hour, 17% one to 2 h and 3% more than 2 h. Stories are written in Superdesk, a software programme that allows quick sharing, editing, and publication. Since Belga does not employ line editors, reporters revise each other’s articles.
As mentioned earlier, the desk managers and news managers decide what will be covered. News typically comes in via email in the form of press releases and other information subsidies, on which Belga is heavily dependent. Freelancers can pitch their own stories, but, at least for science news, these are again mostly based on press releases or briefings. Indeed, as the next section will argue in more detail, this workflow seems to work very well for science news, an ecosystem in which press releases by academic institutions are abundantly available (Autzen, 2018; Comfort et al., 2022; Vogler and Schäfer, 2020).
Science news sources
Depending on the angle, science news can fall under any of the four thematic desks (national-general, national-economy, national-politics, international). Apart from the politics and sports staff, Belga has no beat writers on board, though some reporters specialise on the job. This has not really been the case for science, save for one senior reporter with a keen interest in astronomy whose colleagues consider him an expert on the matter. Overall, most Belgian (Flemish) science news comes from local (freelance) correspondents near universities, with several of them participating in this study. Next to research, they cover a wide variety of other, local topics. When asked what would need to happen for Belga to invest more in science writing, the Deputy Editor-in-Chief mentioned that they needed more budget, immediately adding that there instead had been budget cuts. If the budget were to be significantly increased, they would (in order of priority) raise the freelance rates, invest in quality control by hiring line editors, and invest in specialised reporters, especially people with a law or science background. Overall, we deduct from this and from the fact that only a small percentage of news at Belga discusses science that science news has a relatively low priority for the company.
Sources by language and country
The survey reveals that two-thirds (66%) of respondents used source material in Dutch only, meaning that their articles did not involve any interlingual translation, and one-third used source material from other languages. A further 20% also used English-language sources. The remaining 14% (also) used sources in French and/or German. This suggests that either there is a preference for news in the native tongue of the reporter, and/or a strong focus on news about Belgium (in Dutch), and Flanders specifically (as well as the Netherlands). Indeed, only six articles (out of 260) produced by the national news team were translated reports written by the French-speaking colleagues at Belga, and of the 75 articles about Belgian universities, merely two discussed research performed at French-language institutions.
Belga’s Dutch-language reporters indeed have a notable focus on national (and regional) science news, with 71% of articles in our corpus coming from the national news desk (of which 84% by the general national desk, 10% by the economy desk and 6% by the politics desk). Since Belgium’s contribution to the global research landscape is proportionally much smaller, this demonstrates a large overrepresentation of Belgian (mostly Flemish) science news.
Nearly one-third (29%) of articles thus originated from the international desk. When looking at the location mentioned in the metadata of these articles, most were located in the United States (21%), Belgium (news considering European or other international institutions; 17%), Switzerland (14%, generally about UN institutions), the United Kingdom (9%), the Netherlands (8%), and France (7%), reflecting the language skills and focus of Belga’s reporters, the focus on news from (international) governmental bodies, as well as the dominance of the Anglo-Saxon world and English in the global research landscape. This is in line with Van Doorslaer’s (2009) finding that Flemish reporters tend to focus less on French sources than their Walloon colleagues, even, apparently, within the same media organisation. We also conclude that news from parts of the world where other languages than English, French, or German are spoken, including Denmark, Peru or Finland, has likely been funnelled through one of these other languages before reaching the Belga reporters, diffusing the connection between the original source and Belga’s news report. This was the case in an article about the monkeypox outbreak in Spain and Portugal, which was based on a German report by DPA, whose reporter had based themself on coverage by two newspapers: the Spanish La Vanguardia and the Portuguese Publico. There are also rather idiosyncratic news and translation flows. For example, most articles about the US were sourced from non-American news agencies. We were able to obtain the source text of some of these, and they show that Belga reporters use indirect German-, Dutch- and French-language sources to cover US news. In some cases, it seems that this news was issued from US-based offices of the news agencies involved, though further research is necessary to unravel these transnational and multilingual news flows.
Sources by type
Sources by text type.
Sources by organisation.
Since Belga is itself often used as an indirect news source, it is striking that its news again is largely based on other texts - with at least 77% of articles in our corpus based on coverage by another news medium or a press release. What’s more, when reporters use news agency content, this may again be based on a press release, creating a long chain of steps between the original news source and the audience, as in the example about DPA in the previous section. This practice appears to be largely unquestioned: multiple reporters noted that they trusted indirect information from established sources to be factual.
Sources by organisation
In the final step of the analysis, we looked at the types of organisations news in our corpus was sourced from (originally), as this again gives us an idea of news diversity as well as who weighs in on science communication through Belga. This data is shown in Table 2. Again, note that the total number of sources is higher than the total number of articles.
Most notable here is the large proportion of news sourced from government organisations (from Belgium and abroad). For example, there were many reports about news from the World Health Organisation. This is in line with earlier research that found that news agencies tend to rely on government sources (Al-Rawi and Iskandar, 2022). We also want to highlight the proportion of news sourced from commercial organisations. Many of these articles discussed surveys conducted seemingly as a form of content marketing, sometimes in collaboration with universities. Unfortunately, it is beyond the scope of this study to further investigate this type of news.
Discussion and conclusion
In this study, we triangulated different research methods (a quantitative content analysis, a survey, and interviews with newsroom staff members) to shed light on the daily operations within Belga, Belgium’s national news agency. Specifically, we sought to answer the question of how Belga sources its science reporting and how this is embedded within the newsroom structure. This article constitutes a valuable contribution to the small but rapidly expanding body of scholarly research on news agencies. In this final section, we discuss and contextualise our main findings. Finally, we acknowledge shortcomings and propose next steps for follow-up research.
Answering our first research question, we find that science news is not a priority for Belga, despite heightened attention to science-related news as a result of the COVID-19 pandemic (Nelissen and McMartin, 2022; Wang and Lu, 2022). There is no designated science beat and investing in scientific reporting is not acknowledged as a key target by the Deputy Editor-in-Chief we interviewed. One of the more striking findings of our study is that the geographical spread and linguistic diversity of Belga’s science news sources are quite limited, since the agency’s Dutch-language newsroom focuses heavily on national (mostly Flemish) news provided in Dutch. Science news largely depends on regional correspondents in provinces with leading research institutions, with about two-thirds of science news being sourced from within Belgium and almost exclusively in Dutch in those cases. This is quite disproportionate to the scope of Belgian (Flemish) scientific institutions globally. An explanation may be that the news value of proximity trumps others, in line with the finding of Dumas-Mallet et al. (2019) that Danish newspapers preferentially cover national research. Moreover, Belga likely focuses on news from Belgium to differentiate itself from international news agencies, which its customers may get from other sources. However, we also found a surprising lack of Walloon sources, with most science news not involving any interlingual translation. This was rather surprising, considering Belga’s bilingual operations and national orientation. Further research is needed to explain this phenomenon. It may be that Belga reporters interpret proximity as news from the same region. Or perhaps there is an (unconscious) preference not to translate. International science news, then, was almost always sourced from other news agencies and largely based on the Anglo-Saxon world. Moreover, it appears that international news is sometimes translated multiple times before ending up in Flanders. This answers our second research question and its sub-questions. Press releases and news agency reports constitute almost three-quarters of the source texts used for Belga’s Dutch-language science news (RQ2.1), predominantly focusing on Belgium itself (RQ2.2) and sources stemming almost entirely from universities, research institutes and hospitals, governments, and for-profit organisations (RQ2.3).
These findings build on the existing literature and at the same time create new questions to be answered by follow-up research. In their seminal study, Lewis et al. (2008) found that British newspapers heavily rely on press releases and news agency copy for their reporting. This conclusion was later shared by other scholars assessing news content in Australia and the Netherlands, with more recent studies adding that predominantly online news was more prone to repurposing press release and/or news agency content verbatim (Boumans et al., 2018; Johnston and Forde, 2011; Welbers et al., 2018). Our contribution to this body of scholarly work is that international science-related news agency copy too depends largely on news agencies from institutions and organisations. This begs the question of how this affects the general flow of information. In line with previous research, our data show that the assumption of news agencies as reputable news sources is somewhat tarnished by the finding that press releases are influencing not only news organisations, but also news agency copy. Moreover, the role of press agencies in the coverage of science news has so far been ignored in science communication research, which typically assumes a direct relationship between press releases and news coverage. Intermediary organisations, such as news agencies, should be studied further in this context for a more complete view of the full science news ecosystem.
We position our research results within a multi-dimensional framework of news diversity (Hendrickx et al., 2022). Belga’s ownership structure of mainly being funded by legacy news organisations allows it to solidify its uniqueness in brand diversity, as well as the power to steer content and overall news diversity in Belgium for better or for worse. Moreover, applying a translational lens to study source diversity reveals the cultural influences and orientations of news organisations and the pervasiveness (or lack thereof) of interlingual translation in the newsroom. This provides a more complete picture of where news comes from and how it is altered along the way, opening new conceptual doors for scholars interested in the intersection of translation studies and journalism studies. Our mixed-methods study design mainly focused on aspects of production and content diversity. Since the majority of the science articles in our sample were sourced from press releases and other news agency reports, this returning news production practice risks negatively impacting the overall diversity of news content as Belgian journalists too have previously acknowledged readily using Belga copy for their own reporting (Van Leuven et al., 2013). The findings of our case study paints a picture of news agencies' contribution to overall news diversity, although this study only constitutes the starting point for future research and debate on this until now underdeveloped yet significant relationship.
We believe our study introduces many new trajectories for future research. First, Belga launched a designated English-language news service in March 2022. At the time of writing still in its infancy, the service is explicitly funded by the Flemish government with an overt agenda to promote Flanders abroad to combat the dominance of French-language sources for foreign reporting on Belgium (Verbergt, 2022). As a result, further scholarly attention to the role of translation and the treatment of source material within the news agency is desired. We are also particularly interested in gauging how Belgian news outlets (re)use Belga’s news copy, given our findings of how Belga itself sources its (science) content.
Ultimately, we must acknowledge a few shortcomings. The scope of our case study was limited to the sources used in science news reporting by the Dutch-language component of the Belgian national news agency. As a result, we cannot claim with certainty that our conclusions are easily transferable to other types of news, nor other media markets. Also, our content analysis was limited to the original sources of news agency copy without an in-depth homogeneity analysis as has previously been the case in the Flemish media market (Hendrickx and Ranaivoson, 2021). We are thus unable to verify or make statements regarding the extent to which Belga repurposed content. Additional research can take this approach to further contribute to the ongoing debate about the role of news agencies in media ecosystems and societies.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the KU Leuven Impulsfonds (IMP/20/012).
