Foreign reporting and sourcing practices in the network sphere: A quantitative content analysis of the Arab Spring in Belgian news media

Introduction

In the contemporary context of globalization and cross-national exchanges, foreign coverage is an increasingly important information source for citizens ‘to expand their knowledge about the world beyond the immediate horizons of lived experience’ (Smith, 2005: 1471). Yet ample studies have shown that the narration of international events provided by traditional news outlets is often distorted by national lenses and an overrepresentation of elite countries and elite sources (Van Leuven et al., in press; Galtung and Ruge, 1965; Joye, 2010).

However, in light of recent advancements in digital technologies and within a global arena of news provision, news from across the world increasingly reaches audiences through many more channels, including social media platforms (Heinrich, 2011, 2012; Castells, 2011). The characteristics of social media platforms, more specifically their interactivity, connectivity and flexibility, enable users to hook up with others in virtually any connected spot on earth.

Within the interactive spheres of such networks, links are shared, information is distributed and news is commented on, on a scale not seen before. Many of the information providers who run these blogs, Twitter accounts or Facebook pages are citizen journalists, pressure groups or private persons whose intention is not necessarily to act as journalists. However, they have access to pieces of information that potentially add to the overall picture of news stories. And as stories unfold in virtually no time online, this new breed of information providers is capable of adding viewpoints, story angles or background information that might otherwise go uncovered (Heinrich, 2012: 766–767).

Yet, how can traditional media organizations seize these kinds of information for the purpose of journalistic coverage? Several studies illustrate that in the everyday news production process, social media are rarely used to source information material as many journalists admit they struggle with information overload, language hurdles and the doubted reliability of online information (Van Leuven et al., in press; Hafez, 2009; Lariscy et al., 2009). In contrast, case studies point out that journalists might be more inclined to consult social media sources in the context of breaking news or media restrictions where they cannot (immediately) access the area themselves and therefore try to ‘learn from on-the-ground sources’ (Lotan et al., 2011: 1376). The Egyptian uprisings in January 2011 serve as one example in this regard. Bloggers or Twitterers sent messages from inside the crowd on Tahrir Square, with each of them being just one voice amongst many who added information pieces to the unfolding story. The sheer amount of information pieces available via a multitude of platforms made for a unique mix of text and context: ‘At times domesticated for a specific national audience, yet accessible for a global community of consumers, the information pieces taken together formed a complex news map of the events’ (Heinrich, 2012: 768).

How traditional news organizations tried to incorporate this ‘global news map’ through sourcing social media in their coverage of the grass roots uprisings in three Arab countries is the focus of this article. The research presented here rests on a quantitative content analysis to discuss if and how Belgian news coverage adapted to the network sphere and displays diverse sources. More specifically, the focus is set on four Belgian newspapers and two Belgian broadcasters and their coverage of the street protests in Egypt, Tunisia and Syria in 2011. Do news organizations allow information providers such as bloggers or Twitterers into their source networks or is news coverage of the uprisings still dominated by traditional sourcing practices and mainstream sources? Does Belgian news coverage mainly build on local information sources and does it provide an ‘on/from the ground’ perspective or is the news domesticated for a Belgian/Western audience? Are there cases in which journalists are more inclined to use unofficial or non-mainstream sources? Theoretically based on the ideas put forth by the network journalism paradigm and in research on sourcing practices, this article sets out to shed light on these questions. In the literature review, we first discuss the implications of the evolution towards a network sphere for the future of foreign reporting, and especially sourcing practices.

Conceptual framework

Foreign reporting was until recently to a large extent a monopolized business in the hands of major traditional news organizations. Spearheaded by international news agencies (Boyd-Barrett, 1980) and an economically expensive operation, very few players dominated the field of international news provision for decades (Hamilton and Jenner, 2004). ¹ First, major changes in the foreign reporting business occurred with the emergence of 24/7 news outlets such as Cable News Network (CNN) (Cushion, 2010; Volkmer, 1999). These outlets targeted global news audiences and contributed to a continuous news and information flow across borders. The next step in the evolution towards a more fragmented map of news providers came with the emergence of the Internet and the proliferation of digital technologies. These developments significantly lowered production costs, granted access to audiences around the globe and enabled individual users or smaller news organizations to produce and distribute news globally. In the networked era, everyone connected to the online world can at least potentially produce pieces of information and distribute them not only locally, but globally.

Along with this, scholars attest that the sphere in which news is being gathered, produced and disseminated is significantly changing (Hermida, 2010b; Sambrook, 2010). Networked digital media mark a media environment in which consumers become active contributors within a ‘participatory media ecosystem’ (Hermida, 2010a). The gates of information flows formerly controlled by mainstream media have become permeable. With the ‘emergence of a new electronic communication system characterized by its global reach, its integration of all communication media and its potential interactivity’ (Castells, 2000: 357), digital networks carry the potential to delineate physical borders and enable connection, interaction and collaboration between professional informants such as journalists, just as well as various alternative information providers, including Twitterers or bloggers. Globalization trends, here, contribute to the erosion of temporal and spatial distances (Appadurai, 1996; Beck, 2000; Hannerz, 1996) and in combination with digitalization have given rise to an increase in global information flows.

In line with the arguments put forth by scholars of globalization and digitalization, the paradigm of ‘network journalism’ (Heinrich, 2011, 2012) attests a shift in the organization of information exchange. It conceptualizes the spheres in which journalistic organizations operate today and aims to capture the structures of information flows of the network age. Where Castells speaks of a ‘network society’ (2000), the paradigm of ‘network journalism’ sketches the evolution of a complex network of global information nodes. Here, digital networks are characterized as structural patterns of information exchange that support a revised organization of information gathering, production and distribution. The network journalism sphere is made up of an interconnected system of information nodes and journalistic organizations that have become just some network nodes among many. Nodes can differ in size and reach (as small as an individual blogger and as big as a news organization such as CNN), but they all roam in a shared information space. Information strings criss-cross this sphere and connect nodes in a non-linear fashion. Some nodes might have more impact on information provision than others, but nevertheless they all can potentially build or extend connections with other nodes via digital paths.

Methodology

The analysis presented here discusses if and how Belgian news coverage of the uprisings in three Arab countries displays diverse sourcing practices (RQ1–3). These events are related to a wave of democratic grass roots uprisings in many North-African countries in 2011 − often called the ‘Arab Spring’ ² – but we specifically focus on three cases. First, we selected the countries Egypt and Tunisia because they were the scene of major protests whereas other countries such as Sudan and Saudi Arabia only knew minor protests. These countries were also the only two where the government had fallen at the time this research was being conducted. Second, we selected Syria because of the violent response of the Syrian regime to the protests that took place resulting in a civil war. Besides, in Syria foreign correspondents were/are banned, which may have prompted journalists to rely more on network journalism practices as an alternative to gathering information about the uprisings. The absence of Belgian or Western tourists, implying that news about the Syrian street protests is less appealing for Belgian audiences (proximity news value, see Galtung and Ruge, 1965), further alienates it from the other two cases, and may be an extra trigger to practice network journalism (RQ4). These assumptions lead us to the following research questions:

We developed a quantitative content analysis to examine coverage of the selected cases in four Belgian newspapers and at two Belgian broadcasters. The first focus of the study lies with the sources in the news, or the people and organisations that are quoted in the news (RQ1). We developed a categorization of nine types of sources, grouped along mainstream and non-mainstream sources and a category of undefined sources (anonymous or unnamed sources). The group of mainstream sources categorizes politicians, government institutions, companies (and umbrellas), journalists, experts (scientists, universities), and a remainder category (celebrities, religious sources). The group of non-mainstream sources consists of ordinary citizens and non-mainstream groups (non-governmental organizations (NGOs), groups of demonstrators). A second focus of the article lies with the use of social media (Facebook, Twitter, weblogs, YouTube, WikiLeaks, etc.) as an information channel to access non-mainstream sources (RQ2). A third focus of the articles lies with the level of domestication (RQ3). In accordance with Joye (2010) we distinguish between ‘foreign news’ (e.g. news about Egyptian participants in the street protests, news about the actions of the dictators) and ‘domesticated news’ when foreign news events are adapted for a Belgian or Western (European or North-American) audience. For example, news about the uprisings can be domesticated by linking the street protests to their ramifications for Belgian or Western stock markets, or by interviewing Belgian or Western tourists that have been evacuated from the conflict area. The level of domestication can be limited (≤50% of the news article is domesticated) or strong (>50%).

We applied selective sampling (Wester and Van Selm, 2006) to collect all articles in the time frame of the actual street protests because at those moments demonstrators took reporting matters into their own hands, therefore acting as a potential networked news source for foreign reporters (Heinrich, 2012).

The self-cremation of Tunisian Mohamed Bouazizi out of despair with long-term unemployment and inequitable treatment by the Tunisian police force sparked the street protests from 18 December 2010 that resulted in the resignation of President Zine El Abidine Ben Ali on 27 February 2011. Following the example of the Tunisian people, Egyptian protesters stationed themselves on Tahrir Square from 25 January 2011 leading to the resignation of President Hosni Mubarak and the clearing of Tahrir Square on 14 February 2011. Finally, despite mass protests from 15 March 2011 onwards, Syrian President Bashar al-Assad refused to resign and tried to suppress the insurrection by military force and the peaceful demonstrations ended in a lasting civil war. To allow for comparisons between the selected countries, we delimited the Syrian data collection to the start of a new phase in the uprising, more specifically the association of opposition groups in the Syrian National Transitional Council on 17 July 2011.

We selected two popular (Het Nieuwsblad and Het Laatste Nieuws) and two quality (De Standaard and De Morgen) newspapers, Belgium’s four most important Flemish language newspapers (De Bens and Raeymaeckers, 2010). We furthermore included both Flemish newscasts at the public broadcaster Eén and the commercial broadcaster VTM. All newspaper articles were collected via Mediargus (Belgian equivalent of LexisNexis), all broadcast items via the Electronic News Archive. ³

In total, 1121 news items were selected and analysed by a team of four trained coders. A coding guide and registration form ⁴ were developed to ensure uniformity in the selection and analytical choices. A critically composed sample of 40 articles was tested for intercoder reliability with an outcome of Cohen’s Kappa values ranging from 0.76 up to 1.00. Analysis was carried out using PASW Statistics 18.

Results

The main body of 1121 analysed items consists of newspaper articles (79.5%) compared to 20.5% broadcast pieces. Despite the shorter sample period, the uprising in Egypt generated the highest percentage of articles in the total sample (45.3%) followed by Syria (24.4%) and Tunisia (16.4%). A total of 13.9% of the articles combined coverage of one of these countries with coverage of street protests in one or more other Arab countries and are therefore more generally categorized as ‘Arab Spring’.

RQ1 – The role of non-mainstream sources

With a total of 723 ordinary citizens in 1121 articles, or 36.9% of all 1961 sources, ordinary citizens outnumber all other sources (see Figure 1). In contrast to traditional news coverage (Van Leuven et al., in press), then, news about the selected cases seems to respond to Gans’ (2011) call for multiperspectival news that represents the general public. Political sources (26.3%) follow at the second rank. Non-mainstream groups such as NGOs or groups of demonstrators are ranked third (10.0%). The top three is followed by government sources (7.0%), journalists (6.6%), companies (6.0%), experts (5.4%), a remainder category of mainstream sources (1.3%), and undefined sources (0.5%). The fact that ordinary citizens and non-mainstream groups quantitatively outnumber most categories of mainstream sources in this study is a first indication of their importance in Belgian coverage of the selected uprisings. A more broad comparison of mainstream (52.6%) and non-mainstream (46.9%) sources further reflects the balance between different types of sources. ⁵

OPEN IN VIEWER

Figure 1.

Country versus % of sources (N=1961).

Most sources in the sample are Arabs (54.4% of all sources), followed by Belgian (18.8%) and Western (18.1%) sources. In accordance with the distribution of articles over the three countries, most Arab sources are Egyptian (30.7% of all sources), followed by Syrian (12.6%) and Tunisian (7.0%) sources. Most mainstream sources are of Arab (38.5% of all mainstream sources, mainly political sources, government institutions and journalists) as well as Belgian (19.9%, mainly companies and experts) and Western (33.4%, mainly political sources, journalists and experts) origin, while most non-mainstream sources are Arabs (72.8% of all non-mainstream sources) or Belgians (17.8%).

The majority of sources are personally involved in the event (45.9%) or provide eyewitness reports (34.8%). Most ordinary citizens are not quoted as vox pop (11.5%) merely to add colour to mainstream sources’ standpoints. The finding that most ordinary citizens cited are Arabs and that non-mainstream sources are significantly (p=0.000) more prominent in quality (48.7% of all sources in quality media) compared to popular media (40.3%) indeed goes against complaints about tabloidization in news coverage (Hauttekeete, 2004). The relatively higher prominence of ordinary citizens in quality media seems to be inspired by network journalism practices rather than by a commercial strategy to attract a bigger audience that is often applied by popular media. Instead most ordinary citizens are consulted as eyewitnesses (78.6%), many of them Arab demonstrators testifying about their participation in the street protests or Belgian tourists talking about their evacuation from holiday resorts. Although we cannot conclude solely based on these findings that every eyewitness account brings more viewpoints and story angles on board (as envisioned by the network journalism paradigm), the findings reveal that ordinary citizens are more often heard compared to previous research (Van Leuven et al., in press; Gans, 2011).

Nonetheless we also found signs of traditional sourcing practices as, in accordance with previous research (Van Leuven et al., in press), the analysis shows that ordinary citizens are relatively more present in background articles. ⁶ In contrast, most mainstream sources (besides experts who are almost exclusively consulted to provide background information) are relatively more present in factual news reports where journalists have less room for active news gathering and prefer to rely on official sources (Gans, 1979). Moreover, if we only consider the 763 sources that are firstly quoted in each article (‘dominant’ sources), our original optimism is tempered as 32.1% of them are political sources. This finding points out that – although ordinary citizens are most prominent in the overall coverage – journalists still prefer to place ‘authoritative’ sources first to introduce an issue and thereby set its interpretation framework, which is in line with traditional literature on source selection (Gans, 1979; Hall et al., [1978] 1999). This finding is confirmed if we make a more broad comparison between mainstream (61.9% of all dominant sources) and non-mainstream sources (37.9%), where the balance lifts more towards mainstream sources compared to the overall coverage (where ‘only’ 52.6% of all sources are mainstream sources). There is nonetheless reason for optimism about the importance of non-mainstream sources in coverage of the selected uprisings: ordinary citizens (25.7%) and non-mainstream groups (12.2%) follow at rank two and three as dominant sources. This finding is especially important as NGOs, social movements and groups of demonstrators are more important as dominant news sources compared to the overall coverage (10.0%), which especially for NGOs points to their development as authoritative news sources ‘in given fields of discourse’ (Davis, 2000: 50).

Country comparison

The analysis shows a significant coherence (p=0.000) between types of sources and country of uprising. Figure 1 shows that non-mainstream sources are better represented in coverage of Tunisia (52.3% of all sources in the Tunisian case) and Syria (53.5%) compared to coverage of Egypt (45.4%) and items that discuss more than one Arab Spring country (38.6%). Coverage about the uprising in Syria contains many more quotes from non-mainstream groups, mainly groups of demonstrators or NGOs that condemn the disastrous humanitarian situation and the many civilian victims in the conflict. Coverage of the Syrian uprising also quotes relatively more local Arab sources (64.6%). A look at the dominant sources further confirms that Syria is a distinct case as it is the only country where mainstream and non-mainstream sources are balanced (51.0% non-mainstream sources). In all other cases, mainstream sources are more often dominant (27.3–42.1% non-mainstream sources). It thus seems that media restrictions prompt journalists to refer more to non-authoritative sources, even as prominent as the first quote in the article/item. Journalists, here, seem to be forced to revise their sourcing practices to ensure coverage, find stories and viewpoints and to validate news.

Mainstream sources are especially more dominant in coverage of two or more Arab Spring countries, which can be related to the fact that these news items quote more experts (15.1% of all sources in ‘Arab Spring’ items) compared to coverage of the separate countries (2.0–4.2% of all sources in each case). A closer investigation of the sample reveals that many of these articles are background (35.9%) or commentary (22.4%) pieces that try to sketch the bigger picture of the protests, discussing possible reasons for the grass roots mobilization, the consequences for the countries in question and for the international community, the stance of the international community towards dictators and protesters, etc. In these instances journalists more often rely on background knowledge provided by experts (Albaek, 2011). This assumption is strengthened as we found that statements by experts are significantly (p=0.000) longer than those of other sources, which indicates that experts are often contacted to add more extensive pieces of context and analysis to news reports.

RQ2 – The role of social media and amateur footage

As far as we can observe in the news output (as journalists do not always explicitly mention their sources), social media are consulted in 10.3% of the articles which is a substantially higher proportion compared to previous research (Van Leuven et al., in press). The most important social media sources are Facebook (43 uses in 1121 articles/items), followed by Twitter (35), YouTube (28), WikiLeaks (18), nawaat.org’s Tunileaks (5), and several personal blogs (11). A total of 79 of 128 quotes (61.7%) sourced via social media are attributed to ordinary citizens. In other words, a substantive amount of 11.0% of all ordinary citizens cited are sourced via social media. Eleven social media quotes are attributed to non-mainstream groups. Most sources that were contacted via social media are Egyptian (34), Tunisian (7) and Syrian (12). It is nonetheless clear from the analysis that Belgian newspaper journalists still struggle with social media sourcing. Examples where social media sourcing adds value to the coverage are exceptional (e.g. references to nawaat.org, a website run by Tunisian activists outside the country). In many cases, social media quotes are not integrated in the news article, but instead displayed in a separate column to illustrate the content of the news article. For example, on 3, 5, 11 and 12 February, De Standaard added Twitter and Facebook quotes to articles about the Egypt uprising in a separate column (44 quoted sources in total). Most of these quotes express the experiences and emotions of demonstrators and supporters (e.g. ‘You can do it!!! Tunisia supports u! Good luck ☺’ or ‘Pfffffff hurry up Moebarak, we still have 20 countries to liberate’). One article adds a few tweets in a separate column to illustrate the discussion on the role of social media in the Egyptian uprising (e.g. ‘It smelled tear gasses, but mostly freedom #jan25 #Egypt #Tahrir’). Viewed from the perspective of network journalism, one could interpret this shift in sourcing practices as a move to allow more non-mainstream sources in the coverage (Dahlgren and Gurevitch, 2005; McNair, 2009). Conversely, one can still question how far these personal comments add depth and new viewpoints to the coverage, especially when they are not integrated in the news article but instead ‘packaged away from traditional media coverage’ (Heinrich, 2012: 769).

One important thing to note is that we did not count ‘amateur videos’ in broadcast news as social media sources because it was not explicitly mentioned if these videos were distributed via social media channels. It is however highly presumable that most of the amateur videos are distributed via Facebook, Twitter, YouTube or other social media platforms. As 17.5% of broadcast news or 3.6% of the total coverage contains amateur footage, the combined count of social media sources and amateur footage results in a considerable 13.9% of all coverage where Belgian journalists expand their sourcing practices.

Country comparison

The analysis shows a significant coherence (p=0.000) between country and social media use. Social media are more often used as sources in coverage of the Syrian uprising compared to the other cases (see Figure 2). Moreover, 35 of 40 amateur videos appear in coverage of Syria. This finding leads us to the assumption that journalists in the Syrian case more often relied on alternative sources and amateur footage (e.g. live reports and video images) to compensate for the access restrictions. It seems that the use of social media and amateur videos is inspired by network journalism practices, as images and information are provided that normally would not reach beyond Syria because of heavy media restrictions. Therefore, these sources clearly add depth and viewpoints to the coverage of the Syrian uprising that go against the official statements of the Syrian government. They give voice to the Syrian people even though they did not have face-to-face contact with journalists.

OPEN IN VIEWER

Figure 2.

Country versus role of social media (N=1121, % of items).

In contrast, Figure 2 shows that social media are more often discussed as a topic and merely mentioned (without further discussion) in coverage of two or more Arab Spring countries. This can again be ascribed to the larger amount of in-depth analyses, commentary pieces and background coverage that, for example, examine the role of social media in the uprisings.

RQ3 – Foreign or domesticated news?

The analysis shows that 46.7% of the articles are foreign news; 19.7% of the items are to a small (5.2%) or large (14.5%) extent domesticated for the Belgian audience; 33.5% of the articles are to a small (18.1%) or large (15.4%) extent telling the story from a Western perspective (see Figure 3). When compared to previous research (Van Leuven et al., in press) our case study can be considered as an exceptional news event where Belgian journalists were less inclined to domesticate the news. As domestication ensures that foreign events are more understandable, attractive and relevant to a domestic public (Clausen, 2004; Joye, 2010), the lower level of domestication may indicate that the Arab uprisings are often reported from the perspective of local sources which points to practices of network journalism and multiperspectival news (Heinrich, 2011; Gans, 2011).

OPEN IN VIEWER

Figure 3.

Country versus level of domestication (N=1121, % of items).

Country comparison

Figure 3 shows a significant coherence (p=0.000) between country of uprising and level of domestication. News about Syria is less often domesticated (61.9% of all items about the Syrian uprising are foreign news) compared to news about Egypt and Tunisia that more often discusses the Belgian interests in the area (24.4% and 29.9% respectively). This can be explained by the fact that contrary to Syria, Egypt and Tunisia are important touristic destinations, which automatically results in a higher level of domestication as Belgian interests are involved. Moreover it may be related to new sourcing practices as the network sphere allows journalists to get more easily in touch with local sources in Syria. These assumptions are confirmed by the finding that Belgian sources are heavily underrepresented in coverage about the Syrian uprising compared to the other countries: while only 2.3% of all sources in coverage about Syria are Belgian, their presence ranges from 16.7% to 28.9% in coverage about the other countries. Furthermore we see that news items that discuss more than one Arab Spring country more often domesticate the news for a Western audience (46.2%). As mentioned above, many of these articles are background pieces that discuss the ramifications of the uprisings for Western countries, for example in terms of oil prices or stock markets.

Furthermore we found a significant coherence (p=0.000) between sources and level of domestication (see Figure 4). More specifically, 60.2% of all sources in foreign news count as non-mainstream sources, compared to only 47.0% of all sources in the total sample. Moreover, non-mainstream sources are also more prominent (66.7%) in coverage that is only to a small extent domesticated for a Belgian audience which further points out that more non-mainstream sources are consulted in coverage that (mainly) focuses on the country of uprising. This finding seems to suggest that Belgian journalists made use of the new possibilities in the network sphere to rely more on non-mainstream sources in their reporting of foreign events. In contrast, 94.5% of all sources are mainstream in articles that are, to a large extent, domesticated for a Western audience.

OPEN IN VIEWER

Figure 4.

Source versus level of domestication (N=1961, % of sources).

Conclusion and discussion

The analysis presented in this article focused on the question whether Belgian news coverage of the uprisings in three Arab countries shows signs of diverse sourcing practices (RQ1–3) and whether we can find differences between the countries that are related to issues of proximity and media restrictions (RQ4). The findings are mixed.

On the one hand, we found signs of more diverse sourcing practices compared to previous research (Van Leuven et al., in press). Ordinary citizens and non-mainstream groups are important news sources, social media are relatively often used in the news gathering process, and the news about the uprisings is less often domesticated (RQ1–3). On the other hand, our original optimism needs to be toned down, as it appears that Belgian journalists mainly turn to official political sources to introduce the news and set the interpretation frame for the article. Moreover, non-mainstream sources may be important news sources in quantitative terms, but it seems that ordinary citizens (as eyewitnesses) and social media platforms are mainly consulted to illustrate the information in the news article. This finding is in accordance with previous studies that showed how journalists still value traditional sourcing practices (Bélair-Gagnon, 2011). In a recurring critique on sourcing ordinary citizens in the news, some authors state that this approach does not add depth and new viewpoints to the message but rather dumbs-down the information in the news article (Habermas, 1974). In contrast, Gans (2011) demands that journalists and news media should represent the general public and make their views and voices heard to foster public discourse. Other authors stress that this journalistic approach allows more ordinary people to understand the news from real-life experiences (Dahlgren and Gurevitch, 2005; McNair, 2009). In this sense, even when eyewitness accounts do not add a rational and well-considered argument to the public debate about the uprisings, they can still add a new layer to the discussion in telling the story from different real-life experiences.

However, we did find more characteristics of network journalism practices in news about Syria (RQ4), with a bigger role for non-mainstream sources and social media platforms, and with a lower level of domestication for the Belgian audience. First, this finding can be related to factors of proximity regarding news values, as no Belgian tourists were present in the country to provide a national framework of the street protests. Second, it seems that media restrictions inspired Belgian journalists to rely more on on-the-ground, non-mainstream sources that circumvented the restricted information access by means of digital networks and social media platforms. This finding indicates that in a situation of information chaos (McNair, 2013), the network sphere offers opportunities for sources and journalists to connect on a global scale and exhibit power abuse to a global audience. In this sense, our analysis challenges traditional political-economic conceptions of international news coverage as a means of ‘global dominance’ by those in power (Cottle, 2009).

Nonetheless, De Dobbelaer et al. (2013), who conducted a similar content analysis of the Arab uprisings in Belgian TV news, found that most journalists do not actively search for user-generated content on social media platforms, but instead copy this footage from international news agencies and international media brands (e.g. BBC, CNN). In other words, they rely on international news media to verify the reliability of user-generated content before they incorporate it in their news output. As journalists often fail to mention their sources, further research is necessary to assess whether Belgian newspapers equally passively relied on international news media to provide user-generated content. Nonetheless, although this finding points out that most Belgian TV journalists did not yet actively ‘practice’ network journalism during the Arab uprisings, it cannot be denied that a change is already visible in the news output. Network journalism is still in its infancy, and thus further research should follow-up possible changes in sourcing practices.

When analysing our findings, one should also take note that the ‘Arab Spring’ dominated the media agenda for several weeks and therefore most news media had their own correspondent at the spot (except for Syria). This may be one of the reasons why journalists had so many contacts with local, non-mainstream sources. Further research is needed to examine how far sourcing practices in this study can be explained by network journalism practices or by the deviation from everyday desk-bound sourcing practices where journalists may be more inclined to rely on mainstream sources (e.g. news agencies, government). Second, ordinary people played a central role in the Arab Spring as they were demonstrating in the search for political change. In contrast, most everyday foreign coverage stems from an institutional background such as economic negotiations between government leaders, election coverage, or intergovernmental meetings, where mainstream sources are the key players. Ethnographic studies in newsrooms may be a good starting point to examine how far journalists in their everyday newsgathering practices also make use of online networks and how far non-mainstream sources can be consulted to add depth to this kind of coverage.

Finally, even though we did find indicators of more diverse sourcing practices, our study does not reveal if the content provided through sourcing social media does deserve the label of a ‘global’ news outlook (Berglez, 2008). ‘Foreign’ news about street protests does not necessarily equal ‘global’ news with regard to the perspectives provided on a news story. Or to put it this way: an increase in the use of foreign and non-mainstream news sources does not automatically equal the provision of more ‘global’ perspectives. Berglez explains that a global outlook ‘seeks to understand and explain how economic, political, social and ecological practices, processes and problems in different parts of the world affect each other, are interlocked, or share commonalities’ (2008: 847). A global outlook, thus, consists of a richer tapestry in perspective and depth of a news story.

This article took a quantitative approach to sourcing practices at traditional news organizations. It lies in the nature of such research that it can reveal the balance in sourcing practices and the plurality of voices heard in a news story. It does, however, not tell us much about the content provided by social media sources. Does a higher quantity of citizen and non-mainstream contributions really equal a higher quality of foreign news coverage? Or are these sources mainly used to ‘colour’ or illustrate the information in the article? Qualitative research on social media sources could, here, complement our results and reveal insights with respect to ‘global’ perspectives and if and how the plurality of voices is mirrored in the news. The paradigm of network journalism suggests that news organizations develop new and continuous links with alternative sources. Social media can in this respect add to a more diverse global news map and further research will help to assess how journalistic organizations cope with the new sources at hand, how much of an influx they allow into their network and if we can witness a shift from ‘foreign’ to ‘global’ reporting.