Online discussion threads as promotors of citizen democracy: Current opportunities and challenges for small- and medium-sized media organisations in Finland
, e | 12 mar 2024
Pubblicato online: 12 mar 2024
Pagine: 173 - 192
DOI: https://doi.org/10.2478/nor-2024-0012
Parole chiave
© 2024 Klas Backholm et al., published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
In terms of the journalism–audience relationship, news journalists have traditionally considered themselves as the producers of newsworthy mass media content, and the audience as the main news target group, and ideologically as those having the right to know what is happening in society (Council for Mass Media in Finland, 2011; Kovach & Rosenstiel, 2021; Thomas, 2022). However, technological development has meant that the role of citizens has shifted. Instead of being passive consumers, they may become active participants in the public debate or producers of media content in online commenting systems or similar journalistic publication platforms (Deuze et al., 2007; Marchionni, 2013; Nip, 2006).
In relation to citizen democracy, this shift ideally promotes the fundamental rights of, and possibilities for, the citizen not only to be informed about societally relevant issues but also to actively participate in discussions – and make one’s voice heard – in the public sphere about such issues (Habermas, 1986; Strömbäck, 2005; Zamith & Lewis, 2014). In this article, we focus on one subcategory of journalist–user interaction: how journalists and personnel in charge of moderation strategies in small- and medium-sized regional media outlets can use deliberative norms to promote a democratically sustainable discussion with citizens, within the limited space in user-generated discussion threads published on websites or similar journalistic publication platforms (Canter, 2013; Santana, 2011).
The study was conducted in Finland, a country with a well-developed technological infrastructure and extensive digital skills among the population (European Commission, 2022; Kohvakka & Saarenmaa, 2022), and where the societal prerequisites for citizen participation in online public spaces are thus good. We used participatory journalism (Borger et al., 2013; Deuze et al., 2007) and deliberative democracy as main study frameworks (Cohen, 2002; Fishkin, 2011; Strömbäck, 2005). We look both at to what degree journalists are currently using deliberative norms when they are moderating discussion threads and at the more general opportunities and challenges journalists encounter with user-generated content in relation to central citizen democracy principles such as user participation and interactivity. This article thus contributes to the field by adding to recently published works linking deliberative theory to journalistic practices (Masullo et al., 2022; Ziegele & Jost, 2020).
Some aspects of journalists’ viewpoints on the changing journalism–audience relationship have been studied extensively. For instance, scholars have focused on how journalistic moderation of user-generated discussion threads has evolved (Bergström & Wadbring, 2015; Boberg et al., 2018; Canter, 2013) and how journalistic choices regarding pre-publication content moderation may or may not improve user participation (Singer & Ashman, 2009; Wolfgang, 2016). On the other hand, research related to post-publication participatory discussion-thread interaction between journalists and the audience has rather recently received scholarly attention (Masullo Chen & Pain, 2017; Salonen et al., 2023; Salonen & Laaksonen, 2023). We aim to contribute with new knowledge about whether inclusion of deliberative norms in post-publication interaction may promote citizen democracy in journalistic moderation of discussion threads.
The shift towards growing citizen participation and inclusion in journalistic outlets has also been accompanied by more critical viewpoints, highlighting that scholars need to be realistic and have a sober viewpoint in considerations of the degree to which interactive technology and participation may promote citizen democracy (Peters & Witschge, 2015; Robinson, 2010; Thomas, 2022). Studies have also highlighted directly harmful effects that negative user input – such as hate speech or uncivility – may have on citizens as well as journalists (Quandt, 2018; Ziegele & Jost, 2020), and the potentially added workload for journalists that may come with handling user-generated content (Wolfgang et al., 2020). Some current studies have linked such negative phenomena to media organisations choosing to discontinue their online user-generated commenting systems (Liu & McLeod, 2021). In this article, we discuss to what degree small- and medium-sized regional media organisations, with limited resources available for moderation work, can limit the negative effects of potentially problematic user input in discussion threads by applying deliberative strategies.
The basic idea of online user-generated discussion threads originally was to provide a venue on journalistic websites where the public could discuss current topics (Santana, 2011; Schultz, 1999; Wolfgang et al., 2020) in a collective shared space, gathering varying opinions about a news topic (Bergström & Wadbring, 2015; Masullo Chen & Pain, 2017; Singer & Ashman, 2009). Currently, the publication platforms for discussion threads have naturally expanded to include websites as well as social networking sites. Discussion threads are categorised as a form of participatory journalism, which includes processes where citizens in various ways actively contribute to the news content produced by professional journalists (Borger et al., 2013; Deuze et al., 2007; Morrison, 2017), and where these processes take place within some form of publication frame designed or administered by these professionals (Abbott, 2017; Nip, 2006).
From a deliberative democratic viewpoint – whereby the value of discussion between lay citizens is considered crucial for a well-functioning democracy (Dahl, 1989; Gutmann & Thompson, 2004) – interaction in discussion threads may ideally support the empowerment of citizens not only by promoting active participation but also by creating opportunities for citizens to discuss, or follow others discussing, specific current topics in a public space (Liu & McLeod, 2021). Furthermore, such interaction may increase proximity to societal and political decision-making processes by allowing interaction with the journalists who have produced the news content and who thus have an insight into underlying facts, publication choices, or similar relevant parts of the content (Borger et al., 2013; Marchionni, 2013; Strömbäck, 2005; Zamith & Lewis, 2014).
A well-functioning interaction assumes that the discussion takes place within a safe space, where participants experience, for instance, that their opinions get heard, and where all are equally included and respected. These norms are central within deliberative democratic theory (Fishkin, 2011; Strömbäck, 2005), a normative framework that underlines the importance of discussions for promoting citizen democracy. According to the theory, participants in such discussions should adhere to certain norms, including reciprocity argumentation, sincerity, inclusion, facts-based argumentation, and respect, if they want to strive for a democratically sustainable discussion climate (Chambers, 2003; Fiskhin, 2011). Empirical studies have illustrated the benefits of including supportive tools or actions, such as moderating actors and discussion rules, when striving for a more deliberative discussion climate (Strandberg et al., 2019).
However, some journalism scholars have pointed out that the digital shift towards a “participation paradigm” ideal may have created an overly optimistic view on the democratic potential of, for instance, discussion threads (Kreiss & Brennen, 2016; Peters & Witschge, 2015; Thomas, 2022). This potential is affected by, for instance, underlying digital platform technological boundaries and journalistic trade-based structures and norms, and the opportunities and limitations these have in relation to available participation possibilities and underlying deliberative democracy norms.
Applying this to media organisations’ online discussion threads, in these, the media outlets and their journalists are the administrators of discussions. Thus, their strategies and approaches to the interaction become central for creating safe spaces, promoting deliberative norms in discussions, and avoiding negative participatory issues of uncivility, polarisation, or other forms of dark participation in the online debate (Peters & Witschge, 2015; Quandt, 2018; Stroud et al., 2015). A large part of this administration consists of moderation, which is when journalists on their websites or in social networking sites supervise, control, and enable user comments, to make sure that published content aligns well with the requirements of acceptable platform standards (Masullo et al., 2022; Wolfgang, 2016).
Moderation practices related to supervision and choices regarding allowing or stopping specific pieces of user comments have been studied extensively (Bergström & Wadbring, 2015; Boberg et al., 2018; Canter, 2013; Wolfgang, 2016). For instance, several scholars have investigated journalists’ views on such moderation, often finding a subgroup representing an optimistic or enthusiastic view of the benefits of commenting systems (Abbott, 2017). Robinson (2010) called these convergers, who see a democratic potential with including user comments and interaction. On the opposite side, traditionalists (Robinson, 2010) are those who tend to have a more pessimistic or disappointed view, and highlight risks related to factors hampering the democratic value of online discussion threads (Abbott, 2017; Borger et al., 2013).
The traditionalists’ scepticism or unwillingness to take on an active role as moderators has been attributed to several factors. For instance, journalists have stated that moderation becomes difficult due to the vast amounts of uncivil or dark content (Masullo Chen & Pain, 2017; Santana, 2011); the lacking organisational resources or massive workloads that set challenging time constraints (Boberg et al., 2018; Canter, 2013) or may affect journalists’ psychological well-being (MacDonald et al., 2016); or when moderation takes away resources from journalistic tasks considered more important, such as news production (Wolfgang et al., 2018). A relatively current development related to the pessimistic view is that some major news outlets are choosing to remove their online commenting systems due to, for instance, the challenges of uncivil user-generated content (Nelson et al., 2021). Liu and McLeod (2021) pointed out that such decisions are usually made based on media organisations’ views on the above-mentioned problems with content in and moderation of discussion threads, while we know little about what users themselves think about removing such features.
In addition to the basic moderating practices of supervising, controlling, and enabling comments in discussion threads, more recent scholars have investigated the use of so-called post-publication moderating activities (Masullo et al., 2022; Salonen et al., 2023; Stroud et al., 2015). Scholars studying such post-publication moderation can perhaps be said to represent the more optimistic view of the benefits of discussion threads for citizen democracy, as they often focus on finding ways to improve the journalist–user interaction or the discussion quality of threads.
To date, studies on post-publication moderation have introduced several concepts to describe and emphasise different moderation aspects or underlying journalistic tasks. For instance, Masullo and colleagues (2022) used engagement moderation, while Ziegele and Jost (2020) talked about interactive comment moderation. Others have highlighted the more general concept of gatekeeping, by using terms such as conversational (Salonen et al., 2023; Salonen & Laaksonen, 2023) or non-linear gatekeeping (Bro & Wallberg, 2015). In this study, we use Ziegele and Jost’s (2020) term, interactive comment moderation, to describe this function.
Some studies have outspokenly linked interactive comment moderation to key concepts of deliberative democratic theory. For instance, Masullo and colleagues (2022) and Ziegele and Jost (2020) showed that strategies where journalists outspokenly aim to engage with or choose specific response styles when approaching uncivil content, such as factual-based responses or acknowledgment of problematic commenters, may improve discussion quality. Stroud and colleagues (2015) found that establishing a relationship between users and moderators matters: Involving a recognisable journalist rather than an unidentified moderator will promote a deliberative discussion climate. These examples show that efforts can be made to strive for more democratically sustainable strategies in interactive moderation practices.
Building on the current knowledge base of journalistic strategies for pre- and post-publication moderation of user-generated discussion threads, and on a more general backdrop of the potential benefits and challenges with this form of interaction with users, two research questions were developed. Both aim to specifically investigate the role that deliberative norms may play in promoting citizen inclusion and democracy in journalistic work with basic and interactive comment moderation practices in small- and medium-sized regional media organisations:
To answer the research questions, semistructured interviews were carried out during 2021–2022 with 18 journalists or media workers, with a working experience between 3 and 32 years, from regional media organisations in Finland. We used two strategies to obtain the sample, mainly reflecting a snowball sampling approach. We either started by approaching existing media organisation contacts, or in cases where no such contacts existed, by examining discussion threads to identify active moderators or by directly contacting media organisations. Then, contacts either agreed to participate or forwarded us to others in their organisation that better fit participation criteria.
A majority of participants were female (
The interviews lasted between 38 and 82 minutes and were all carried out by the same project researcher, either as online or physical meetings. Regarding the main publication format, the sample was equally divided between those indicating audio and video broadcasting or written journalism, with nine participants representing each. Furthermore, participants published news-related content on the main social media platforms: All created news for Facebook and Instagram, while roughly 50 per cent had a similar presence on Twitter (now known as X). None of the participants worked with TikTok.
The sample represented public service as well as commercial media companies and consisted of two subcategories. Six interviews were conducted with personnel in charge of media organisations’ moderation and online strategies. This subsample included people with various titles and scholarly backgrounds, ranging from editors to technological development experts. All had either formerly worked with or had a detailed insight into journalistic tasks carried out in their organisation, and thus they felt comfortable discussing the research topics. This subsample is hereafter referred to as the “personnel in charge” group in this article.
Twelve interviews included news journalists directly involved in the production of online media content and handling of audience-produced comments on a continuous basis. The first subsample, personnel in charge, partly worked within the same media organisations as participants in the second group. As data were collected with two subsamples, two sets of interview guides were used. Both guides included themes focusing on how practices reminiscent of deliberative norms are used in the journalist–audience online interaction today, and on the relevance of developing new deliberative practices. The interviews with the personnel in charge encompassed two themes, the first focused on current challenges with negative comments and the second on how media organisations could develop their current practices. Interviews with journalists focused on how they moderate online threads and relate to such work, and on how the discussion climate could become more interactive and deliberative.
Thus, discussions with personnel in charge focused more on broadening the understanding about which technology platforms the media organisation uses, how these are used, the main challenges regarding, for instance, uncivil content, and how the organisation is currently developing their work with user-generated online discussion threads and related technology. Interviews with journalists focused more on defining how they work with such tasks, how they apply organisational guidelines about these tasks in their own work and potentially combine these with individual strategies, and how they prefer to handle uncivil user-generated content.
The interviews thereby used a relatively broad approach to the research topic. This was done as a reflection of the various possible combinations of, for instance, technological platforms, work tasks, and available economical resources at regional media organisations. Interviews with personnel in charge were carried out before data was collected from journalists working with moderating content, since the former could provide involved researchers with organisation-specific internal information not otherwise publicly available about, for instance, preferred technology or strategies. This information was useful in the subsequent interviews with news journalists, although, as mentioned above, all interviewed journalists didn’t work in organisations represented in the subsample of personnel in charge.
Furthermore, while interview questions were built on the current knowledge base of topics, such as application of deliberative norms in pre- and post-publication moderation of discussion threads (Fishkin, 2011; Masullo et al., 2022; Salonen et al., 2023), challenges with uncivil or dark participation (Quandt, 2018; Ziegele & Jost, 2020), and journalists’ general views of pros and cons with participatory journalism (Abbott, 2017; Borger et al., 2013), the questions used in the interviews were developed specifically for this study. This approach was chosen to come as close as possible to understanding the above-mentioned topics of interest in the two subsamples. Gathered data was analysed using a combined inductive and deductive approach, to allow for identifying unexpected as well as established themes. Conceptual frameworks (as described in the transcendental realism approach; Miles et al., 2019) were used in the construction of interview themes and when analysing and comparing interviews.
Finland was used as a case for several reasons. The country has been at the forefront regarding technological infrastructure, Internet access, and digital skills among the population (European Commission, 2022; Kohvakka & Saarenmaa, 2022). Therefore, it can be assumed that the societal conditions for participatory journalism are good. As illustrated by, for instance, Salonen and colleagues (2023) and in the public service Finnish Broadcasting Company’s guidelines (Finnish Broadcasting Company, n.d.), media companies in Finland have included user-generated discussion threads in news production practices. According to Ahva (2013) and Heinonen (2011), Finnish journalists recognise the usefulness of including active participation from users, while they simultaneously acknowledge a need to align such activities with established journalistic practices and norms to ensure high-quality content in their media outlets.
Furthermore, more generally, the Finnish media system acts in a context strongly rooted in basic democratic principles, promoting the basic rights of both working journalists and citizens interacting with the media. For instance, Finland is part of what researchers have called the Nordic media welfare system (Syvertsen et al., 2014), in which journalists are able to work in relatively stable conditions without external interference (Reporters Without Borders, n.d.). Finnish journalism also applies a national and publicly available self-regulatory system of media ethics: the Journalistic Guidelines (Council for Mass Media in Finland, 2011) based on democratic principles and that allow any citizen to make a complaint about journalistic content. The media system thus includes established guidelines about rights and obligations regarding creating and handling media content in the public space, supporting both traditional and participatory forms of journalism.
The current study focuses on small- and medium-sized regional media organisations. Small- to medium-sized in this context is related to the number of employees in the editorial offices where study participants worked. In the included organisations, roughly 20–100 employees worked with journalistic tasks (exact numbers not provided to ensure participant anonymity). All offices, however, also belonged to larger media companies, ranging across several regions (e.g., a regional newspaper belonging to a larger company consisting of several media outlets) and in which specific editorial offices can benefit from, for instance, technological development or content strategies drawn up at the head office.
In addition, small- and medium-sized regional offices were chosen as they, in comparison with larger media outlets, have received less scholarly attention. Regarding work with online user-generated content, smaller offices may be limited in different ways, for instance, when it comes to available resources or work staff. On the other hand, they may benefit more from increased proximity to and a lower threshold for interaction with their users.
We begin the results section with a short description of how the participants worked with user-generated content in online commenting systems. Then, results related to each of the two research questions are presented, and finally, we conclude with some practical and scholarly implications as well as a discussion on limitations. In the second part of the results section, where the research questions are revisited, results from the two subsamples (personnel in charge and journalists) are treated separately where relevant.
If looking at basic work tasks related to online discussion threads in Finland in general, all study participants worked in organisations that had established continuous contacts with users either on their organisation website or via chosen social networking sites, mainly Facebook. This reflects previous studies showing that media companies in Finland usually have some form of established practice for how user-generated content is included in news production (Ahva, 2013; Salonen et al., 2023). A majority of the participants worked with user-generated content in both formats. Two journalists represented companies that had recently moved away from using discussion threads on their websites but had such content on their social networking sites.
Regarding online discussion-thread moderation tasks in general, the sample was divided into three groups: those representing organisations where moderation of both the website and social networking sites was carried out in-house, those who had outsourced one of these channels to an external commercial company but moderated the other themselves, and those who had outsourced moderation of both channels. In media organisations that carried out moderation of one or several channels in-house, this was usually done by journalists who had a work shift that specifically included such tasks – while the journalist who had written the original news story linked to the discussion thread very seldom was expected to participate in the moderation. Furthermore, in organisations where moderation was outsourced, the external actor handled basic moderation activities such as supervising and controlling user comments, while possible interactive comment moderation activities were carried out by journalists in-house.
Thus, in addition to the division between the personnel in charge and journalists, participants in both groups represented a broad variety of combinations of work tasks and different content-handling setups. This variation is a reflection of the choice of study design, as we aimed to focus on understanding work with online commenting systems from a broad perspective rather than investigating a more homogeneous group, for instance, at the same editorial office.
The first research question focused on the degree to which journalists currently use deliberative norms in their practical work with moderation of online discussion threads. Deliberative norms in citizen discussions may generally include factors such as respect towards other participants’ viewpoints, sincerity, facts-based argumentation, and a willingness to potentially revise one’s preferences in light of new information during debates (Chambers, 2003; Cohen, 2002).
In general, both subsamples saw online user participatory features such as discussions on their websites or social networking sites as useful tools for promoting citizen democracy, as these may ideally create communities where readers can meet and debate current topics within a supervised environment. This is in line with the intended participatory functions of online commenting systems (Schultz, 1999; Wolfgang et al., 2020), which thus still hold, at least in the views of the participants. In addition, the result reflects the strongly rooted democratic principles and citizen rights in Finnish society and in the Nordic media welfare system (Syvertsen et al., 2014).
On the other hand, participants simultaneously acknowledged more commercial media organisation benefits with creating or keeping up online user communities as equally or more important. Such benefits included communities being helpful tools for steering user traffic to journalist-produced news content or advertisements, and for potentially supporting long-lasting bonds between the media organisations and their users. This result is hardly surprising, due to the well-established challenges with diminishing revenues or trust issues among, for instance, young audiences (Backholm et al., 2022; Newman, 2022), which media organisations currently struggle with.
When looking at how deliberative norms are included in work outspokenly with moderation in a journalistic online commenting system framework, such norms may be applied both strategically, in publicly available or internal guidelines and recommendations for acceptable user behaviour (see, e.g., Finnish Broadcasting Company, n.d.), and in the practical moderation work tasks journalists carry out when handling user-generated content pre- or post-publication. Both subtypes were mentioned by the participants, as well as the fact that the former reflects a democratically sustainable way of thinking about work tasks developed by the organisation, while the latter is about the practical application constructed and continuously revised by individual journalists. Furthermore, the latter is naturally limited due to the technological format of and related narrow textual space for pieces of content in discussion threads.
The research question focused specifically on the practical application of deliberative norms in moderation work, and the collected data showed that journalists currently apply several such norms in their ongoing work. These norms were mainly mentioned in the subsample of journalists who continuously work with such practices. Moderation work was usually carried out as interactions with users in the discussion threads and should thus be seen as interactive moderation practices (Ziegele & Jost, 2020). Furthermore, journalists described that they personally used such interactions relatively seldom, but applied deliberative approaches every now and then. Deliberative norms were seen as especially useful in relation to comments balancing in a grey zone between acceptable and uncivil content. While clearly identifiable hate speech or similar dark content was excluded directly, participants described several subtypes of deliberation that could be applied as a first step when handling uncivil content.
Journalists identified two main categories of uncivil content for which deliberative norms seem to work well (see Table 1 for a summary). The first was cases where arguments between users with varying opinions are taking form within dialogues, and where the tone is becoming heated or uncivil. In such cases, the participants mainly used friendly reminders about what the media organisation considers an acceptable discussion tone for interactions in their discussion threads. Journalists emphasised the relevance of a friendly tone in their reminders, both since they represented the administrating body and as a way to not aggravate discussion counterparts further.
How those working with online moderation use deliberative norms in interactions with users
| Arguments between users. | Friendly reminders about discussion tone. |
| Criticism towards other users, topics, or interviewees in the published journalistic product or the media organisation. | Friendly reminders about discussion tone, acknowledgement of negative criticism, facts-based clarifications. |
| Questions/corrections about the journalistic product or media organisation activities. | Facts-based clarifications, and a respectful tone, sincerity, and honesty in responses and corrections. |
As a second step, if such deliberative methods did not have any effect, the participants followed up with more drastic moderation methods, such as removing comments or closing the discussion thread. Some participants would have preferred to combine these methods with deliberative approaches to a larger degree by, for instance, providing users with personally directed explanations about why the decision to unpublish or stop new comments was made. However, technological challenges (e.g., difficulties with finding contact information for users behind discussion thread nicknames) and time constraints limited such efforts.
The second category of uncivil content where deliberative strategies had proven useful was in cases of user criticism. Such criticism could be directed towards other users in the discussion thread, an interviewee in the news piece, or the journalist who had produced the news product that the thread was linked to, the topic of the product, or the media organisation in general. In such cases, depending on the situation, the participants used a combination of strategies, including a balance between acknowledging the user-generated negative criticism, friendly reminders about the discussion tone, and facts-based clarifications of criticised topics.
A third, overlapping type of user-generated content where interactive moderation with deliberative norms worked well had to do with user questions or corrections about the journalistic product or media organisation activities, such as publication choices (see Table 1). Study participants underlined that the tone in such input may range from uncivil to civil, and from intended or unintended misunderstandings to well-needed remarks. In such cases, regardless of whether the journalist considered the user question or correction relevant, deliberative norms were applied in the forms of facts-based clarifications and using a respectful, sincere, and honest tone in responses and corrections.
The strategies listed for all three subtypes of user comments above are in line with the limited previous research that has focused on the application of deliberation in interactive moderation. According to this, journalists’ choices regarding, for instance, response styles and overall type of presence in interactions may contribute to a more democratically sustainable participatory climate in online commenting systems (Masullo et al., 2022; Stroud et al., 2015; Ziegele & Jost, 2020) and to an improved mutual understanding between discussants (Salonen et al., 2023). Similar experiences were also clearly present in the current sample. Furthermore, as mentioned above, interactive moderation, in general, was relatively uncommon among the participants, which echoes previous results with small- and medium-sized media organisations by, for instance, Carter (2013).
The second research question focused on the degree to which the participants considered that deliberative strategies could be further applied to promote citizen inclusion and democracy in work with user-generated comment moderation. Participants mainly discussed this question in relation to what they considered challenging with moderation work today, and how such challenges can be solved. To begin with, the main impression of the sample was that they in the last decade have shifted from an optimistic and perhaps somewhat naïve view about the potential of user-generated content, similar to Robinson’s (2010) convergers, towards what could be described as a more crude or realistic attitude, wiser from lessons learned in recent years.
Their current realistic viewpoint should not, however, be directly categorised as Robinson’s counterpart to convergers, traditionalists, as the participants were still carefully optimistic about the usefulness of user-generated comments, especially the personnel in charge. The sample positioning may more reflect the sober or realistic viewpoint found in, for instance, Peters and Witschge (2015) or Kreiss and Brennen (2016). In our study, three general dimensions of challenges that may affect the relevance of online discussion threads as promotors of citizen democracy took form; they are summarised in Table 2 and discussed in relation to deliberative norms below.
Three challenges that may affect the usefulness of online discussion threads
| Uncivility, hate speech, or dark participation in user-generated comments. |
Minor problem. Some degree of uncivil content present in threads. Hate speech not present. |
Disrespectful or mean-spirited comments hamper user participation or interactivity. |
| Lack of citizen representativeness among active commenters. |
Major problem. Certain age and gender groups dominate discussions. Certain individuals repeat and argue in similar ways across threads. |
A majority of citizens see discussions as biased and repetitive, and irrelevant for them. |
| Lack of time or economical resources for journalistic moderation of discussion threads. |
Major problem. Organisation size requires prioritisation. Moderation not prioritised in relation to other news production tasks. |
Pre-publication moderation and post-publication interaction toward sustainable discussions suffer. Lack of peer support creates discrepancies in moderation decisions between co-workers. |
One challenge revealed by the sample was about the already mentioned potential problems with negative content, such as hate speech or dark participation, as a hindrance of citizen empowerment in user-generated commenting systems (Bergström & Wadbring, 2015; Quandt, 2018). According to the sample, some content, usually of a less serious but still mean-spirited and disrespectful uncivil nature compared with direct hate speech (Masullo Chen & Pain, 2017), may still find its way through. Uncivil content is somewhat more common on organisations’ social media accounts than in website discussion threads, as pre-publication moderation of social media is not possible. However, the personnel in charge saw that they have managed to develop relatively well-functioning strategies for both platforms to avoid the vast part of negative content. These naturally included pre- and post-publication moderation and supervision of their commenting systems, as partly exemplified in the discussion of the first research question above.
A majority of participants also listed increased post-publication interaction with users as central for further development towards more sustainable commenting systems. While such responses did not necessarily refer to outspoken deliberative norms as a part of interactions, the main point was clear, and does reflect underlying deliberative principles (Chambers, 2003; Fiskhin, 2011) and recent studies highlighting positive effects of journalist–audience interaction (Salonen et al., 2023; Ziegele & Jost, 2020). If media organisations want to keep up participatory fora such as online commenting systems for their users, interactions in the form of well-articulated initiatives from journalists are key. This may be especially true in countries such as Finland, where citizens’ digital skills are high and opportunities to actively participate in online public spaces are good.
Furthermore, such efforts should, according to the sample, not be limited to responses provoked by user comments but expanded to, for instance, journalist-initiated broadening of news topics or providing better insight into the underlying processes related to political or journalistic decision-making related to the published news content. If realised, such efforts would promote citizen deliberation by providing a better understanding of complex or hidden circumstances and increased proximity between citizens and decision-makers (Strömbäck, 2005; Zamith & Lewis, 2014).
Another strategy for handling uncivil content mentioned by some personnel in charge was a refinement of choices regarding which types of news content that is linked to user comments. This was seen as especially relevant in light of user comments on organisations’ social media platforms. Participants described how this refinement is about moving from allowing comments on most of their news content to only on a small part of the published news. In more detail, companies previously allowed comments on most types of news, and only avoided this for a small subset, including news content that according to their editorial experience was most likely to provoke dark participation such as racism or personal attacks on interviewees. Currently, the approach has changed to allow comments on a more limited subset of news content.
Decisions about what to open up for comments were described by participants as a balance between choices due to the need to avoid dark content as described above and choices based on the news topic’s societal importance and timeliness (i.e., that users should be allowed to discuss an issue if it is seen by the organisation as societally important). Related to this, the participants also described having moved towards using various forms of user account requirements, especially in website commentaries. Such accounts force users to provide personal information to the media company before being able to comment, with the underlying idea that users may behave in a more civil manner when they can be identified by moderators.
In terms of citizen democracy, both strategies listed above seem to be general approaches that indirectly promote the functions of citizen participation and deliberation in the online public sphere (Abbott, 2017; Strömbäck, 2005), as they diminish the extent of dark participation and hate speech (Masullo Chen & Pain, 2017; Quandt, 2018). However, they may also have a hampering effect – for instance, opening up less content for comments naturally steers public discussions towards certain kinds of topics only, while other themes, perhaps hot topics that would need more public debate, are left out. User account requirements can complicate citizen participation, as starting such accounts may become a threshold for citizens and thus lead to fewer citizens being able to participate in discussion threads.
A second dimension of challenges that may affect the potential of user-generated threads for citizen democracy was discussed extensively among the participants. This challenge had to do with the overall citizen representativeness among active user commenters in the discussion threads. Participants considered representativeness as key to the trustworthiness and relevance of such fora. They underlined that in order for an online discussion venue to be a relevant public space for citizen deliberation (Fishkin, 2011; Strömbäck, 2005), its active participant constellation should at least be somewhat representative of the demographics of the media outlet’s users (and thus implicitly of Finnish citizens in general).
However, study participants were convinced that currently, this was not the case. Although they lacked detailed data to prove this, the participants mentioned that certain age and gender groups (often referred to as “middle-aged men”) dominate, while others, particularly young citizens, are more or less invisible. In addition, the subsample of journalists working actively with ongoing moderation described how, at least in the small- and medium-sized regional media context they represent, some dominating individuals often seem to move from one discussion thread to the next, bringing forth similar arguments or viewpoints – and arguing with the same counterparts from time to time.
The lack of representativity and a seemingly small group of active users may, according to the sample, create among non-active users negative connotations towards discussion threads as a venue for a very specific, and to them distant, subgroup of users. For journalists in the organisations, the representativity issues lead them to think of a cost–benefit imbalance with discussion threads: Why put time and energy into these practices if they are only seen as relevant or useful for a small part of one’s audience? In some responses to this theme, participants even called discussion threads a lost cause.
Taken together, the representativity issues were considered a major challenge and a factor that contributes to a pessimistic view of the potential benefits of discussion threads as deliberative public spaces in the future (Abbott, 2017; Borger et al., 2013). Clearly, this issue reflects the crossroads media organisations currently seem to be in regarding this type of content. Some major news organisations are choosing to remove their comment sections (Nelson et al., 2021), and Liu and McLeod (2021), for instance, highlight that more in-depth studies focusing on citizens’ viewpoints about removing commenting systems are needed to better understand the potential impact of such choices and underlying issues for citizen participatory democracy.
A third type of current challenge that may affect how deliberative strategies can be further applied in discussion-thread moderation had to do with the practical challenge of not always having enough time or financial resources for moderation and similar work. While mentioned in both subsamples, personnel in charge related the issue to the size of their organisations, which forces them to prioritise. Moderation-related tasks are thus considered of secondary importance, in comparison with more important work tasks such as producing news content.
Journalists actively working with moderation themselves expanded on this. In small- and medium-sized companies, a moderator may have less collegial peer support available or be working alone when moderating decisions need to be made. For instance, participants explained that moderation is one of several work tasks one is expected to do during a work shift, which means that moderation mainly consists of the most acute actions related to putting out fires, instead of preventive or more thorough supervision or post-publication interaction. This was also considered frustrating by some, in relation to the effort made by commenting citizens, as they may have put a lot of thought and time into formulating relevant content that then may be handled poorly or carelessly by the moderator.
Related to this, several of those in the sample who carried out basic moderation tasks mentioned problematic discrepancies between co-workers regarding publication decision-making. Even though established organisational guidelines exist, the decision about whether to allow a user-generated piece of content in a grey zone is still up to an individual journalist. In their experience, co-workers at the same organisation seem to vary regarding their criteria and thresholds for stopping uncivil comments. Some participants highlighted this as a central problem, underlining that such instability in basic processes related to commenting systems impacts the relevance and usefulness of this feature as a whole.
An obvious prerequisite for online discussion threads as public spaces for citizen deliberation is that moderators are provided the resources to steer discussions in a sustainable direction (Fishkin, 2011; Strömbäck, 2005). Canter (2013) referred to this as a contradiction between a lack of resources and a sense among journalists that user participation would be important as a democratic function. The current data shows that in small- and medium-sized media organisations, this is a challenge that remains to be solved before efforts can be focused on developing practical deliberative strategies for moderation tasks.
For citizen democracy, it is also central that online participatory venues provide a safe space for commenters (Deuze et al., 2007; Fishkin, 2011), so that risks for uncivil assaults on individual citizens can be minimised (Stroud et al., 2015; Wolfgang, 2016). Thus, it is problematic if the boundaries of such safety vary depending on which moderator happens to be at work at the moment. Furthermore, the issue of lacking resources may increase stress-related psychological impairment among journalists allocated to such tasks (MacDonald et al., 2016). It may also add to the above-mentioned challenge with a lack of representativity. If too few journalists are assigned to these tasks, and active users in one’s discussion threads only represent a small part of overall users, the benefits of discussion threads may be of poor counterbalance.
In this study, we aimed to investigate how discussion threads in small- and medium-sized journalistic media organisations may be useful for promoting central citizen democracy principles such as user participation and deliberation, with Finland as a case study. Some limitations should be kept in mind when interpreting the results. A qualitative research design with a limited sample was used, and therefore generalisation of study results should be made with caution. Furthermore, the sample represented a specific media system and worked with varying tasks related to online commenting systems within organisations of different sizes, with varying strategies regarding, for instance, outsourced moderation practices. Thus, further studies focusing on other media systems are needed to confirm the current findings.
With these limitations in mind, a main study conclusion is that media organisations, at least in the studied context, indeed seem to be at a crossroads situation regarding future directions of discussion threads. The organisations are clearly aware of the main benefits and challenges of this type of user-generated discussion system in relation to both citizen democracy and journalism practice, while the uncertainty lies in whether, in a near future, the pros will outweigh the cons, or vice versa.
While some recent studies (Liu & McLeod, 2021; Nelson et al., 2021) have reported that newsrooms are choosing to shut down their user-generated discussion systems, the current sample seemed hesitant regarding any major decisions to potentially develop or phase out such systems. In relation to citizen democracy and the (in Finland) well-established rights of citizens to openly provide feedback to and criticise authorities and institutions, journalists in the sample saw a clear role as providers of public spaces for user-generated input and deliberation. Thus, another conclusion is that the need for media organisations to engage in active interaction with the public will not in any way disappear in years to come. Rather, the sample saw a lot of potential in investing in this, especially by developing improved strategies for interaction by, for instance, applying deliberative norms to the discussion context – if current challenges can be overcome.
A related practical implication of the study for media organisations is thus that increased interaction with users, in various forms, may be a key factor for developing well-functioning and democratically sustainable participatory journalism practices. Applications of deliberative norms within journalistic interactive moderation by, for instance, active presence, acknowledgment of varying opinions, and an honest and sincere discussion tone promote positive meeting experiences and should be encouraged in varying forms of journalist–user interactions.
A main issue that the study raises is, however, in which venue such deliberative interaction will take place. In other words, have user-generated discussion threads, as they are currently defined, become too contaminated due to challenges with, for instance, uncivility, representativeness, or resource availability (especially in smaller media organisations), or can they still be redefined into the sustainable interactive meeting spaces they ideally could be? And related to this, were perhaps the journalists and scholars who have had a more sceptical viewpoint about the relevance of the user participation paradigm (Robinson, 2010; Thomas, 2022) right all along? If so, how can we avoid mistakes made in past participatory efforts within online commenting systems when new spaces for deliberative interaction between journalists and users take form?
Regardless of which form the participatory spaces for such journalist–user interaction may take, future research in the field should continue to identify how interactive features can be designed to promote sustainable democratic discussions by, for instance, applying deliberative norms in relevant ways. If, or when, the public spaces where journalists and users meet will take new forms in the coming years, for instance, within the Metaverse concept (Park & Kim, 2022), an additional future research direction could be to identify which current best practices for democratically sustainable interaction can be transformed in a relevant way for such new venues, and where new applications of, for instance, deliberative norms, need to be developed.