Online News Audiences as Co-Authors? The Extent and Limits of Collaborative Citizen-Professional Journalism on Newspaper Comment Threads Online News Audiences as Co-Authors? The Extent and Limits of Collaborative Citizen-Professional Journalism on Newspaper Comment Threads

Recent research has demonstrated how comment threads published beneath online news articles are being transformed into fluid interfaces between professional journalists, their work and their audiences. Today’s audience-members are not only able to respond to pub- lished narratives but to embellish and, potentially, contest them: by posting comments based on personal knowledge about an issue and even using eyewitness testimony to directly affirm or challenge a story’s details. Though often stylistically “messy,” such comment posts go beyond merely manifesting and magnifying news discourses—let alone simply reacting to them. Rather, as on social media, posters can publicly discuss and debate the meaning and significance of stories, with the more informed and assertive among them contributing content so significant that it reshapes the texts themselves. In so doing, such claims-makers and counter claims-makers become hybrids of journalists (news producers), audience- members (news consumers) and claims-makers (news sources). Drawing on the author’s recent empirical findings, this chapter argues that online news has entered a dynamic but disruptive new phase in which journalistic authority may increasingly be contested, as “audience-members” begin to compete with “reporters” for authorship of news narratives.


Introduction
Much has been made of the opportunities, and tensions, arising from the expectation that journalism in the Web 2.0 era would become ever more "participatory" [1]: less a "one-way lecture" by reporters and commentators than a "dialogical conversation" between news professionals and their audiences [2]. If the traditional model was largely top-down-with privileged gatekeepers selecting, packaging and projecting stories to the public based on institutionalized judgments about what made something newsworthy-early optimists held that the newfound ability of those once dismissively termed "consumers" to contribute their own comments, eyewitness accounts and supporting evidence would transform journalism into an information-gathering and publishing process far more democratic, even bottom-up.
Nowhere was the possibility of reimagining journalism as a multi-directional, even citizen-led, exchange of news and views more clearly articulated than in its near-spontaneous occurrence during the Arab Spring. This cascading wave of popular uprisings across north Africa and the Middle East saw first-hand testimony, backed by digital photographs and video footage, escape through the cracks of creaking authoritarian censorship regimes via Twitter, YouTube and other social media platforms to spread "democratic ideas across international borders" [3]. In this case, the status of gatekeepers fell to a network of previously disenfranchised activists, transformed into (often inadvertent and involuntary) citizen journalists. Without them, Western media outlets would have known little of what was happening on the ground-and had meagre source material to inform their own reports.
Yet, 8 years on from the Arab Spring, there are grounds for arguing that it may have been both a singular event-or sequence of connected events-and (in media terms at least) a false start. For all the industry hype about transforming audiences into vibrant, engaged "communities" and newsgathering itself into a "collaborative" process led as much by lay "produsers" as trained reporters [4], recent studies point to a continuing resistance among professional journalists toward interacting directly with their publics, let alone surrendering their gatekeeping authority to them. A well-cited early newsroom-based study into a local newspaper negotiating the transition from print to online-only publication drew a marked distinction between the attitudes of younger, more digital-native, "convergers" enthusiastic about interacting with readers to enhance their output and wary "traditionalists" [5]. Though editors encumbered by dwindling newsroom budgets embraced the chance to harvest newsworthy content at little or no cost, while enriching stories already running by crowd-sourcing eyewitness accounts, photographs and video footage from audience-members, they were initially much more reluctant to recognize the importance of allowing readers to post comments and reactions (e.g., [6,7]). More recent work has found that, even when news organizations demonstrate an enthusiasm, in principle, for encouraging their audiences to spend time on their sites engaging in debate and discussion, and even contributing their own original material to inform and enhance their output, they are invariably hampered by inadequate resources. The "small size of online teams at news organizations" makes it "difficult" for hard-pressed journalists to "moderate," curate or engage in more meaningful ways with threads or other forms of audience-generated editorial contributions-together popularly known as user-generated content (UGC) [8].
Whatever cultural and organizational obstacles might be frustrating the wholesale, industry-wide adoption of collaborative newsgathering and reporting practices, however, there are intriguing (if still embryonic) signs that they are starting to emerge-even if this fact is not always fully recognized, or capitalized on, by news publishers. Nowhere is the potential for this more evident, moreover, than on the below-the-line comment threads routinely run beneath articles on news websites. Though comparatively under-researched to date, compared to Twitter, Facebook and other (external) social media platforms, these deliberative spaces are uniquely significant in two crucial respects. Firstly, they offer online audiences an opportunity to comment directly-and publicly-on news texts, by doing so on the sites to which they are native. In so doing, they capture a raw and immediate record of public reactions to news events, and the ways they are represented by specific outlets (and journalists), while offering intriguing insights into how individuals process media narratives and derive meaning from them, based on both their own prior knowledge, experiences and worldviews and the interpersonal dialogue in which they engage with fellow audience-members. More significantly, though, comment threads are important because they allow those with relevant knowledge and/or experience to contribute to narratives, not just respond to them, by adding evidence-based posts that confirm what a journalist has reported; qualify it with a more balanced or nuanced account; or even contradict the published one. When backed by firstperson testimony, credible and clearly cited statistics and examples, or other verifiable truthclaims, such posts have the potential to affirm, enhance, dispute or even debunk the substance of "journalist-led" articles on which they "comment"-helping to complete and/or correct otherwise "unfinished stories," and transforming audience-members into co-authors of the "finished" versions [9].
Based on discourse analysis of an extensive corpus of UK newspaper comment threads gathered during two recent studies into popular representations of low-income social groups carried out over a two-year period between 2016 and 2017 (much of it previously unpublished), this chapter explores the evidence thus far for considering below-the-line threads as extensions, rather than adjuncts, of the articles they accompany. Conceptually, it argues that the journalistic potential of threads is considerable, if (as yet) largely untapped. This is because they offer opportunities that go well beyond monetising audience engagement in the most superficial sense: for instance, by encouraging people to linger and debate the news to boost short-term advertising revenues. By contrast, more sophisticated evidence-based comment posts-though often obscured by the noise of routine chatter-have the power to provide nuggets of rich additional detail, background and context, alternative angles and follow-up ideas. In so doing, these dynamic interfaces between journalists, their articles and their audiences have the potential to be used as engines for promoting much more systematic, valuable and widely adopted forms of crowd-sourced, collaborative reporting.
their utilitarian function, as thinly disguised profit-generators that enable owners, editors and sales managers to demonstrate the breadth and/or depth of their user engagement to promote their brands and attract sponsors and advertisers. Among the most influential optimists in the user empowerment debate has been Axel Bruns, who has crystallized the explosion of interplay between publishers and their audiences as a "hybrid form of simultaneous production and usage" he defines as "produsage" [10]. Considering news websites specifically, he suggests these have fostered the rise of a new form of "citizen journalism" that "fundamentally disrupts the industrial journalism model by employing its users as journalists and commentators": a collaborative and participatory movement led by (though not confined to) digitalnative outlets such as IndyMedia, technology news site Slashdot and the South Korean-based "opinion leader" OhmyNews.
Such positive visions of audience and/or consumer empowerment have, however, been tempered by increasing recognition of the ways in which "the producing consumer" has been "co-opted" into "corporate strategies regarding the marketing and creation of content"-with everything from UGC published on news sites to "comments, bug reports and other feedback" fed by video-gamers to "developer-funded platforms at company websites" demonstrating how online forums are being hijacked for naked commercial purposes [11]. As Jonsson and Ornebring have argued, "strengthening and in different ways exploiting the relationship between users and their media (through 'community-building' on newspaper websites, for example) is an increasingly important part of monetising strategies used by newspaper companies." Quoting real-world examples, including ex-USA Today president and publisher Craig Moon's description of its comment and community features as devices for deepening audience "engagement" with the paper's "brand" while also creating a "stronger environment for" its "advertisers' messages," they also emphasized how contributing users act as "a form of 'free' labour" for budget-conscious editors [12]. Moreover, empirical studies of newspapers that were relatively early adopters of below-the-line comment threads suggest that, far from being used as sites of interaction between journalists and their audiences, or acting as leveling mechanisms to empower the public to inform or contest articles, they were seldom even visited, let alone read or responded to, by news professionals. Twelve out of 19 respondents to a survey of New York Times journalists by Schultz in the late 1990s confessed they rarely, if ever, looked at their own paper's comments forums, suggesting that, far from being the "people formerly known as the audience," as Jay Rosen memorably characterized them [13], active online media-users were largely sidelined, by being left to "discuss among themselves" [14]. Meanwhile, in relation to the deeper question of whether the rise of digital media has redefined "the boundaries" of the "public sphere," as some anticipated it would, the likes of Trenz have cautioned against accepting "the mainstream assumption" that it necessarily has "a new emancipatory potential"-arguing instead that, in political communication and other areas, "by and large" the Internet "continues to reproduce the national public sphere " [15]. Early evangelism about a citizen-empowering "interactive revolution" [16] and "grassroots journalism for the people, by the people" [17], let alone the "techno-deterministic optimism" and "web 2.0 ideology" succeeding it [18], have, then, been supplanted by more conflicted, less optimistic assessments of the liberating democratic potential of the Internet.
For all this fast-building scepticism, however, in the realm of online news publishing specifically, recent years have seen promising signs that comment threads and forums-dismissed by some in the journalism industry itself as "relics of web 1.0" [19]-are slowly being transformed into sites of meaningful inter-user (and user-producer) discourse that take us a step closer to the kind of journalistic co-production or "collaborative produsage" [10] envisaged by Bruns and others. Rambling, structurally cumbersome, stylistically messy and less technically agile or user-friendly than dedicated social networking sites, threads nonetheless provide an increasingly dynamic space for engaged and proactive audience-members to post material capable of embellishing, qualifying and/or contesting the articles to which they relate-and (by using their more interactive latter-day features) to "liking" and "rating" each other's comments in much the same ways as users of more conventional social media "share" and/or "retweet." In so doing, they have the potential to nudge the balance of power in the dynamics of news (and reality) construction a small step further away from being the preserve of privileged journalist-knowers and their (typically) elite sources, and into the hands of informed citizens as influential social actors.
As indicated previously, although traditional news organizations were relatively quick to take up forums and comment threads in the early stages of migrating online, adopting them well before they began experimenting with more sophisticated forms of UGC, research suggests user comments were introduced with little enthusiasm-and scant regard for any meaningful editorial contributions they might make. A succession of ethnographic studies stretching back to the early 2000s testify to a widespread normative aversion among news professionals to giving audience-members too much of a say on their websites. These were based on everything from proprietorial concerns that doing so might erode their privileged gatekeeper roles (e.g., [20,21]) to fears that threads could open the flood-gates to abusive, ill-informed and/ or defamatory comments [6,7], thereby necessitating continual, time-consuming and costly, refereeing to promote civility and avoid litigation.
More recent studies have continued to highlight the wariness and indifference of many journalists (especially those trained in the analogue era) toward the value of comment threads. A revealing content analysis and newsroom ethnography carried out by Canter at two British local newspapers found levels of direct interaction between audience-members posting comments and journalists on whose work they were commenting were "virtually non-existent at 1 per cent" [22]. Nevertheless, there were signs, even here, of threads starting to serve a useful deliberative function, as well as offering audience-members opportunities to post "comments" that potentially went well beyond merely reacting to articles: by making evidence-based truth-claims affirming or contesting them. Around half of all exchanges on these websites involved two or more posters debating issues with one another, while 34 per cent of posts used asserted knowledge or experience to confirm, challenge or otherwise "interact with" news texts (if not their authors). Though more optimistic about journalists' willingness to engage with audiences, an earlier newsroom ethnography by Robinson spotlighted a marked difference of opinion about the value of UGC (including threads) between a younger generation of digital-native "convergers" and older "traditionalists" keen to preserve conventional gate-keeping hierarchies [5].
For all these undoubted tensions, however, a growing current of academic opinion has come to view threads and forums as actual or potential sites of audience empowerment-allowing them to engage in peer-to-peer debate and information-sharing and actively contribute to narratives around which their dialogue revolves. Focusing on the ways in which fellow users exchange news and views on threads, in the manner of other social media, several scholars have reconceived of them less as sites of comment than "discussion" [23] or argumentative "dialogue" [24]. More positive still was a 2008 study suggesting that, at an even earlier stage, visitors to the websites of Britain's biggest-selling national daily tabloid, The Sun, and Sweden's Aftonbladet were increasingly creating "news/informational content," with threads offering them the opportunity to "manipulate existing content" so as to change "the nature and character" of the news material they were accessing [25].
This was a highly significant finding, in that it recognized that the effect of people posting comments containing fact-based claims, rather than mere opinion, was to alter, edit or supplement the articles themselves, rather than simply respond to them-in so doing, conceivably helping to inform other audience-members who read them. As Meyer and Carey would later observe [26], such posts "represent a way to continue the conversation about important community topics" and "can help journalists see that the story does not end once it is published," by identifying "new avenues for examination and new perspectives to include." However, by far the most persuasive evidence to date of informed audience-members using threads as a way of participating in the construction of stories themselves-at times directly contesting the authority (and by extension authorship) of the journalists whose by-lines they carried-was contained in Secko et al.'s 2011 study into the ways readers with expert knowledge engaged in sophisticated levels of "narrative interaction" with an online science journal, through posts either directly "questioning…the journalist's authority" or "contradicting essential elements of the journalistic narrative" [27]. Equally significant in this case was the finding that, far from ignoring this audience input (as had Canter's refuseniks and Robinson's "traditionalists"), journalists on this publication "in no cases" considered their own articles, standing alone, as "a completed package." Instead, they would routinely "read the audience commentary after their articles," to get "a sense of what people were concerned about, and even details on breaking stories." These, then, are the concepts that form the theoretical and argumentative basis for the coming empirical sections. Firstly, it is argued that articles researched and authored by journalists are (and can only ever be) incomplete, in that they are constrained by limitations of time, space, budgets and other organizational pressures-which invariably prevent them from offering fully rounded, let alone comprehensive, accounts of news events or issues. In addition, the authority and accuracy of journalists' articles depends on their gaining access to trusted knowers: individuals, organizations and other sources equipped with the information and expertise they need to make sense of (often contested and unfamiliar) subject matter and translate this material into terms that are understandable and meaningful to the public. In circumstances when there are questions about the credibility or suitability of sources, or whether a sufficient range of knowers has been consulted, moreover, articles can be said to be even less finished than otherwise. Flowing from these issues, it is further argued that one way in which journalist-authored narratives can be made more complete is to incorporate the testimony and background knowledge of (informed) lay knowers: in essence, by reconceiving of user "comment" threads as sites of narrative negotiation and knowledge exchange, rather than mere reaction, and users themselves less as audience-members than citizen sources and co-authors. Most importantly, the chapter moves beyond the narrow purview of Secko et al.'s study, which focused on exploring the co-authorship potential of threads on specialist websites, to build on arguments introduced by this author elsewhere that they might just as usefully be viewed as tools for finishing "unfinished" stories in the context of mainstream online news media as those aimed at niche audiences [9].

Comments and conversation as co-authorship: constructing case studies
The sources of primary data on which this chapter's analysis is based are supplementary findings drawn from two recent studies by this author into the role played by comment threads in amplifying and consolidating UK newspaper narratives portraying low-income children, parents and/or families as problematic. The main body of data is extracted from threads analyzed for a 2016 conference paper into the ways in which press reports about children and parenting, and the comments accompanying them online, tend to distinguish between "good" and "bad" families, communities and neighborhoods [28]. The overall dataset constructed for this study consisted of all articles about children and/or families published in a cross-section of British national newspapers over 10 weekdays between 1 January and 4 March 2016 inclusive, as well as the online comments they generated. Three titles-a broadsheet, mid-market tabloid and red-top-were sampled on each date, beginning on the first Friday in January, followed by the second Thursday, third Wednesday, etc. This sequence was reversed from the start of February (i.e., first Monday, second Tuesday and so on). In addition, different sets of papers were sampled over each of the two periods: The Guardian, Daily Mail and The Sun in January and Daily Telegraph, Daily Express and Daily Mirror during February and early March. The threads selected for qualitative analysis here are those published in response to three of the most widely reported (and debated) stories in this dataset: sites of UGC that offer especially clear and rich illustrations of the range of narrative and discursive possibilities presented by comment posts.
The additional findings derive from research carried out for a forthcoming monograph focusing on press and popular discourses that frame households experiencing poverty as relatively more or less "deserving" of public sympathy and social assistance [29]. For this project, six datasets of articles were assembled based on a series of Lexis Library database searches of all UK national and regional newspaper articles using the terms "benefits," "welfare," "unemployed," "dole," "claimant" and "poverty" during 2016. In-depth framing analysis was reserved for a selection of six key stories that generated both some of the most widespread newspaper coverage and lengthiest discussions on comment threads. The threads explored in this chapter are those that were published in response to one of the most widely reported of these six stories: the tale of an unemployed couple who had been denied social assistance (welfare benefits) after spending a £50,000 cash prize.
Of the numerous user posts generated on the threads explored here, our analysis focuses on those that contributed additional detail, background or context based on claimed eyewitness testimony, other forms of personal experience and/or asserted expert/insider knowledge. The most intriguing and impressive of these elevated themselves well beyond the status of mere reactive opinion, by making truth-claims that were arguably of equivalent (if not greater) news value to those included in articles on which they "commented"-often backing up these claims with strong supporting evidence. Though high-quality evidence-based posts of this kind accounted for only a small minority of the total of all posts analyzed for the two studies, a number were identified in each case, with the most extensive threads-often those published on Mail Online, the world's most visited news site [30]-at times opening up into energetic debates between fellow posters about disputed "facts" and details.
The following analysis centers on four discrete samples of evidence-based posts that emerged from the inter-user dialogue generated by these threads. The samples are presented in the form of conversational snapshots concerned with aspects of the stories concerned that provoked high levels of debate and, at times, dissent. Each sample is presented as a case study, accompanied by explanatory context illustrating how the nature of the post(s) generated introduced additional material pertinent to the articles over and above that included by the journalists. This evidence had the effect of strengthening, qualifying or contesting the framing and/or substance of the original texts. It is argued that, by contributing to journalistic texts in such ways-and modifying, amending or even "correcting" them-audience-members participating in these conversations were engaged in a process of factual and discursive negotiation that saw them, at times, compete with the journalists for overall authorship of the narratives. In this respect, their agency in the communication process was closer to co-authorship than any form of simple news reception.

Case study 1: using personal testimony to affirm/reinforce a story's framing
The most common form of evidence-based comment to be found on the threads sampled across the two studies were those that saw audience-members transform themselves into additional news sources for journalists' stories-by contributing personal testimony that affirmed, and therefore consolidated, the way articles were framed. A vivid illustration of this process in action was the succession of posters effectively crowd-sourcing themselves in support of a 19 January 2016 Daily Mail story about the UK's high rate of stillbirths, which opened with an alarmist introduction blaming "the needless deaths of 720 babies a year" on "Britain's failure to provide proper care for pregnant women" [31]. The value of such posts was particularly marked in this case, given that the report itself largely relied on truth-claims made by professional and/or official sources, rather than women themselves: principally the respected medical journal The Lancet (which had published the study it cited) and a mix of other academics, leading clinicians and spokespeople for stillbirth charities.
Several women responding to the story affirmed its angle in the loosest sense, by focusing on the alleged shortcomings of antenatal and maternity services they had experienced personally. "Jjj20, London, United Kingdom," a self-styled "first time mum," decried "the lack of information and advice" from her midwife, remarking that she had "not yet seen the same midwife twice," and "wojdy, manchester" relayed how, despite her placenta "showing signs of packing up" and advice from her doctors that she "needed to be induced asap," her hospital had no beds available and her local maternity unit was "closed." Others went further, however, to reflect specifically either on their own experiences of suffering stillbirths or how they or someone close to them had narrowly avoided one, despite having also been let down by professionals. While "Geordie Lass 71, Newcastle Upon Tyne" confined herself to recalling her child's stillbirth as "the most heartbreaking thing that has ever happened in my family," "Sarahmum, Hull" angrily rebutted several posts highlighting lifestyle factors, like smoking and obesity, as bigger causes of stillbirth than poor healthcare, by urging fellow readers not to "blame mums." Describing herself as a "slim" non-smoker, she recounted how she had lost her own baby 28 hours after the birth because her midwives "could not interpret" the signs of "severe distress" on her fetal heartbeat monitor (CTG).
The most effective evidence-based posts, however, were those containing anecdotes that could be verified by referring to authoritative third-party online sources-even if their posters rarely provided these details or signposts themselves. A commenter using the alias "My point not yours, wales" described how, when her friend's midwife ignored the "severe itching" she complained about during late pregnancy, she referred herself to "a different midwife" who recommended that her baby be induced "three weeks before term" to avoid a stillbirth. The potential link between itching and stillbirths is confirmed by the official website of Britain's National Health Service (NHS). Although it states that mild itching is common in pregnancy, it advises that persistent itches can be a sign of the liver condition Intrahepatic cholestasis of pregnancy (ICP), which can increase the risk of stillbirth [32].
By way of balancing the picture, it is worth noting that the verifiability of truth-claims also applied to more contentious posts-including those that generated debate and disagreement by criticizing mothers themselves for the high rate of stillbirths, rather than poor and/ or underfunded health services (in line with the way the same story was framed by at least one other outlet: the Daily Express [33]). Mail reader "VerySeriousPerson, Peebles" suggested provocatively that the main risk was the fact that "the UK has the fattest mothers with many health problems." His/her claim for the obesity level itself is broadly supported by readily searchable academic studies (e.g., [34]), while the association between overweight mothers and increased stillbirth risk is backed by the US National Library of Medicine [35].
In sum, then, evidence-based posts generated by this story contributed valuable additional testimony and factual context that was missing from the original report. When combined with the more generalised account given by the journalist, the (often detailed) anecdotes provided by women who claimed to have experienced inadequate antenatal care and/or the trauma of stillbirth added substance and legitimacy to the overall journalistic text: that is, the original journalist-led article taken together with its accompanying comment thread. Moreover, some posts introduced truth-claims that went beyond merely volunteering experience-based testimony: affirming the story's substance while also adding human-interest value that potentially made the story more involving for other readers. The most significant were posts containing additional scientific/medical information that introduced substantive context which could easily be independently checked.

Case study 2: using personal observation/experience to widen scope of story
Another way in which evidence-based posts can help consolidate the framing of an article is by contributing testimony based on background knowledge and/or experience that widens the purview of the original story-giving it added legitimacy by demonstrating that a problem or issue it identifies is more widespread and/or serious than might be apparent from the report itself. By way of illustration, a 16 February 2016 Mail Online article, headlined "Primary schools are forced to tell parents to stop smoking CANNABIS and using foul language as they take their children to school," focused on two schools-one in Greater Manchester, the other in Devon-but dialogue on its accompanying comment thread prompted a number of posters to regale fellow readers with anecdotal tales based on alleged similar antics at those in their own neighbourhoods. These included "sayitwithasmile, Isle of Wight," who claimed such behavior was "happening nationwide," based on his/her experience of working at a school where "the 'F' word" was "in most of the sentences spoken among a large percentage of the students," and "Ternet, Edinburgh," who had seen a notice "regarding foul language" posted on the door of a local nursery school. "Bev Burrows, Hull" described "similar issues" at her local primary, ranging from parents turning up at the gates in their "pyjamas" (a behavior complained about at one of the schools named in the article) to "foul language AND cannabis abuse by parents waiting for their kids, both before and after school." She was also one of several posters to assert that such problems were typical of (if not endemic to) the Manchester area, informing fellow readers that she lived "in a village on the outskirts" of that city, but was "desperately" trying to "move away" to escape the conduct of "long term residents." Another, going by the curious nom-de-plume "I MUST SAY," claimed that, while working as a "learning mentor" at a different Manchester school, she had had "to deter parents from swearing and FIGHTING"-often over "what was said on Facebook the night before." A further notable feature of inter-user discourse in relation to this story was the frequency with which people combined their disdainful comments about the parents concerned, and/or anecdotes about similar behavior they had witnessed, with sweeping generalizations about their assumed social class and/or income status-often explicitly stigmatizing them as welfare claimants. Though many such comments were unevidenced, they are relevant here in that they were sometimes answered by evidence-based counter-claims challenging their prejudices. One notable post of this kind widened the story's scope much further than those complaining of identikit behavior in unsavory areas of their own towns and cities, by arguing that uncouth parents were not confined to working-class schools. Dismissing the notion that antisocial behavior was unique to "benefit families," "NorfolkBroad, Norfolk" claimed to "recall teaching in an Army school," where she told a mother about her "concerns regarding a child's violent attitude and foul language"-only to watch as she "turned to the child, slapped him around the head" and swore at him.
In some ways, then, scope-widening evidence-based posts can be problematic. The numerous posts shared by audience-members within the comforting echo-chamber of a Mail Online comment thread, in which they shared superficially similar anecdotes affirming the story's "problem parent" narrative, bear all the hallmarks of what Iyengar calls "episodic framing": a de-contextualized litany of antisocial antics that simply reinforce lazy social stereotypes [36]. In other respects, though, they offer the potential to be used for quite the opposite end: to add more varied, at times surprising, observations and examples that can contribute toward a more rounded and representative "thematic framing" approach [36].

Case study 3: using firsthand knowledge to add detail specific to the story
Perhaps the most exceptional form of co-authorship to appear on threads accompanying dayto-day news reports is a strain of evidence-based post purporting to contain additional details, context and/or testimony that directly pertains to the news event, incident or individual(s) that are the subject of the story on which it is "commenting." While crowd-sourced eyewitness testimony and/or expert observations have been normalized during recent coverage of major incidents, from the uprisings of the Arab Spring to terrorist attacks and extreme weather events, it is much less usual to come across user-generated content of this kind in relation to less high-profile, spectacular and/or breaking stories.
The datasets from which these case studies were drawn, however, did include a handful of (more mundane/routine) stories that produced evidence-based posts of this kind. Of these, by far the most commented on was the story of an unemployed couple from Guernsey, a self-governing UK dependency, who attracted widespread national media attention when their local paper, the Guernsey Press, reported that they had been refused social assistance after spending a £50,000 cash prize they had won in a Christmas Lottery. Though evidencebased posts only accounted for a tiny minority of the 4400 comments their story generated on the Mail thread alone, some of the longest of these introduced additional details about the couple's background absent from the journalist's report. Others, meanwhile, emphasized and developed points that the article itself had downplayed.
Given the nature of this story, and its widespread framing as the tale of two irresponsible and entitled young people-the Mail's headline alone read "Jobless lotto couple who won £50,000 and then spent it all in eight months face eviction after being refused benefits" [37]-it is perhaps unsurprising that many comments were negative and highly judgmental. Beyond those that were merely reactive, however, a number contributed details and observations that might well have added to any feelings of derision and disgust the report provoked in their fellow readers. A background fact included in the Mail story, but amplified and expanded by posts from readers with local knowledge of the couple's history, was the fact that, as a younger man, the husband had been convicted and gaoled for an unsavory assault on a woman, during which he had spat in her mouth while infected with Hepatitis C. Though the Mail wove key details of this conviction into the second half of its report, at least two posters copied and pasted whole chunks of their local paper's original account of the court case into their posts, adding details such as the name of the judge and the fact that an earlier offense by the defendant, in which he was caught "brandishing an axe and a hammer," took place during a meeting he arranged in a car park [38].
More significantly, several posts on the threads of both the Mail and Daily Mirror were used to inform other readers of sensitive personal details omitted from the news articles themselves, but which arguably had a bearing on key comments attributed to the couple in the reports. To illustrate, the Mail's story described the couple as having two children and quoted a Facebook post in which the wife had responded to criticisms from the public that the lottery money should have been spent on them by retorting that her "kids have everything" and were "very spoilt" [37]. However, a succession of posts from audience-members claiming to have local knowledge of their situation qualified these remarks by asserting that the couple's children were no longer living with them, as they had been taken into care by child protection services. Referring to the husband's convictions, a poster going by the geographically revealing alias "Guernsey mum, Guernsey, Guernsey" remarked that he had "done plenty and been done for it" and that his "kids have been in care for a while because of it," before posting a further comment querying why his wife was "playing the kid card," given that "their two sons are in someone else's care." This refrain was echoed by Mirror poster "GSYLady E2," who confided in fellow readers, "as far as I'm aware (from word of the mouth in Guernsey with it being small) they don't even have their children," before adding this incisive (if brief) critique on the unfinished nature of journalist-led stories: "you never get the full stories on these websites." Other sensitive, and potentially contentious, details added by posters included references to Facebook posts by the wife (unmentioned in the news coverage) that some interpreted as indications that she and her husband were recovering drug addicts. Reacting to comments on social media accusing them of wasting their lottery winnings on drugs (allegations repeated by a number of those contributing to the newspaper comment threads), Mail reader "blah blah, manchester" quoted her retort to critics on Facebook that she had "spent nothing on drugs coz we're on 'script'" as an admission that "she hasn't spent it on drugs coz they get it free on prescription!" Likewise, Metro contributor "Carrie Little" encouraged others to "read her Facebook"-which she said explained why the couple "do not work" and also proved that they were "smackheads." Besides adding intriguing, if demeaning, details about the story's subject(s), evidence-based responses also introduced valuable context about Guernsey itself: in particular, the island's disproportionately high property prices and general costs of living. This information was relevant because the premise of the story-the family's impending eviction following their failed social assistance claim-was predicated on the fact that, despite having managed to pay the initial deposit on their new flat, they had insufficient income and savings available to cover the cost of their rent and other essentials, like food and heating. By way of example, one poster to the Metro's thread, "Thomas Foxen," informed his fellow readers that "the Guernsey housing market averages above London's for both purchase and rent sadly" [39]. This is a claim which independent data shows to be broadly historically accurate, if not at that precise point in time [39]. However, while another poster with insider knowledge, "WizzleTeats E1," agreed that it was "hard going keeping a good job and home" on the island, based on the experience of her "family and friends on Guernsey," he/she claimed to know "a few who have at least a few children to their families and live off around £26,000 a year and they manage ok!" The figure of £26,000 added a layer of context that will have been significant for readers at the time (in the UK, if not Guernsey), as it was the level of a then new cap on the maximum amount individual working-age households in Britain were eligible to receive in welfare benefits [40]. A more banal (if persuasive) defense of the couple's financial predicament was offered by Mirror reader "Jonathan Sebire" in response to a dismissive post by "Pash Pash" challenging them to "join the rest of the working class and earn a real wage" by applying for jobs at McDonalds, who were "hiring-with free meals." His confident, drily worded retort simply stated that there was "no McDonald on the rock," because the "foot fall needed to sustain a store" was "more than the population" of the island.
It is worth briefly noting that evidence-based posts drawing on asserted local knowledge to confirm, embellish, qualify or contradict the detail included in articles (and/or added by other readers commenting on them) were not confined to threads responding to the Guernsey story. Among the many posters who affirmed the substance of reports about cannabis-smoking, pyjama-clad parents was "jason69, Durham," who claimed to have direct knowledge of antics at one of the schools specified in reports. "I live in Darlington," he explained, "and this poor head has been subjected to vile abuse because of her sensible comments." He added that he had "many friends" content to "spend all day smoking weed in their council-funded properties," so it was "no wonder these parents do as they please!"

Case study 4: using personal knowledge to contest inaccurate reporting
One of the most powerful uses of evidence-based audience testimony is not to affirm and reinforce the truth-claims made in an article, but to contest them. A story in the conference paper sample which provoked a particularly strong evidence-based counter-discourse from readers disputing its accuracy and the manner of its framing by journalists (and many other posters) concerned a supposed plan to adjust the dates of that year's school exams timetable to avoid it clashing with Ramadan. The story's premise rested on the fact that the Islamic festival was, unusually, due to coincide with the period when 16-year-olds sat important GCSE exams-meaning that young Muslims might be disadvantaged if their concentration suffered because they were observing the custom of daytime fasting. As a result, the Joint Council for Qualifications (JCQ), the body that represents UK exam boards, was said to be planning to reschedule exams in key subjects, including English and Mathematics, to earlier dates, which (in Mail Online's words) would give all children "fewer days to revise" [41].
Coverage of this story (spread across a wide range of newspaper titles) prompted an initial deluge of comments condemning the reported decision. However, these were met with a swift and vocal fightback, particularly on more liberal news sites-not only from people defending the merits of the proposals but evidence-based posters self-identifying as practising Muslims, teachers or other expert knowers who disputed the story's entire basis.
Perhaps the most direct and unambiguous attempt to debunk the framing of the Mail's reportprovocatively headlined "Sit exams early to fit in with Ramadan: Pupils taking GCSEs and A-levels face timetable shake-up to accommodate fasting Muslims"-was mounted by "cherriesandoranges, Doncaster," who began with the blunt analysis, "DM this is shoddy reporting," before going on to argue that, as "a GCSE English lecturer in a college which delivers GCSE to over 1000 students," he/she had known "the exam date for both Maths and English for 2014 in spring 2015"-meaning that the practical difficulties of rescheduling the nation's exams were such that "there's no way they will change the exam days now," because "it's too late." Intriguingly, however, this particular truth-claim was contradicted by "John Cantelo," another self-identifying schools expert (this time on the Huffington Post comment thread), who responded to a previous poster's tirade against a decision to "screw over all the other students because of the muslim [sic] students' religious problems," by arguing that, "as a former secondary school examinations secretary," he doubted that "examination boards have yet fully firmed up the precise timetable for summer 2016 (let along published it for the students)" and that this was probably "why this is being debated now." Importantly, though, he was one of several posters to criticize the way the story had been presented-i.e., as a proposal that would negatively impact non-Muslim students-by arguing that, far from "being given an 'unfair advantage'" over everyone else, those observing Ramadan were merely being protected from "any potential disadvantage they may suffer from their beliefs," and that "all that is happening" was that all students would be sitting Maths and English exams in the morning, rather than the afternoon. In the event, the first of these two counter-claims would prove to be nearer the truth, based on a succession of subsequent pronouncements by the exam authorities exposing the extent of the misinformation contained in articles about the subject. These included an official statement issued over the following days by the JCQ-which pointedly identified "a clear misunderstanding in some parts of the media as to how the GCSE and A level timetable is set and the impact religious events, such as Ramadan, Easter and Passover, have on it" [42]-and a public letter from Glenys Stacey, chief executive of the UK exams regulator Ofqual, to then Chief Inspector of Schools Sir Michael Wilshaw, reiterating that "the [exam] timetable for 2016 was drafted over a year ago, is published, and won't be changing" [43]. Nonetheless, for all its comparative limitations, "John Cantelo's" counterclaim still offered evidence-based testimony of real material value, if only because it directed his fellow readers to pay closer attention to (underplayed) details in the report to which they were responding: namely that, even if some timetabling adjustments were made, these were only likely to involve bringing forward some exam sittings from one end of a particular day to the other.
Perhaps the most unified front against misleading impressions conveyed by articles about this story, though, was presented by posters who self-identified as Muslims. Many of these commenters went beyond disputing details of whether, or how, exam timetables might be changed to accommodate Ramadan, by instead rejecting the entire premise of the story, i.e., any suggestion that Islamic religious-leaders regarded observing Ramadan as somehow incompatible with sitting exams. One Mail poster, "pray for peace, forever-london," said he/she lived in "the Middle East" and had Muslim children, but that exams times had "never changed due to Ramadan"-despite the fact the region was "far hotter than in the UK." Meanwhile, "AN333, South Yorkshire" claimed to have "done exams during Ramadhan whilst fasting" and "Peru123456789, London" asserted, "I'm Muslim and going to do my GCSE exam," adding that he/she was "fine with fasting and sitting exams." But perhaps the most significant strand of truth-claim mounted in defense of the Muslim community was that emphasizing the fact that there was scope within both interpretations of the Quran itself and the official advice issued by Britain's leading Islamic authorities for exam-aged children to be exempted from fasting. While "Mother of Two, At home" bluntly told her fellow Mail readers that "Muslim children are exempt from fasting," "slight32" contextualized this claim by adding important further detail for those on the Independent thread: notably the explanation that there was "no requirement in Islam for children to fast," and that "those choosing to" could either delay it until after their exams or instead observe "an alternate form of piety," such as "charitable works." Once again, these confident and fulsome posts added genuine material value to the journalist-led narratives-at least for anyone prepared to take the time to check them out-because the truth-claims they made could easily be verified by referring to authoritative third-party online sources. The websites of both the Muslim Council of Britain (quoted in several articles welcoming the supposed rescheduling plans) and the Association of School and College Leaders stressed "the flexibility Islamic Law offers" children to "delay or exempt themselves from fasting and late night prayers if they believe their performance in exams could be affected"-adding that "Islam encourages critical reasoning" and individuals' rights "to make their own decision" [44].
Oppositional evidence-based posts contesting the way journalists and/or other audiencemembers had framed a narrative also surfaced on threads accompanying reports about the "problem parent" story. In an impassioned and personalised riposte to the anti-Manchester, anti-welfare sentiment expressed by many Mail posters, "Mrsxsx, Manchester" railed against the "middle class snobs that read and write articles such as this," arguing that "not all teenage parents or sink estate kids are like this." In the course of two lengthy posts detailing her own background as someone from a low-income, working-class family, she told fellow readers she had been "brought up on a council estate, had a child at 17, married his father, who was also brought up on the same estate," but had gone on to "own a home, 2 new cars, and earn in excess of £80k per year between us" at the age of 30-thanks to "nothing but hard work" and "a determination to prove snobs such as yourself wrong."

Problems with evidence-based posts: contentious or inaccurate truth-claims
As with any argument over truth-claims-including those advanced by journalists themselves, and by the competing "he said/she said" sources on which they often rely [45]-some evidence-based claims inevitably proved less credible and/or verifiable than others. Responding to the Mail's take on the stillbirth story, "Michael Haymar, Oxford" drew attention to the assertion, "not mentioned in the article," that the annual flu vaccines offered to some adults in Britain were "a cause of stillbirths." However, a glance at the research literature shows this to be a highly disputed link, with some studies suggesting that these jabs can, in fact, reduce the stillbirth risk, rather than increasing it [46]. The importance of subjecting such claims to careful fact-checking was also demonstrated in this case by one of the few posts that directly cited its source. This was a statement by "amj127, Stockholm" that Icelandic mothers were "on average older than mothers in the UK": a retort to a previous poster's remark that had insinuated a connection between the age at which British women became pregnant and the fact (mentioned in the story itself) that the country's "stillborn rate" was "eight times" higher than Iceland's [31]. Although this poster cited his or her source as "Eurostat," the European Commission's statistics database, a visit to its breakdown of maternal age ranges across the 28 EU states shows that, in the then most recent year (2015), the mean age of UK first-time mothers was actually older than Iceland's [47].
Spurious evidence-based claims were far from confined to threads responding to the stillbirth story. A sweeping statement by a poster using the (misspelt) nom-de-plume "Spolit Rich Kid, Chelsea" described Manchester as the "drugs, gun and crime capital of UK" on the thread accompanying the Mail's rendition of the "problem parents" story. More verifiably untrue was the wild claim by "Fed up tax payer, london" in response to the same story that the cost of "obesity-related illnesses" (a problem he/she associated with the feckless families portrayed in these articles) now accounted for "50% of the NHS budget." Swift fact-checking against an online report by the respected free-market Institute of Economic Affairs thinktank showed the "net cost" of UK obesity to be £2.47 billion-equivalent to "0.3 per cent of the UK government's total budget in 2016 or 1.8 per cent of the NHS budget in the same year" [48].
Other evidence-based posts were problematic because they contained truth-claims that were misleading, rather than inaccurate as such. Responding to the story about the Lottery-winning couple, Mail poster "Listen_To_Me, bristol" confidently talked his fellow readers through the rules set by the UK's Department for Work and Pensions (DWP) to determine eligibility for social assistance-apparently oblivious to the fact that they were residents of Guernsey, not Britain. The criteria he set out regarding the UK system were broadly accurate: principally the fact that "you are not eligible for benefits if you have more than £16,000" in savings and that, "if the DWP believe that you have deliberately disposed of money over this limit" to make yourself eligible for benefits, they had "the right to refuse" you [49]. However, as "Anon, Anon, Jersey" pointed out elsewhere on the same thread, posters complaining about subsidising their lifestyles did not "have to worry about paying toward their rent," as "the Channel Islands aren't part of the UK for tax/social security purposes."

Conclusion: the scope and limits of evidence-based comments
As the above case studies show (however briefly), hidden among the blizzard of comment posts that are wholly or largely reactive in nature, the threads published below online news articles offer rich pockets of asserted experience, expertise and local/privileged knowledge which can enhance, embellish and, at times, authoritatively contest the way stories are framed by journalists. But for such evidence-based contributions to truly "come of age," by realizing their full narrative and discursive potential, three pre-conditions will first need to be met. Primarily, journalists (and the news organizations for which they work) must be prepared to acknowledge the value of such audience-to-journalist truth-claims and to integrate them more meaningfully into the process of news production. In short, threads should be recognized as forums offering crowd-sourced information and testimony about day-to-day stories and issues that, if properly incorporated into newsgathering and reporting routines, can be every bit as useful and important as the more dramatic user-generated footage and eyewitness accounts contributed by lay sources to coverage of rarer, more explosive, "news events" like terror attacks or natural disasters. The second precondition is that the value of evidencebased comments needs to be recognized not only by journalists but also (other) audience-members-and factored into their understanding and appreciation of the overarching narratives they are reading/viewing, and to which they are responding (whether publicly, through their own posts; in discussions with peers; or through their own private interpretations).
It is the third pre-condition, however, that is most important: that relating to the responsibilities of the audience-sources contributing the evidence-based posts. If their role is to be taken seriously, they have a duty to ensure that the truth-claims they make are honestly expressed, clearly and credibly sourced-and (as far as can be ascertained) factually accurate. As this chapter shows, evidence-based posts often contribute additional context, detail and layers of meaning with the potential to fine-tune and make more complete the journalist-led articles to which they relate, in so doing improving public understanding of their subject matter. As with news articles themselves, however, it is unwise to take posts at face value: though many may be trustworthy, some contain truth-claims that are misleading, if not downright inaccurate. What is needed, then, to make the most of the opportunities offered by threads for substantiating and finessing stories is a new form of three-way "contract" between journalists and their two levels of audience-member: people who contribute their own source material to consolidate/contest the finished narrative, and those content merely to "receive" the truthclaims relayed through this emerging hybrid of professional and citizen journalism.