{
 "blyg": "0.4",
 "id": "1at84fqbvrm4txsn6dqpvtg39m",
 "kind": "thread",
 "origin": "https://notionparallax.co.uk/",
 "page": "https://notionparallax.co.uk/2021/commendary-cultures",
 "author": {
  "name": "Ben Doherty",
  "url": "https://notionparallax.co.uk/"
 },
 "created": "2021-07-26T03:18:00Z",
 "updated": "2021-07-26T03:18:00Z",
 "version": 1,
 "content_md": "# Commendary Cultures\n\n\nSocial sharing of news has a critical economic and cultural function in an attention economy. It reveals our personal interests, beliefs and tastes to the array of publications trying to win our interest and identifies to them and our social networks what is valuable to us in the cascades of data that we negotiate every day. Through our shares, likes and favourites we help the social platforms determine what is popular and profitable in advertising terms. We also elevate certain events and issues in public debate, and reinforce opinions and attitudes in our social circles. Thus, if most mainstream media users are more likely to redistribute, like or recommend news than to create it, then news sharing is potentially as important a phenomenon to study as participatory or citizen journalism.\n\nThis makes it imperative for reporters, editors and news executives to understand, beyond the metrics they might receive about Facebook or Twitter trends, what motivates news sharing in different contexts, as well as how and why we are likely to share stories we come across online. It is also important to recognise how the platforms have ensured news sharing is imbricated in the rhythms of everyday life online and integrated into the editorial processes of news rooms. This shift is driving the reorganisation of news production, consumption and distribution, and redirecting economic, political and cultural power away from the legacy media and towards the platforms.\n\nIt’s easy to focus on social media platforms as the primary site of news sharing, as they are designed precisely for that purpose: the easy personal exchange of information with what passes for a self-built social network (although recommendation algorithms play a big role in helping us constructing that sociality). Social media companies make sharing simple to encourage us to disclose more about our everyday lives, so they can analyse that routine evidence of activities and preferences to create more effective forms of advertising. It’s a twenty-first-century design for a radical, globalised consumer sovereignty that seeks to make as much social behaviour as possible accessible to corporate scrutiny or, as Facebook’s most durable first decade mission statement more altruistically puts it, “To give people the power to share and make the world more open and connected”\n(Reagan 2009).\n\nYet, news sharing also takes place via email, apps, forums and other online channels, just through the posting of hyperlinks—a practice as old as the World Wide Web, and which has its basis in earlier hypertext systems. Each time we post a link to a news story, we are curating from our own feeds what might be of interest to others, recommending what they should attend to and demonstrating something of our own concerns and biases. Link sharing is not always an endorsement of the content, but demands those who click through consider its meaning and worth. So, this chapter explores how social news sharing has become central to the creation of commendary cultures online, where ordinary people take on the representational and distributive roles of the news media, as well as something of journalism’s filtering and evaluative functions. It explores the pre-history and evolution of news sharing technologies and the emergence of different ideas about what sharing means, why we do it and how it’s affecting digital journalism.\n\nIf link sharing is the basis for online news exchange, how then did social media become synonymous with this practice, to the extent that people around the globe now see these channels as key news sources? To answer this question, we first need to establish what qualifies as ‘social media’. As new media scholar Graham Meikle notes this term evokes “a particular set of technological affordances, a particular set of business models and corporate practices, a particular set of organisations, and a particular set of cultural habits, practices and expectations” (2016: x). The organisations are well recognised worldwide and becoming virtual monopolies in their various niches by virtue of their transnational scale and market dominance. By late 2016, Alphabet (owner of Google and YouTube), Facebook (with Whatsapp and Instagram) and Microsoft (LinkedIn and Yammer) were among the largest companies internationally by market capitalisation, bigger than the finance and energy companies they had succeeded on that list (Flew et al. 2018). Facebook is now the world’s most popular social media platform, with around 2.19 billion monthly active users in the first quarter of 2018. In the same period, WeChat has reached 1 billion Chinese users alone, more than Twitter has worldwide, while YouTube dominates video sharing and video on demand advertising globally.\n\nWhile each company has a distinct business model, on the whole, the biggest players’ economic success has been based in five factors:\n\n1. broadening the base of production to every subscriber;\n2. making global communication and publishing easier through templating and automation;\n3. reducing or removing the geographic boundaries to social and business transactions;\n4. aggregating and analysing user created data to create more efficient, individually targeted advertising sales models; and 5. operating as international information technology companies rather than national media publishers, which has until recently put them outside traditional media content regulations and allowed them to publish content that others could or would not.\n\nEconomist Yochai Benkler (2004) suggests internetworked social sharing of information has proved a more efficient form of distribution than historic market forces. It is decentralised, avoiding the strictures, conventions and routines of corporate news gatekeepers; autonomous, allowing anyone and everyone with network access to take part, and personalised, in tying the act of recirculating news to communicative rituals and social conventions of information gifting and reciprocation.\n\nSocial media companies’ technological affordances have enabled us to increase our everyday news sharing, extending it to large, curated social networks. Starting with Six Degrees, launched back in 1997, and then Friendster in 2002, social networking applications allowed users to create public or semi-public personal profiles, to build a list of social connections that they could easily navigate, and to communicate publicly or privately with those people using templated publishing tools (Hinton and Hjorth 2013). Later social media technologies also enabled users to create, aggregate, evaluate and share media, to create groups and audiences for this content, and to monitor user analytics based on that service interaction. From 2006, Facebook and others provided widgets or limited functionality scripts for publishers to embed in their sites so users could subscribe to news publishers’ social media channels, or share a story link directly on reading it on a branded website, and later, mobile apps for continuous access to this sharing capacity.\n\nAll of these strategies were designed to simultaneously extend our customised communicative reach and to automatically co-opt more users and publishers into the orbit of platform analysis, in what critical media studies scholar Jose van Dijck and Poell (2013: 8) calls a “double logic” of connectivity. Platform analytics are not only “perpetually operational” and so pervasive that we don’t notice them (as [Kennedy 2016: 25] suggests), but are also designed to reward and reinforce our participation; to make it addictive using a series of behaviour reinforcing communication strategies (Alter 2017). Alerts or notifications from our social media feeds tell us how much attention our posts are getting, and detail our network members’ activities, to encourage our reciprocity and to routinise our sharing activity. Sometimes, gamification is built into the systems, so that we are notified where in a hierarchy of users we are placed as a result of our activity. In these ways, platforms cultivate habits of documentation, declaration, discussion and debate, through constant visual and auditory reminders of their presence. They recommend related posts and stories we should like, given our past consumption and interaction history, and make others invisible, through algorithmic processes that maximise our exposure to popular content, personalised advertising and online services. So, while social media are not the simplest or only ways to share news stories, they do provide a widely accessible suite of portable tools for customising our news consumption, accelerating the scale and scope of its distribution and exposing it to the scrutiny of algorithmic tracking, classification and assessment which allows them to monetise any insights into our behaviour and interests.\n\nIn positioning themselves as IT companies and carriers of content rather than publishers, social media companies historically have avoided having to conform to the content regulation laws and accountability measures set up for legacy media, or to worry overly about what their users post on their platforms. For example, the U.S. Digital Millennium Copyright Act includes a safe harbour provision, which protects them from liability for the content their users might post—as long as they act to remove any illegal content about which they are informed (Flew et al. 2018). There is, as yet, no legal mechanism for forcing platforms to reveal how their algorithms structure the content we see, and that we don’t see, or what human editorial decision making takes place in content management. Recent national and regional moves to regulate the content we share, like Germany’s Network Enforcement Act (German Law Archive 2017) and the European Commission’s (2017) Code of Conduct on Countering Illegal Hate Speech have put some pressure on the major platforms to remove violent and extreme political speech (Goggin et al. 2017). However, on the whole, what we share online is determined by the guidelines that they set, and which they are not answerable for—even to us.\n\nFrom this relatively unassailable legal position, and in possession of audience knowledge that the old media could only dream of, major social media platforms have been in a position to influence both cultural production and the production of culture, especially news media culture.\n\n## The Importance of Commendary Culture\n\nAs a source of traffic to news websites, social media’s direct influence on news distribution may well have peaked and passed. Certainly its 2015 dominance of referred news traffic was a blip on the metrics radar. Since then, Google search has reasserted its place as the number one online news referral source. Following the 2016 election misinformation scandal, Facebook also demoted news page content in people’s feeds (Facebook 2018), leading some industry observers to argue that publishers should reorient their strategy away from that platform (Filloux 2018) and engage more deeply with their audiences rather than using social media channels to push out links to stories (Elizabeth 2017).\n\nYet, some media commentators still argue that social media sharing may have profound effects on the future of journalism. They note that young people’s turn to social media as a primary news source—and social media companies’ capacity to imperceptibly select, filter and promote the news that gets seen—may have three immediate impacts: eroding journalism’s power to connect directly with audiences, to deliver what they need to know in the public interest and to ensure the accuracy and quality of the news they consume (Bell 2016; Bell and Taylor 2017). They worry that there is little platform accountability for what news gets seen, and by whom, and how much revenue is shared as a result (Newman 2018). Some see journalists as being locked into reporting what gets likes and shares, tailoring the news to social media metrics and “dancing to the popular music of consumer driven logic” (Tandoc 2014: 572). Others, like digital media theorist Jodi Dean (2005, 2014), have positioned social media users as unpaid labourers in the platforms’ bid for informational dominance, questioning whether their circulation of ideas can do much to affect the democratic process (as well-informed citizens might hope to), or is attenuated instead in technological fetishism. From this perspective, news sharing may be just so much symbolic chaff, spread in lieu of political engagement or action. Christian Fuchs has called out Facebook CEO Mark Zuckerberg’s claim that “the goal of the company is to help people share more in order to make the world more open and help promote understanding between people”, as a contradictory ruse that belies a ruthless commercial business model (Fuchs 2018: 172).\n\nThe more hopeful call social media news sharing a new form of media democracy and creativity, enabling a greater diversity of people to signal to newsmakers and our social networks what’s truly significant to them, in both local contexts and globally. Ostensibly, social signalling should guide journalists to report on what we value most, and on more varied conceptions of that value. In an ideal world, that would make journalism more socially responsive and culturally diverse. This techno-optimist shift is characterised in Henry Jenkins (1992) notion of participatory culture, and his newer concept of “spreadability”, where “new tools have proliferated the means by which people can circulate material, word of mouth recommendations and the sharing of media content” and facilitated the diffusion of cultural artefacts and practices (Jenkins et al. 2013: 2–3). It is also embraced in Bruns’ (2005, 2018) idea of “gatewatching”, where ordinary users publicise news, rather than reporting or publishing it. Bruns sees news sharing as a demotic, habitual practice for most people, and core to the emergence of new forms of expertise. But while these claims may have some basis, we need to know far more about the political economics and cultural forms of shared news, as well as the real-time network dynamics and contexts for its production and consumption, before we can understand its ramifications for the news media or its role in our societies.\n\nUp until now citizen journalism has had far more attention in media research than sharing. Most book-length studies of participatory news cultures to date have focused on novel forms of news creation: blogging, produsage and gatewatching (Bruns 2005, 2008, 2018), user-generated content (Singer et al. 2011), citizen journalism (Allan and Thorsen 2009; Miller 2014), citizen witnessing (Allan 2013) and maker and hacker culture (Usher 2016), rather than the new modes of distribution such as aggregation and sharing. Alfred Hermida’s #TellEveryone: why we share and why it matters, a lively journalism studies account of how everyday social news sharing works, concentrates on the sociology of sharing, leaving the broader questions of cultural change and platform power, their shape and impacts largely unaddressed. In Axel Bruns’ more theoretically engaged _Gatewatching and News Curation_, he devotes considerable energy to analysing news sharing, but tends to focus on the relationships between users, journalists and platforms, rather than the more complex industrial and regulatory relationships emerging in the social media news ecology. While these books are important interventions in social media research, the study of news sharing remains a relatively understudied transformation of cultural production, and one whose complex power dynamics remain obscure.\n\nRather than taking one or other polar position on how our social media engagement will affect the future of news, we adopt a critical media perspective on news sharing’s meanings for journalism in society. That is, we want to examine how our online news sharing has been commodified, what role our actions, beliefs and proclivities have played in that process and what impact it will have on the institutional notion of journalism and the new cultures of digital news consumption.\n\nThe rest of this chapter examines the rich meanings of sharing, and how they have shifted to accommodate our new socially networked interactions and the commodification of that sociality. It then investigates why we share news online and how, exploring how our motivations for promoting certain types of news and the nature of our interactions is reshaping the way news generally is selected, reported and distributed, and in different ways on different platforms. In these senses, it is an account of the cultural politics of news sharing, or the ways that our cultural practices affect the exercise of news media power, and vice versa. As Angus and Jhally (1989) argued, it is important to understand “the manner in which institutional and ideological structures act as limits to the possibilities of cultural practices” and how cultural activity supports the emergence of new forms of power such as those of the transnational social media platforms.\n\nSo, instead of focusing on people’s impulse to create media online or to participate in public debate, as most participatory journalism studies have done to date, we are exploring the rise of a more everyday use of internet technologies—to commend, recommend or condemn the news. In the first instance, news sharing seeks to draw people’s attention to a news item by extending its reach. Sharers often tell others what they think about their shared item to rationalise the reposting, and seek or expect some sort of response from their social networks. They may hope to engage others in discussing or debating or acting on the judgement they have expressed.\n\nThe term _commendary culture_, used here to describe the rituals, beliefs, norms and practices associated with sharing pre-existing messages online, is drawn from the two fourteenth-century Latin meanings of commendare: to entrust someone or something to the care of others, and to commit to writing. These days, the acts of commending and recommending have the same positive connotation of praising the object in question. Those stronger positive associations have developed in contradistinction to the concept of condemnation, criticism or censure, yet to commend something originally meant simply to note it as worthy of attention. The study of commendary culture then, is concerned with all acts of elevating media to the attention of others, regardless of the conclusion drawn about that material. It understands news sharing as a social and cultural practice which seeks to amplify messages, to capture the attention of interested others, and to engage them in a reciprocal evaluative process. It is also interested in the socio-technical systems that allow this sharing of symbolic goods, preferences and opinions, and the means of making sense of these acts—\nthe techniques of measuring and analysing commendary activities.\n\nAs the title of this books suggests, the commendation of news stories to our social networks is commonly described as news ‘sharing’. Sharing is such a polysemic and widely used term in online contexts that it demands some deconstruction before we look at why it is such a compelling activity.\n\n## What Is Sharing?\n\nAs Nicholas John (2017) reveals in his absorbing cultural history of sharing, the phrase ‘to share’ has three different root meanings: to distribute or divide, to have something in common with and to communicate one’s personal experience or feelings. Like marketing scholar Russell Belk, John is intrigued by the association of sharing with economic behaviours, and particularly prosocial internet use.\n\nBelk (2010), from an anthropological perspective, originally argued that sharing is a form of consumption within practices of commodity exchange and gifting, based in mutual association, shared possession or collaborative ownership, and caring relationships with others. For this perspective in sharing news online, we are not losing our share of anything, but expanding what he calls our sense of an ‘extended self’, the projection of our identity in the world through external means (p. 724).1 However,\n\n1 See also Belk (1988) Possessions and the Extended Self.\n\nhe saw sharing as a non-commodity relationship, which it clearly is not in the contemporary sense, when social media companies reap both advertising data and attention from our information sharing activities. Thus, while Belk later modified his definition of sharing to include industrial models of collaborative consumption and short-term rental (Belk 2014), John’s work is more relevant to our purposes. He sets out to explore how tech corporations have appropriated the idea of sharing to fuel the expansion of digital capitalism via the ‘sharing economy’—a key concept enabling social media companies to appropriate our personal data, our social connections and our interests.\n\nJohn (2017: 62–64) proposes that sharing has acquired new meanings in the age of social networking, associated with the reach, scale, morality and emotional intimacy of online communication. The first has come via the new organisational techniques of early computing, including practices like time sharing and file sharing, which point to sharing’s networked or distributive potential. Wadbring and Ödmark (2016) argue that there is a difference between the news media’s deliberate distribution strategies, which are designed to attract, hold and measure audiences, and social sharing’s ‘circulatory’, participative logics. We disagree with that assessment. As we will argue, platforms have their own deliberate, quantified distribution logics which seek to reach as many as possible with the most popular content their algorithms identify.\n\nThe second meaning for news sharing is reproductive, as online messaging collapses sharing’s two primary meanings of communication and distribution. In sharing news, we are simultaneously engaged in an interpersonal act of telling others about ourselves and publishing that information for an audience to read. Sharing in these reproductive senses is also enabled by what Meikle (2016: 26) calls the “intrinsic” capacity of computational technologies to produce perfect copies, which means we can give information without losing it and in a non-rivalrous manner, as the original is still around for others to consume. Despite the culture industries’ ongoing legal proceedings to assert their digital rights in music and film file sharing (Borschke 2017) and in news aggregation (Martin 2015), the capacity to make and exchange digital copies has established online media sharing as a commonplace, even virtuous act—especially where it challenges the constrictive licencing regimes of media and publishing conglomerates. Meikle quotes programmer and former Reddit co-owner Aaron Swartz arguing that “sharing isn’t immoral—it’s a moral imperative. Only those blinded by greed would refuse to let a friend make a copy” (Swartz 2015).\n\nThis _reproductive potential_ underpins the third and most potent reason for sharing’s expropriation by Silicon Valley—what John calls “its positive connotations of equality, selflessness and giving” (John 2017: 63). This beneficial sense of online sharing, he notes, accords neatly with origin mythologies of the internet as a democratic force, a network for free, transparent and non-hierarchical exchange of information. So, news sharing has an _egalitarian potential_ that makes us feel good about commending coverage of events and ideas to others.\n\nFourthly, a factor that John briefly alludes to and Meikle develops, is sharing as intimacy, and immediate personal exchange. We most often share online with our friends, family and workmates, people with whom we have some close, emotional or intellectual connections, and these bonds influence what we share and why we share it in networked circumstances. We share to build, affirm, maintain or repair relationships (Goh et al. 2017; Meikle 2016: 26). As we will see in Chap. 6, the users of an online news service who we surveyed primarily share news to show they care about a topic. They do it to express their feelings about a subject and sometimes to provoke a reaction from their friends or followers. This _affective potential_ of sharing is key to understanding the topics people will respond to, and what types of stories might go viral on a social platform (Fig. 2.1).\n\n<figure>\n\n![]()\n\n<figcaption>\n\nFig. 2.1 The semantic associations of social media sharing. Image: Fiona R Martin\n\n</figcaption>\n<figure>\n\nOver time, social media companies have recognised how to capitalise on these semantic potentials, with Facebook in particular promoting our information sharing as a “mechanism for improving human relations and making the world a better place” (John 2017: 65–66). As John critically observes, they also tend to use the term sharing to underplay the commodity exchanges they have with third-party advertisers or developers, with whom they also ‘share’ our personal information. This appropriation of the positive meanings for sharing to legitimate purely commercial and wholly privatised trade activity is meant to divert user attention from the potential privacy implications of this activity.\n\nThe corporate intention is to promote sharing as collectively oriented, authentic and socially productive. Reddit’s 2018 homepage states it is “home to thousands of communities, endless conversation, and authentic human connection”. Twitter too wants us to “spark a global conversation”. Facebook is “giving people the power to build community and bringing the world close together” (Zuckerberg 2017). In this focus on dialogue and community, the platforms are tapping into an old understanding of sharing as a means of intensifying social relations through personal and emotional disclosure. According to media studies researcher Jenny Kennedy (2016), social intensification is one of the three main discourses of sharing in literature on technologically networked exchange.2 She uses Wittel’s (2011) sociological analysis to argue that the exchange of immaterial goods, such as news, is ‘inherently social’, as it only serves to invite responses from those in the sharer’s network and to accelerate interaction (2016: 468). In short, we share to connect more and more often. Platforms certainly seek to intensify our connective impulses by engineering rituals of information sharing. They send us regular, repeated daily alerts—who is tweeting what, who you should follow, how many responses you have had. To ignore these is to deny a basic drive to sociality, or so they would have us believe.\n\n2 The other two being sharing as an economic exchange, which embraces different types of research on how we allocate access to resources, and sharing as scaled distribution, which explores the extension and contestation of ownership in sharing systems.\n\nThe concept of social intensification also alludes to a marriage of old and new meanings for sharing online: the search for connection with like minds, and the increased speed and scale of messaging possible in networked systems. Social media analytics now measure ‘social velocity’ to understand the intensity of news sharing (and other social signals) as a measure of informational relevance. The proposition is that the faster and more often we share, like or favourite a breaking story, the more engaged we are with the topic, and the more important it is to the news media’s agenda. In 2016, Chartbeat subsidiary NewsWhip patented its social velocity algorithm, which it claimed at the time was used to identify breaking news and guide editorial decision making in 300 newsrooms worldwide (PR Web 2016).\n\nThis industrial interest in the gauging and valuing of social intensity speaks to the way in which our map of sharing’s semantic evolution is still incomplete. Our research suggests that social media sharing is interesting to journalists and editors for the way it signals the popular worth of stories, and magnifies their reproduction and distribution, supporting the so-called virality of highly shared content. Online we can not only share stories but with more people than we ever could face-to-face, and more quickly. Due to the exponential distribution effect enabled (and encouraged)\nby online networks (Kelly 1997) sharing can extend the reach and amplify the impact of false, shocking or scandalous news, with potentially significant effects. In India, for example, social media rumours and disinformation campaigns have recently resulted in riots and lynching (Tharoor 2018). In social media contexts, then, news sharing then not only distributes and reproduces, but _amplifies_.\n\nFurther, to share something immaterial is to seek wider recognition of that report, analysis, idea or opinion. We recommend (and hope) that our friends and followers attend to what we post. From a Bakhtinian perspective, we speak with anticipation of an audience and a response.3 In a social media system, we may add comments to a link or tag it so it is discoverable as part of larger discussions on a platform. We lift this item out of the flows and try to make it more _visible_. Our success of course is contingent on our playing the social game—sharing regularly, interacting lots with other users, sharing what is already popular, tagging posts with popular keywords, ‘optimising’ them for search-engine discoverability. And while Google argues that social signals do not influence a news story’s visibility in search rankings, they do influence, for example, a post’s Facebook ‘relevancy score’ (Oremus 2016). At Menlo Park, Facebook’s Silicon Valley headquarters, your past performance in signalling what’s important to you is one of the ‘hundreds’ of factors that play into the platform’s algorithmic prediction of what will then show up in your news feed, and be subject to on-sharing.\n\n3 See Mikhail Bakhtin (1981) The Dialogic Imagination: four essays, (Ed. Michael Holquist), University of Texas Press, Austin.\n\nSo, in sharing news on social media systems we are not only signalling what we think about that material, but making it available for evaluation by others.\n\nSharing is a collaborative evaluative process: we amplify and draw attention to a story, our contacts register their assessment (thumbs up, smiley face, heart, downvote), the platform registers, collates and analyses our reactions, and shows us more of what we like. We are entangled in a process of promoting and assessing information, a commendary circuit, which feeds social media companies’ understanding of how audiences behave and express desire, and enables personalised targeting of its behavioural advertising (Fig. 2.2).\n\n<figure>\nNews Feed Alert News Consumed Consumer Evaluation and Annotation News Sharing Social response Corporate algorithmic/human evaluation News Feed Story Placement Targeted Advertising\n<figcaption>\nFig. 2.2 The commendary circuit of news sharing. Image: Fiona R Martin\n</figcaption>\n</figure>\n\nIn socially networked environments, news sharing accrues meaning through its annotation and iteration. Users will normally frame the link to, an excerpt or copy of, a story with either an iconic evaluation, a like or an emoji, a comment about it, or a discursive categorisation such as a hashtag or category tag. Each type of annotation has different effects in terms of the way that the news item is then indexed, and its relative value is measured by the social system involved. Comments give companies richer data about our opinions and attitudes, so Facebook users are urged to “Say something about it”, while Twitter gives us the option to retweet with a comment. These evaluative gestures set the agenda for any subsequent interpretation and discussion of that material within a social network, and provide the data for the algorithmic placement of those items in the news feeds of the users’ social network. Hashtags, which are now used on Twitter, Facebook and Instagram to organise information flows, also act to amplify the reach of any original share, by making the item visible to interested parties. Thus, it is difficult to interpret the impact of any one act of sharing without also analysing its diffusion, and the accumulation of these incremental social signals.\n\nThe other problem for news organisations trying to promote this commendary activity is that social media users seem to favour the entertaining over the newsworthy. In 2017, there were no news organisations on YouTube’s top ten most viewed and subscribed channels, none in Instagram’s most viewed stories list, and only four news articles and two research features in the top fifteen most shared Facebook posts. One study of the best performing WeChat accounts found 24% were ‘media’ companies, but another noted that only 24% of users posted trending news content on Moments (tagline: “Share your life with friends”), and only 16.6% wanted to view this type of content, compared to 66.8% who posted personal life records and 62.7% who wanted to see that material. Only Twitter and Reddit stand out as systems that sustain significant reposting of news journalism (Bastos 2015; Barthel et al. 2016). The Pew Research Center found that 78% of Reddit users relied on its threads for news, either via information from other users or links and excerpts on the site. What then do we know about why people might share news rather than celebrity shots, funny pet videos and political memes, and why they might gravitate to particular news platforms?\n\n## Why We Share News Online\n\nAs many journalism texts will argue, news fulfils a basic human need to know about the actions of others, our place in the society and our capacity to act in the world. Exchanging news with others is fundamental to establishing and maintaining our social connections, to building and demonstrating social (and cultural) capital—and beyond those motivations is a way of projecting ourselves into the world, as self-determining individuals and agents of change. A recent and extensive survey of literature looking at news sharing motivations suggested they fell into three broad categories: we seek to attract attention, promote our reputation and gain social status; to socialise with others, discussing and developing ideas; and to share important information, not just altruistically but in order to process it better and in the hope that others will reciprocate (Kümpel et al. 2016).\n\nWhile study participants are sometimes unwilling to characterise themselves as self-serving, research by boyd et al. (2010), Lee et al. (2012) and Ihm and Kim (2018) found that there are strong presentational and impression management aspects to news sharing. People will carefully choose what they share to construct an ideal self-image, one that will attract approval from friends and peers. Some are keen to build their followers, or to get a response from a more visible actor in the network. Those who are highly motivated by self-presentation are more likely to share news, and on several platforms, than those who are less motivated by attracting recognition and acknowledgement (Ihm and Kim 2018: 14). This is likely because the more people we share new, credible and arresting information with, the higher our social visibility and the potential for building influence. The need to develop a reputation as a valuable news source applies most strongly when people are building professional networks on platforms like Twitter or Linkedin. But self-presentation factors may not be as broadly important in news sharing as social motives (Picone et al. 2016).\n\ndanah boyd and her colleagues found that people retweeted news for a range of social outcomes: as “an act of friendship, loyalty or homage”, to validate others’ ideas, and to recognise less visible participants in a conversation (boyd et al. 2010). A widely cited New York Times study also found the primary motives for their readers to share news were social: to improve the lives of people they cared about and to build and maintain relationships (NYT 2011). They hoped to help others have good experiences (or to avoid bad ones), to save money and to give them information about products that they’d be interested in. Readers also derived enjoyment from giving people useful news, from the responses they received and the sense of connection they felt to the world. Eighty per cent used news sharing as a way of keeping in touch with others, while 73% wanted to connect with those who shared their interests.\n\nThe NYT study also found evidence of a strong informational motive for news sharing: three in four of their surveyed readers “process information more deeply, thoroughly and thoughtfully as a result of sharing it with others” and said responses from their network helped them to interpret its meaning (NYT 2011: 6). Importantly, the survey suggested politics was key, as people share news about causes they want to promote. This rationale is supported by many studies that show the centrality of social sharing in mobilising political activity (see Hermida 2016; Papacharissi and de Fatima Oliveira 2012; Halpern et al. 2017; Kahne and Bowyer 2018).\n\nHowever, context is clearly central to why we share. In cases of natural disasters, we share to express our emotions and to establish how we can best act to respond to the crisis (Shaw et al. 2013). In therapeutic communities, where people gather to self-manage illness, they exchange news about research which might aid their treatment. Pedagogical groups share news about teaching, and which may be of use in the classroom (Swart et al. 2018). In an extensive survey of news sharing literature, researchers also found that people who most often share news on social media “perceive themselves as opinion leaders and tend to have lots of friends or followers”, consume a variety of news, especially on social media and follow news organisations (Kümpel et al. 2016: 5).\n\nWe certainly share news to influence others’ opinions, and intimate sharing is very effective in this sense. Both political and sociological studies have shown social recommendations have significant power to shape our belief in news, and the likelihood of us commending news to others. An American Press Institute study (Media Insight Project 2017) argues that Americans evaluate the trustworthiness of news based on who shares it, rather than where it comes from. If they trust the sharer, they are more likely to “recommend the news source to friends, follow the source on social media and sign up to news alerts from that source” (p. 1). This is particularly the case for young people. Boczkowski and his colleagues (Boczkowski et al. 2017) found that much of the news young people encounter is ‘incidental news’, seen in the process of following their social feed rather than accessed from a trusted brand as part of a deliberate news search strategy. In this environment, news appears as part of an undifferentiated flow of social and informational content, its value mediated by social recommendations.\n\nAs the fake news moment testifies, some people share news to deceive. In analysing the false stories that she saw circulating online during the 2016 Trump election campaign, digital journalism researcher Clair Wardle found seven different kinds of ‘fake news’—or mis-, dis- and mal-information, as she more accurately labels it (Wardle and Derakhshan 2017). There were items that made false connections or presented false context, as well as manipulated, misleading, imposter and fabricated content, and satire or parody (Wardle 2017). However, it is a moot point whether much of this deceptive material stemmed from ordinary individuals with unhappy intentions. Subsequent academic research suggests that politicised institutional and economic rationales have been driving the circulation of deliberately deceptive news—not in the least the notorious Internet Research Agency of St Petersburg, once Russia’s most prominent trolling operation, which a Guardian investigation (Hern et al. 2017) showed was posting supposed witness accounts from disasters and crime scenes that were picked up by UK and US publications including the Telegraph, BuzzFeed, the Daily Mail, the Huffington Post and the BBC.\n\nJournalists being fooled by imposter accounts is one problem, but a larger concern, which Wardle and Derakhshan (2017) highlight in a report for the Council of Europe, is whether and how ordinary social media users can identify a hoax post. The templating of social media items and their random appearance in our feeds, means that a post from a legitimate news source can look similar to a story from a conspiracy site, and equally shareworthy. They cite political scientists Messing and Westwood’s (2014) observations that social media content is decontextualised, since the focus is the story content, not the source. The Reuters Institute found that less than half of those they surveyed in 2017 usually recognised the brand of news content when they access news from social platforms (Kalogeropoulos and Newman 2017). Thus, without rich indicators of quality or origin, people are left to their social networks to help them identify and critique what’s authentic and what’s not.\n\nUnfortunately, misinformation is far more likely to be recirculated than the truth because it makes a claim that is unexpected and contradicts what people know about the world. Its novel, unusual nature makes it more interesting to share (Vosoughi et al. 2018). Researchers at MIT’s Laboratory for Social Machines found that the average ‘fake’ news item took 10 hours to spread to 1500 users on Twitter, whereas a verified story took 60 hours to reach the same number of people. They also found it travelled “farther … deeper, and more broadly than the truth”, partly because of the strong emotional reactions people had to it (Aral 2018). This points to an interesting aspect of virality that will be explored further in Chaps. 6 and 8—that news we feel strongly about is more likely to be shared.\n\nResearch from the last decade indicates that whether we share news will depend on how we feel about what we’re viewing. As Jonah Berger (2011) showed in a ground-breaking marketing psychology study, we are more likely to share information that stimulates or arouses us emotionally, whether we feel positively or negatively about that content. His work with Katherine Milkman (2012), which suggests that arousal increases the speed of news sharing, triggered a wave of studies that have explored how our emotional or affective states influence the sentiment of news we share, and the rate at which it spreads. As Chap. 6 indicates though, this research has provided some confusing and contradictory results.\n\nFor a start, it is not clear whether emotional reactions make us more inclined to share positive or negative stories. Decades of research on legacy news distribution has suggested journalists and consumers pay more attention to bad news than good (Galtung and Ruge 1965; Glasgow University Media Group 1980; Trussler and Soroka 2014), a trend reflected in news sharing on Twitter (Hansen et al. 2011). Yet Berger (2013) has argued that if we have a choice we prefer to pass on good news rather than bad, and this is supported by a recent study showing Twitter users share ‘overwhelming positive’ viral news (Al-Rawi 2017). Meanwhile, another group who set out to determine whether everyday bad news is likely to travel faster than good found that it did—at least as far as Brexit news on Twitter was concerned (Fang and Ben-Miled 2017).\n\nIt is clear, though, that sharing news is very much an affective pursuit, an immediate reaction to socially significant events, issues and actions. It invites others in our social network to identify with or dispute ideas and helps us to mobilise others around causes or campaigns. Zizi Papacharissi (2015) notes affective publics form around networked expressions of sentiment about news, supporting large-scale, pluralised responses to it, and connective—if not collective—action on it. Social sharing is a move away from the controlled, centralised representative agendas of analogue news, which offered limited means to publicly express our feelings, in our ways about important changes in our worlds. As Chap. 6 will argue, sharing news enables users to place themselves in the news matrix, integrating or contesting the concepts, symbols and ideologies of news narratives in ways particular to themselves and their cultural identities.\n\n## How We Share News Online\n\nJust as our motives for news sharing vary, we have many ways to do the sharing. People will use any networked communicative tool they read news on to flag or repost the URLs of content they find interesting or valuable, and there are many competing systems designed to allow us to do that. Much news sharing is done via email clients, from secure forums or using messaging apps like WeChat, Whatsapp, LINE or Facebook Messenger. The problem is that these channels, together with a range of mobile social apps, such as those belonging to Reddit, Facebook and Instagram, don’t generate traffic referral information in their message headers. That means measuring and understanding how people share news via these ‘dark social’ avenues is a difficult pursuit, and one that has become more pressing with the upswing in mobile traffic over desktop access to news websites (Breaux 2015).\n\nDark social is a phrase coined by former Atlantic magazine deputy editor Alexis Madrigal to describe the large volume of traffic coming to his site from sources that he couldn’t track. As Madrigal (2012) noted at the time, this problem exposed not only the range of sharing activities outside social platforms, but one of the great shifts that these platforms had visited on web media consumption—visibility:\n\n> the social sites that arrived in the 2000s did not create the social web, but they did structure it. This is really, really significant. In large part, they made sharing on the Internet an act of publishing (!), with all the attendant changes that come with that switch. Publishing social interactions makes them more visible, searchable, and adds a lot of metadata to your simple link or photo post. (Madrigal 2012)\n\nSocial media has become the main way that we share news online, and are encouraged by news companies to share, because it makes our activity visible and measurable. There are several ways in which Facebook, for example, makes what you share more apparent. It first gives your post a permanent URL, so it is discoverable by search engines. It adds traffic referral data to any link you post, say to a news story. When someone clicks on your linked item, the metadata tells the publisher’s server where your visit came from. Facebook’s software also crawls the web to capture data about that site, and to auto-populate a preview image for that link, excerpting the content and making the share more attention grabbing. This last technique is designed to capitalise on the likelihood that visual material such as photos and videos is more engaging than text alone. Imagery makes your posts more shareworthy. Finally, social media provides a range of automated commendary tools so we can evaluate others’ posts. As a whole, these algorithmic processes greatly elevate social media’s role in promoting media sharing, and produce quantifiable data about the scale of post creation and engagement. In enabling news publishers to collect, aggregate and measure social sharing metadata, or referral data about visits coming from platforms to news sites, social media companies have constructed themselves as key players in the digital distribution of news. As Madrigal also notes, social media platforms are only the latest in a long succession of computational systems for sharing information and news (see Fig. 2.1), each of which presented its own political and economic battles over how we could share information, what was and wasn’t proper use of the system. To get a more complete understanding of these struggles to define the authorised boundaries of news sharing, it is instructive to traverse some of that history, starting with first popular distributed online system, PLATO.\n\nLaunched in 1960, the US educational computing platform PLATO connected display terminals at the University of Illinois to dozens of networked mainframe computers. The platform’s development spawned many concepts now familiar in social computing including forums, message boards and chat rooms (Dear 2017) but also saw intense arguments over what was legitimate (coursework) and non-legitimate (gaming) use of the network (Latzko-Toth and Jones 2017). In a similar sense, France’s successful Minitel videotex system, which delivered news, directory, gaming, ticketing and billpaying services to ordinary citizens across the country, as part of a government sponsored roll-out, was scandalised by the growth of ‘messageries roses’, adult oriented chat rooms (Mailland and Driscoll 2017). US commercial service Prodigy offered news, email and bulletin board systems (BBS), but no chat, and alienated its subscriber base with over jealous censoring of content (Banks 2008). The UK’s centralised Prestel service was simply too expensive and visually uninteresting for most subscribers (Lean 2016).\n\nThe most successful news sharing systems were those that democratised the hosting and exchange of news, like BBS, which allowed any computer hobbyist with a PC and modem to publish or connect with specialised networks via the telephone network (Fig. 2.3).\n\nUsenet, which was developed in 1979 at Duke University so that people could exchange public messages between connected computers, expanded this concept beyond the national. Over its first decade, Usenet gradually linked more and more research institutions, and became host to an extraordinary variety of specialist discussion or ‘newsgroups’. There were struggles about the structural hierarchies, naming and control of these groups, and debates about what constituted permissible speech, all part of Usenet’s evolution as a public access network. it is remarkable for being the first international scale decentralised news sharing initiative, with the user-led alt. newsgroups still being the largest section of the system. Sharing information with other systems however, like the defence research oriented ARPAnet was not easy, with technical and administrative hurdles to inter-mailing list communications (Hauben and Hauben 1997). In 1982, users dreamed of a World Net “that would tie all sorts of computers worldwide together” (ibid.), a vision that would be realised in 1993 with the World Wide Web, a graphical language, system of resource locators and open transmission software protocol that would enable any computer users to share information with another over existing telecommunications networks.\n\n<figure>\n<figcaption>\nFig. 2.3 Online news sharing systems development timeline, adapted from Weber (2017)\n</figcaption>\n</figure>\n\nInternet historian Marc Weber (2017) argues that we lack a deep historical account of many of these early networks which might help us compare factors that are key to the effective distribution of information: the efficiency of different linking systems, of strategies to reduce the incidence of misinformation. However, it is likely that the one core reason we chose to share news via privatised, controlled social media rather than the more openly owned and structured World Wide Web is automation.\n\nThe advent of the World Wide Web in 1993 gave every PC and modem owner the possibility of being an author-publisher, and with a graphical focus that brought a new, if low-resolution visuality to news sharing. Websites were the new way to share your interests, even if much of the content still consisted of URL lists pointing to interesting sites and documents, and blogrolls of favoured writers. Yet, while the web lowered the barriers to sharing information across the globe, beyond the territoriality of analogue media, it still required technical skill, and an understanding of computing and coding beyond the reach of most people (let alone the billion or so worldwide living without electricity).\n\nAutomating the web publishing process made it more accessible to more people. Geocities, the US-based webhosting service founded in 1996, lowered barriers to creating web pages with templates and neighbourhood directories, and later drag and drop publishing tools. Together with Angelfire, Tripod and much later MySpace, Geocities gave users a base from which to learn the principles of html coding and media uploading, but it was templating that simplified, expedited and fully democratised online news sharing. Publishing templates defined the layout and features of many blogging services from Blogger and Live Journal to WordPress, and the first social media services, Six Degrees in 1997, Lunar Storm in 2000 and Friendster in 2000.\n\nIn 2006, Facebook further revolutionised the automation of news sharing by enabling websites to instal a share widget or button to their pages, which automated the posting of content to their accounts or contacts, making it more “more efficient by giving people a simple structure to do it” (Facebook 2006). In 2009, it then introduced its version of the like button, originally introduced by Friendfeed in 2007 before Facebook acquired that company. Liking enabled people to react to others’ sharing activities, but only positively. Zuckerberg recognised the value of this positive social reinforcement in encouraging interaction when he rejected calls for a ‘dislike’ button:\n\n> The like button is really valuable because it’s a way for you to very quickly express a positive emotion or sentiment when someone puts themselves out there and shares something. Some people have asked for a dislike button because they want to be able to say, “That thing isn’t good.” That’s not something that we think is good. We’re not going to build that, and I don’t think there needs to be a voting mechanism on Facebook about whether posts are good or bad. I don’t think that’s socially very valuable or good for the community to help people share the important moments in their lives. (Zuckerberg in Lafferty 2016)\n\nShare and react buttons are purposely designed technological affordances of Facebook that make acting on our impulses to pass on and respond to online information easy, quick and ‘frictionless’ (to use a term favoured in Silicon Valley). They encourage affective responses, but most obviously constructive ones as these are more likely to generate.\n\nFacebook’s resistance to automating critical signals is just one example of _affective regulation_, the process of controlling and directing expressive responses during social sharing. Another is Facebook’s introduction during 2017 of a set of five new emoji reaction buttons—Love, Haha, Wow, Sad and Angry—to supplement the Like. These options were ostensibly introduced in order to reduce incidences of reaction avoidance, where users felt it was inappropriate to ‘like’ content that made them upset. However, Facebook also indicated it considered the new reactions ‘stronger’ signals of relevance than the original like, raising questions about how it weighs them in calculating post visibility, and what affect users might have on news feeds by choosing different types of emotional response.\n\n## How Users Affect What Is, and Can Be, Shared\n\nUsers have changed the way platforms can be used to share news. Software developer and designer Chris Messina (2007) came up with the idea to use keyword ‘channel tags’ as a way to signal that a group discussion was opening, but keeping the signal short enough to be embedded in the message. Hashtags are now used to organise and aggregate shared content on Facebook and Instagram, illustrating the normalisation of socio-technical standards for sharing online. Tumblr users also adopted the GIF or ‘graphics interchange format’ to animate their personal reactions to posted content, drawing on an eclectic store of appropriated pop culture motifs. This playful convention was later adopted by digital-born news publishers like BuzzFeed, and GIFs are now being used by legacy media to give visual interest to stories and to animate simple infographics.\n\nHow we share news and what we share depends not only on a platform’s affordances, but also its commendary culture. This becomes clear in contrasting Reddit and Twitter, the international platforms most clearly focused on news sharing. On Reddit, users take part in discussion groups, called subreddits, devoted to particular topics, and news sharing involves excerpting, explicating and debating, to some extent, stories from existing media sources. The top 5 subreddits in terms of subscribers were humour (funny) world news, and educational (AskReddit, todayilearned, science). Users come from the US, Canada and the UK primarily, and in the US are largely young, 18–29-year-old males (Pew Internet 2016).\n\nTwitter, on the other hand, is more cosmopolitan and gender balanced, with 57% male users and 42% females in total, although this ratio varies by country. 79% of Twitter users come from outside the US, including China, its biggest national market, and India its third largest, while many countries including the UK and Japan are heavier users than North Americans. While its users are diverse, Twitter operates mainly as a professional and expert communications network, with news sharing done primarily through hyperlinks, due to the messaging constraints (280 characters per post, double that of pre-2017 tweets). More than half of the top most retweeted messages in 2017 were motivational, including three from Barack Obama, and two disaster appeals. While there is no recent global survey of account types on Twitter, an Australian study found that aspirational accounts were one of the top four concentrations, alongside teen culture (music and celebrity especially), sport and netizen/technologists (Bruns et al. 2017). That research also showed that user interests had changed over time, with technologists being a more important group in Twitter’s early years, while teens and sports enthusiasts being later adopters.\n\nEach evaluative system then demonstrates different politics of participation and recognition, based on its measures of social visibility, status and approval. On Reddit, which is a community of pseudonymous users, the title, visual content, placement and timing of a post will determine its visibility and popularity more than on Twitter, where real identities are more common and visibility is defined by numbers of followers and popularity of posts. A Twitter user’s profile and status, or social authority, has a distinct positive effect on content sharing. In terms of approval, each Twitter post accumulates a register of signals, including retweets, favourites, replies, follower requests and the addition of the user to expert lists. By way of contrast, Reddit users, in a quasi-democratic gesture, use up and down-voting to register their approval or disapproval of posts. As pseudonymity and anonymity pose particular issues of recognition and behavioural control for Reddit, this service has developed an additional reputation strategy. Users accumulate ‘karma’ points (a Silicon Valley gesture towards the Buddhist concept of an individual’s actions influencing their future), which then determine the degree of freedom they have to post information, the degree of attention paid to moderating their contributions, and their status within the system when users have low karma.\n\nThe sharing of politically contentious or illegal material can trigger internal or external regulation to control sharing behaviours, a subject we look at more closely in Chaps. 3 and 8. While the major social media platforms subscribe to the North American protection of speech freedoms, they nevertheless are obliged to control what we share, and how we speak about it in order to prevent the circulation of hate and discrimination, misinformation, propaganda and criminal activity. As Tarleton Gillespie argues in his investigation of content moderation online:\n\n> Platforms must, in some form or another, moderate: both to protect one user from another, or one group from its antagonists, and to remove the offensive, vile, or illegal—as well as to present their best face to new users, to their advertisers and partners, and to the public at large. (Gillespie 2018: 124–125)\n\nWhat we share and how it frames our relationship to the legal and political frameworks of nation state and platform revenues then affects what we care able to share. States, for example, routinely step in to regulate news sharing, showing they acknowledge its power to shape public opinions and political processes. Countries like China, Iran, Turkey, and Vietnam control to varying degrees both access to social media and activity on the platforms, usually to prevent political dissidence, and mobile internet access is sometimes restricted in areas dominated by religious and ethnic minorities. Governments will also restrict access to social sharing to further their own political agendas, as when the Ukrainian government blocked access to the popular Russian-owned VKontakte social platform on national security grounds (Kelly et al. 2017). As Chap. 3 will argue, these battles over access to news sharing systems represent a major front in struggles for freedom of expression and media freedom.\n\nWhat we share is also, as this book argues, fundamentally reshaping the news business, its relationship to the social media platforms and its audiences, and the production of news. To understand its impact on news journalism, we’ll now examine some of the factors that have affected social news sharing’s integration into newsroom routines, and its part in influencing how news is constructed.\n\n## Reshaping the News: The Likeable Engine\n\nIn 2009, Richard Sambrook, then director of the BBC Global News Division, told the Oxford Social Media conference that he thought “the impact of social media was overestimated in the short term and underestimated in long term” (Bunz 2009). Unlike most media predictions, his was remarkably accurate. The legacy news media have not wholesale adopted BuzzFeed’s social first approach to production, or, as Jonah Peretti’s organisation did early on, tied wages to evidence of social impact, clicks, likes and shares. Yet, nearly a decade later, social media organisations and their users have had a profound influence on how news is identified, selected, produced, distributed, promoted and discussed.\n\nAt peak social sharing, before Google reasserted its dominance and Facebook changed its algorithm to demote news in favour of personal information from people’s feeds, industry reports suggested anywhere from 7–50% of news traffic would come from social media referrals, depending on the site’s business model and the story topic (Mitchell and Jurkowitz 2014; Parse.ly 2015). Those digital-born companies based on social first publishing strategies like BuzzFeed and Upworthy had referrals at the top of that scale. At this time, social media metrics were a proxy for digital news consumption. This conception drove the uptake of social analytics which we document in the next chapter, and a greater editorial focus on monitoring what was being shared online and why. Although there is now a healthy scepticism about the importance of social media for news making, and a move for news organisations to refocus on a more diversified approach to distribution (Cornia and Sehl 2018), it is worth remembering how and why newsrooms sought to integrate social media during this period.\n\nThe growth of commendary culture and the rise in social referrals to news sites saw news organisations become tightly bound into the social media ecology, investing in branded social media channels, advertising and revenue deals, hiring specialised social media teams, adopting digital workflows and analytics systems and exploring occasional partnerships with the platforms (Cherubini and Nielsen 2016). As Bruns (2018) suggests, journalists were under pressure to boost sharing and increase audience trust in news product via personal social media engagement with audiences. In turn, these social branding activities led to power inequities in the way some individual reporters could command public attention and mobilise audiences, and engendered a lesser hold by employers over the opinions and job mobility of those employees (Bruns, 222–223). So, news sharing affected production investment and innovation agendas, while destabilising (to some extent) existing labour models and relations.\n\nFrom 2010 onwards, we were attempting to measure news sharing, first internal sharing of news across co-owned titles in Australia (Dwyer and Martin 2010) and then, together with our Share Wars colleagues, user-led sharing internationally. For the latter project, we used Share Wars’ Likeable Engine, a homegrown news analytics tool. For four years from 2011, Likeable tracked social news sharing trends using data from 140–160 major news sites worldwide and information feeds from Facebook’s and Twitter’s public application programming interfaces (APIs). Like other later commercial analytics applications like Chartbeat, Newswhip or Parse. ly, or the Guardian’s in-house developed Ophan dashboard, Likeable enabled editors and journalists to monitor changing patterns of social media news distribution and consumption (including their own relative performance), to track trending stories and to strategise live editorial decision making in competitive online news markets.\n\nLikeable was created by the Share Wars group, employees of one of Australia’s top-rated online news sites Nine.com.au (formerly ninemsn). Share Wars was a start-up built by former news editor Hal Crawford, former Windows 8 editor Andrew Hunter and former Network Development Director Domagoj Filipovic. They sought to use its analytics capacity to discover “what makes stories shareable and how this new distribution force will change workflows, roles and resourcing” (Crawford et al. 2015). The project marked a period where company executives started to take an interest in the uses of computational intelligence in audience research via their creative agency Mi9, and as Nine Entertainment split with Microsoft Network. The founders, who have all moved on to senior executive positions in large news media organisations and Facebook, maintain a blog of the project findings at http://share-wars.com/.\n\nThe Likeable Engine had four core components:\n\n1. News Feed Processor (extracting story URIs from news websites)\n2. Database (storing story data and matching it with social media data)\n3. Like Processor (extracting commendary data stats from FB and Twitter)\n4. Live reporting Interface (graphing the mashed data)\n\nIts architecture demonstrated the classic public/private nature of the social media ecology. It relied on public internet carriage and open standards web content (news sites), accessed public (but privatised) commendary information, formatted by proprietary social platform feeds, using bespoke scraping and querying algorithms and stored it using open source database software (Fig. 2.4).\n\nUntil its decommissioning in 2016, the Likeable Engine scraped new news story URIs linked from its sample news homepages every 11–20 minutes and then queried Facebook and Twitter APIs for the share counts of those stories. It requested these counts hourly for the first 24 hours, then less frequently for 5 days. Where possible URIs were canonicalised to their Facebook IDs, to avoid the same story being differently counted on its mobile or web domain. The interface then displayed the top ten stories graphed over 22 hours, and listed the top 100 with their share counts. Users could filter by region (AU, UK, US), by date and by satire (with an eye already to post-truth news).\n\n<figure>\n<figcaption>\nFig. 2.4 Likeable Engine information model. Image: Fiona R Martin AU14\n</figcaption>\n</figure>\n\nThe journalists interviewed in the early days of our project used Likeable as a complement to other paid analytics services such as Spike, Chartbeat, BuzzSumo, Omniture, and the in-built or native analytics provided by RedditInsight, YouTube and Google Analytics. Likeable users were primarily the ninemsn editorial staff, although users in several other workplaces, including Fairfax Media (which has since been taken over by Nine Entertainment), adopted the app. Interviews with five online journalists who used Likeable in two commercial newsrooms suggested that it played a useful and important role in their editorial routine when it was released, but had been overtaken by newer, more sophisticated analytics packages which they preferred.\n\nThe major insight from the interviews was the way in which analytics had become an important tool in editorial decision making. Interviewees said that they used analytics daily, and often throughout the day, to filter content, identify story ideas and trigger coverage, and to determine news scheduling and placement. Likeable was often consulted in the morning to identify story ideas based on what social media news was getting most user attention, where and when, to indicate whether journalists might need to produce reaction stories or explore new angles such as local perspectives. Two users regularly looked for ‘pre-viral’ stories or videos that were performing well on smaller news sites, and which had yet to be shared internationally. One sought news which “isn’t prominently placed on the original website, or from sites that are not part of the normal news cycle”.\n\nSharing analytics were also used to monitor the performance of competitor coverage of similar material. One news editor noted that Likeable helped him:\n\n> … to monitor sharing across the day and look at what’s sharing well comparatively—also to think about domestically what might get a push from posting on social media.\n\nAnalytics also emphasised the temporal flows of sharing, with trends suggesting times when stories might be posted so they “move better” and are shared more. One user noted he always looked at the top videos and what time they were posted “live versus the number of shares they get”.\n\nOverall, the interviews suggested social analytics had five distinct uses in editorial decision making. They were regarded as:\n\n-   an evaluative system—to identify stories and topics that were sharing well, or which might go viral, and which stories should stay on the homepage\n-   a news filter—to identify what topics were getting most attention, where and when\n-   an alert system—to signal new stories that are sharing well on smaller sites, or the need for reaction stories to viral content\n-   a scheduling device—to suggest times when stories might be posted to attract more shares\n-   an audience filter—to indicate which on which publications and platform certain stories were being most often shared.\n\nThese interviews also suggested how traditional new values are being reinterpreted relative to social sharing metrics. A story’s timeliness, which was once understood relative to the publication’s production lead time, is now also determined by whether and how long it has had a presence on social media. The need to ensure optimal news sharing is also determining whether stories are posted to social media or the World Wide Web first. The audience relevance of a story is being assessed by where its sharing is initiated (from the article page or via social platforms) and on which platform/s people are posting about it. Controversial or polarising topics tend to be highly shared, suggesting conflict remains an important news value, but exploiting that interest demands journalists include provocative keywords in story straps, excerpts and headlines to attract searches and shares. Shareable items as a whole sit on either end of the impact spectrum: they must be either novel and pre-viral, or ubiquitous, that is, viral and otherwise newsworthy. These evolving ideas about what makes news significant shows just how subtly and comprehensively social sharing analytics are influencing journalists’ normative interpretations and applications of news values.\n\nWhile we didn’t directly ask journalists if they understood how these figures were compiled, they did voice a range of concerns about the transparency of the metadata capture from Twitter and Facebook, particularly how analytics tools weighted views against shares or retweets, and how they should interpret differences in sharing relative to publication readership or audience size. One user requested better indexing of Australian news sites. The majority wanted better graphing of sharing in real time, one indicating that Omniture provided the best model.\n\nThese journalists’ interest in tracking the rate at which news was shared was consistent with the news media’s obsession with speed in news production and distribution. It also suggests that social virality is a central factor in the adoption of social analytics services such as Chartbeat, and a rationale quite distinct to measuring user attention to stories or journalistic engagement.\n\nOur research was never intended to fully map journalistic responses to social news sharing—far better covered in reports like Caitlin Petre’s _Traffic Factories_ (2015) or Fedrica Cherubini and Rasmus Kleis Nielsen’s _Editorial Analytics_ (2016)—but rather, we wished to trigger a critical enquiry into how social media news sharing is analysed, measured and valued. The next chapter takes up this story, particularly questions about the clarity and reliability of the data captured.\n\nMy newsroom encounter also highlighted the diversity of players in the social sharing ecology, with analytics companies based in Ireland, Israel and the US, and a distinct gap in the tracking of non-Western platforms like WeChat, Sina Weibo, LINE and Kakaotalk from our Asian region. This was a reminder that much news sharing knowledge still reflects the late twentieth-century cultural hegemony of transnational Western news domination. The deficiencies of this focus are obvious with the rise of BRICS countries. It was encouraging then to see, in December 2016, that Newswhip recorded IndiaTimes.com as the most shared site internationally on Facebook, part of a shift to it acknowledging greater diversity in the publishers who were achieving high share counts (Corcoran 2017). In an increasingly globalised communications world, we need to understand how our news sharing is shaped by different local, national and extra-national forces, and how commendary cultures might emerge in response to particular economic, political and social contexts.\n\nThe rise and fall of Likeable is one small part of that story. Likeable’s shutdown was partly the result of transnational competition, from analytics start-ups with more venture funding, more viable business models and better designed products, and partly due to Facebook and Twitter’s full commodification of what had previously been ‘freemium’ data subscription models. Likeable originally sampled data from the pre-2015 Facebook Developer API, which provided free access to separate counts of shares, likes, comments, and on-page comments—some of which were no longer available from the amended public API. It also drew on Twitter’s public data feed. This had been narrowing for years as the company slowly withdrew users’ permissions to access certain types of information, but it still provided tweet count data which was valuable to app developers and marketing services. Even if the free APIs had been somewhat unstable and subject to sudden change, they still enabled experimentation and innovation in the analytics space among a raft of small technology companies. The closure/commodification of both streams in 2015 undermined Likeable’s already limited functionality, and as it didn’t have a viable subscriber base, it was not a commercial proposition for Nine to buy that data from the platforms. Yet, Likeable had already fulfilled its ideological promise, introducing a group of journalists to the ground zero of analytics use, its incorporation into their news routines and the alteration of their production norms.\n\n## Commending the News\n\nSocial media news sharing has become a powerful new force in digital news production, in its variety of commendary forms and practices, its everyday adoption, scale of interaction and integration with news media production, distribution and consumption. Platform designs and their affordances operate to enable audiences to make visible the news they value as shareworthy, to marshal attention to their judgements, and to enlist others in responding to their reactions. Their evaluations, when tracked, analysed and represented, affect not only platform advertising, but journalism practice. Commendary culture influences the reporting and editorial practices of journalists in ways that are both obvious and also difficult to quantify, especially given tech companies’ relative lack of accountability for their calculus.\n\nSocial media platforms have, by promising informational and communicative agency, colonised the lifeworld of their users, and captured their commendary motivations and agency in ways and at a scale that have the potential to transform.\n\nRead in light of this analysis, online sharing means much more than redistribution, communication or disclosure. It now encompasses the amplification of ideas, the speed and scale of reproduction, the ritual and affective nature of our acts and the differing levels of access we provide to our shared material. It also presages an era of machine learning and automated news discovery and redistribution.\n\nThere is little doubt that social sharing is reconfiguring news journalism, even where social referrals constitute a small percentage of traffic to a\n\nnews site. That is because social media users now routinely influence what gets seen by others, and what is valued in the news selection process, in a circular and self-reinforcing logic, just by redistributing and evaluating news stories. Their sharing metrics feed the editorial algorithms of the platforms, influencing how news is made visible and prominent and generating the metrics that newsrooms will monitor for signs they are producing what the audience wants. It is this commendary dynamic, quite apart from the gravitation of advertising revenue to the platforms, that is reshaping the future of news and how it will reach audiences.\n\n## References\n",
 "content_html": "<h1>Commendary Cultures</h1>\n<p>Social sharing of news has a critical economic and cultural function in an attention economy. It reveals our personal interests, beliefs and tastes to the array of publications trying to win our interest and identifies to them and our social networks what is valuable to us in the cascades of data that we negotiate every day. Through our shares, likes and favourites we help the social platforms determine what is popular and profitable in advertising terms. We also elevate certain events and issues in public debate, and reinforce opinions and attitudes in our social circles. Thus, if most mainstream media users are more likely to redistribute, like or recommend news than to create it, then news sharing is potentially as important a phenomenon to study as participatory or citizen journalism.</p>\n\n<p>This makes it imperative for reporters, editors and news executives to understand, beyond the metrics they might receive about Facebook or Twitter trends, what motivates news sharing in different contexts, as well as how and why we are likely to share stories we come across online. It is also important to recognise how the platforms have ensured news sharing is imbricated in the rhythms of everyday life online and integrated into the editorial processes of news rooms. This shift is driving the reorganisation of news production, consumption and distribution, and redirecting economic, political and cultural power away from the legacy media and towards the platforms.</p>\n\n<p>It’s easy to focus on social media platforms as the primary site of news sharing, as they are designed precisely for that purpose: the easy personal exchange of information with what passes for a self-built social network (although recommendation algorithms play a big role in helping us constructing that sociality). Social media companies make sharing simple to encourage us to disclose more about our everyday lives, so they can analyse that routine evidence of activities and preferences to create more effective forms of advertising. It’s a twenty-first-century design for a radical, globalised consumer sovereignty that seeks to make as much social behaviour as possible accessible to corporate scrutiny or, as Facebook’s most durable first decade mission statement more altruistically puts it, “To give people the power to share and make the world more open and connected”\n(Reagan 2009).</p>\n\n<p>Yet, news sharing also takes place via email, apps, forums and other online channels, just through the posting of hyperlinks—a practice as old as the World Wide Web, and which has its basis in earlier hypertext systems. Each time we post a link to a news story, we are curating from our own feeds what might be of interest to others, recommending what they should attend to and demonstrating something of our own concerns and biases. Link sharing is not always an endorsement of the content, but demands those who click through consider its meaning and worth. So, this chapter explores how social news sharing has become central to the creation of commendary cultures online, where ordinary people take on the representational and distributive roles of the news media, as well as something of journalism’s filtering and evaluative functions. It explores the pre-history and evolution of news sharing technologies and the emergence of different ideas about what sharing means, why we do it and how it’s affecting digital journalism.</p>\n\n<p>If link sharing is the basis for online news exchange, how then did social media become synonymous with this practice, to the extent that people around the globe now see these channels as key news sources? To answer this question, we first need to establish what qualifies as ‘social media’. As new media scholar Graham Meikle notes this term evokes “a particular set of technological affordances, a particular set of business models and corporate practices, a particular set of organisations, and a particular set of cultural habits, practices and expectations” (2016: x). The organisations are well recognised worldwide and becoming virtual monopolies in their various niches by virtue of their transnational scale and market dominance. By late 2016, Alphabet (owner of Google and YouTube), Facebook (with Whatsapp and Instagram) and Microsoft (LinkedIn and Yammer) were among the largest companies internationally by market capitalisation, bigger than the finance and energy companies they had succeeded on that list (Flew et al. 2018). Facebook is now the world’s most popular social media platform, with around 2.19 billion monthly active users in the first quarter of 2018. In the same period, WeChat has reached 1 billion Chinese users alone, more than Twitter has worldwide, while YouTube dominates video sharing and video on demand advertising globally.</p>\n\n<p>While each company has a distinct business model, on the whole, the biggest players’ economic success has been based in five factors:</p>\n\n<ol>\n  <li>broadening the base of production to every subscriber;</li>\n  <li>making global communication and publishing easier through templating and automation;</li>\n  <li>reducing or removing the geographic boundaries to social and business transactions;</li>\n  <li>aggregating and analysing user created data to create more efficient, individually targeted advertising sales models; and 5. operating as international information technology companies rather than national media publishers, which has until recently put them outside traditional media content regulations and allowed them to publish content that others could or would not.</li>\n</ol>\n\n<p>Economist Yochai Benkler (2004) suggests internetworked social sharing of information has proved a more efficient form of distribution than historic market forces. It is decentralised, avoiding the strictures, conventions and routines of corporate news gatekeepers; autonomous, allowing anyone and everyone with network access to take part, and personalised, in tying the act of recirculating news to communicative rituals and social conventions of information gifting and reciprocation.</p>\n\n<p>Social media companies’ technological affordances have enabled us to increase our everyday news sharing, extending it to large, curated social networks. Starting with Six Degrees, launched back in 1997, and then Friendster in 2002, social networking applications allowed users to create public or semi-public personal profiles, to build a list of social connections that they could easily navigate, and to communicate publicly or privately with those people using templated publishing tools (Hinton and Hjorth 2013). Later social media technologies also enabled users to create, aggregate, evaluate and share media, to create groups and audiences for this content, and to monitor user analytics based on that service interaction. From 2006, Facebook and others provided widgets or limited functionality scripts for publishers to embed in their sites so users could subscribe to news publishers’ social media channels, or share a story link directly on reading it on a branded website, and later, mobile apps for continuous access to this sharing capacity.</p>\n\n<p>All of these strategies were designed to simultaneously extend our customised communicative reach and to automatically co-opt more users and publishers into the orbit of platform analysis, in what critical media studies scholar Jose van Dijck and Poell (2013: 8) calls a “double logic” of connectivity. Platform analytics are not only “perpetually operational” and so pervasive that we don’t notice them (as [Kennedy 2016: 25] suggests), but are also designed to reward and reinforce our participation; to make it addictive using a series of behaviour reinforcing communication strategies (Alter 2017). Alerts or notifications from our social media feeds tell us how much attention our posts are getting, and detail our network members’ activities, to encourage our reciprocity and to routinise our sharing activity. Sometimes, gamification is built into the systems, so that we are notified where in a hierarchy of users we are placed as a result of our activity. In these ways, platforms cultivate habits of documentation, declaration, discussion and debate, through constant visual and auditory reminders of their presence. They recommend related posts and stories we should like, given our past consumption and interaction history, and make others invisible, through algorithmic processes that maximise our exposure to popular content, personalised advertising and online services. So, while social media are not the simplest or only ways to share news stories, they do provide a widely accessible suite of portable tools for customising our news consumption, accelerating the scale and scope of its distribution and exposing it to the scrutiny of algorithmic tracking, classification and assessment which allows them to monetise any insights into our behaviour and interests.</p>\n\n<p>In positioning themselves as IT companies and carriers of content rather than publishers, social media companies historically have avoided having to conform to the content regulation laws and accountability measures set up for legacy media, or to worry overly about what their users post on their platforms. For example, the U.S. Digital Millennium Copyright Act includes a safe harbour provision, which protects them from liability for the content their users might post—as long as they act to remove any illegal content about which they are informed (Flew et al. 2018). There is, as yet, no legal mechanism for forcing platforms to reveal how their algorithms structure the content we see, and that we don’t see, or what human editorial decision making takes place in content management. Recent national and regional moves to regulate the content we share, like Germany’s Network Enforcement Act (German Law Archive 2017) and the European Commission’s (2017) Code of Conduct on Countering Illegal Hate Speech have put some pressure on the major platforms to remove violent and extreme political speech (Goggin et al. 2017). However, on the whole, what we share online is determined by the guidelines that they set, and which they are not answerable for—even to us.</p>\n\n<p>From this relatively unassailable legal position, and in possession of audience knowledge that the old media could only dream of, major social media platforms have been in a position to influence both cultural production and the production of culture, especially news media culture.</p>\n\n<h2 id=\"the-importance-of-commendary-culture\">The Importance of Commendary Culture</h2>\n\n<p>As a source of traffic to news websites, social media’s direct influence on news distribution may well have peaked and passed. Certainly its 2015 dominance of referred news traffic was a blip on the metrics radar. Since then, Google search has reasserted its place as the number one online news referral source. Following the 2016 election misinformation scandal, Facebook also demoted news page content in people’s feeds (Facebook 2018), leading some industry observers to argue that publishers should reorient their strategy away from that platform (Filloux 2018) and engage more deeply with their audiences rather than using social media channels to push out links to stories (Elizabeth 2017).</p>\n\n<p>Yet, some media commentators still argue that social media sharing may have profound effects on the future of journalism. They note that young people’s turn to social media as a primary news source—and social media companies’ capacity to imperceptibly select, filter and promote the news that gets seen—may have three immediate impacts: eroding journalism’s power to connect directly with audiences, to deliver what they need to know in the public interest and to ensure the accuracy and quality of the news they consume (Bell 2016; Bell and Taylor 2017). They worry that there is little platform accountability for what news gets seen, and by whom, and how much revenue is shared as a result (Newman 2018). Some see journalists as being locked into reporting what gets likes and shares, tailoring the news to social media metrics and “dancing to the popular music of consumer driven logic” (Tandoc 2014: 572). Others, like digital media theorist Jodi Dean (2005, 2014), have positioned social media users as unpaid labourers in the platforms’ bid for informational dominance, questioning whether their circulation of ideas can do much to affect the democratic process (as well-informed citizens might hope to), or is attenuated instead in technological fetishism. From this perspective, news sharing may be just so much symbolic chaff, spread in lieu of political engagement or action. Christian Fuchs has called out Facebook CEO Mark Zuckerberg’s claim that “the goal of the company is to help people share more in order to make the world more open and help promote understanding between people”, as a contradictory ruse that belies a ruthless commercial business model (Fuchs 2018: 172).</p>\n\n<p>The more hopeful call social media news sharing a new form of media democracy and creativity, enabling a greater diversity of people to signal to newsmakers and our social networks what’s truly significant to them, in both local contexts and globally. Ostensibly, social signalling should guide journalists to report on what we value most, and on more varied conceptions of that value. In an ideal world, that would make journalism more socially responsive and culturally diverse. This techno-optimist shift is characterised in Henry Jenkins (1992) notion of participatory culture, and his newer concept of “spreadability”, where “new tools have proliferated the means by which people can circulate material, word of mouth recommendations and the sharing of media content” and facilitated the diffusion of cultural artefacts and practices (Jenkins et al. 2013: 2–3). It is also embraced in Bruns’ (2005, 2018) idea of “gatewatching”, where ordinary users publicise news, rather than reporting or publishing it. Bruns sees news sharing as a demotic, habitual practice for most people, and core to the emergence of new forms of expertise. But while these claims may have some basis, we need to know far more about the political economics and cultural forms of shared news, as well as the real-time network dynamics and contexts for its production and consumption, before we can understand its ramifications for the news media or its role in our societies.</p>\n\n<p>Up until now citizen journalism has had far more attention in media research than sharing. Most book-length studies of participatory news cultures to date have focused on novel forms of news creation: blogging, produsage and gatewatching (Bruns 2005, 2008, 2018), user-generated content (Singer et al. 2011), citizen journalism (Allan and Thorsen 2009; Miller 2014), citizen witnessing (Allan 2013) and maker and hacker culture (Usher 2016), rather than the new modes of distribution such as aggregation and sharing. Alfred Hermida’s #TellEveryone: why we share and why it matters, a lively journalism studies account of how everyday social news sharing works, concentrates on the sociology of sharing, leaving the broader questions of cultural change and platform power, their shape and impacts largely unaddressed. In Axel Bruns’ more theoretically engaged <em>Gatewatching and News Curation</em>, he devotes considerable energy to analysing news sharing, but tends to focus on the relationships between users, journalists and platforms, rather than the more complex industrial and regulatory relationships emerging in the social media news ecology. While these books are important interventions in social media research, the study of news sharing remains a relatively understudied transformation of cultural production, and one whose complex power dynamics remain obscure.</p>\n\n<p>Rather than taking one or other polar position on how our social media engagement will affect the future of news, we adopt a critical media perspective on news sharing’s meanings for journalism in society. That is, we want to examine how our online news sharing has been commodified, what role our actions, beliefs and proclivities have played in that process and what impact it will have on the institutional notion of journalism and the new cultures of digital news consumption.</p>\n\n<p>The rest of this chapter examines the rich meanings of sharing, and how they have shifted to accommodate our new socially networked interactions and the commodification of that sociality. It then investigates why we share news online and how, exploring how our motivations for promoting certain types of news and the nature of our interactions is reshaping the way news generally is selected, reported and distributed, and in different ways on different platforms. In these senses, it is an account of the cultural politics of news sharing, or the ways that our cultural practices affect the exercise of news media power, and vice versa. As Angus and Jhally (1989) argued, it is important to understand “the manner in which institutional and ideological structures act as limits to the possibilities of cultural practices” and how cultural activity supports the emergence of new forms of power such as those of the transnational social media platforms.</p>\n\n<p>So, instead of focusing on people’s impulse to create media online or to participate in public debate, as most participatory journalism studies have done to date, we are exploring the rise of a more everyday use of internet technologies—to commend, recommend or condemn the news. In the first instance, news sharing seeks to draw people’s attention to a news item by extending its reach. Sharers often tell others what they think about their shared item to rationalise the reposting, and seek or expect some sort of response from their social networks. They may hope to engage others in discussing or debating or acting on the judgement they have expressed.</p>\n\n<p>The term <em>commendary culture</em>, used here to describe the rituals, beliefs, norms and practices associated with sharing pre-existing messages online, is drawn from the two fourteenth-century Latin meanings of commendare: to entrust someone or something to the care of others, and to commit to writing. These days, the acts of commending and recommending have the same positive connotation of praising the object in question. Those stronger positive associations have developed in contradistinction to the concept of condemnation, criticism or censure, yet to commend something originally meant simply to note it as worthy of attention. The study of commendary culture then, is concerned with all acts of elevating media to the attention of others, regardless of the conclusion drawn about that material. It understands news sharing as a social and cultural practice which seeks to amplify messages, to capture the attention of interested others, and to engage them in a reciprocal evaluative process. It is also interested in the socio-technical systems that allow this sharing of symbolic goods, preferences and opinions, and the means of making sense of these acts—\nthe techniques of measuring and analysing commendary activities.</p>\n\n<p>As the title of this books suggests, the commendation of news stories to our social networks is commonly described as news ‘sharing’. Sharing is such a polysemic and widely used term in online contexts that it demands some deconstruction before we look at why it is such a compelling activity.</p>\n\n<h2 id=\"what-is-sharing\">What Is Sharing?</h2>\n\n<p>As Nicholas John (2017) reveals in his absorbing cultural history of sharing, the phrase ‘to share’ has three different root meanings: to distribute or divide, to have something in common with and to communicate one’s personal experience or feelings. Like marketing scholar Russell Belk, John is intrigued by the association of sharing with economic behaviours, and particularly prosocial internet use.</p>\n\n<p>Belk (2010), from an anthropological perspective, originally argued that sharing is a form of consumption within practices of commodity exchange and gifting, based in mutual association, shared possession or collaborative ownership, and caring relationships with others. For this perspective in sharing news online, we are not losing our share of anything, but expanding what he calls our sense of an ‘extended self’, the projection of our identity in the world through external means (p. 724).1 However,</p>\n\n<p>1 See also Belk (1988) Possessions and the Extended Self.</p>\n\n<p>he saw sharing as a non-commodity relationship, which it clearly is not in the contemporary sense, when social media companies reap both advertising data and attention from our information sharing activities. Thus, while Belk later modified his definition of sharing to include industrial models of collaborative consumption and short-term rental (Belk 2014), John’s work is more relevant to our purposes. He sets out to explore how tech corporations have appropriated the idea of sharing to fuel the expansion of digital capitalism via the ‘sharing economy’—a key concept enabling social media companies to appropriate our personal data, our social connections and our interests.</p>\n\n<p>John (2017: 62–64) proposes that sharing has acquired new meanings in the age of social networking, associated with the reach, scale, morality and emotional intimacy of online communication. The first has come via the new organisational techniques of early computing, including practices like time sharing and file sharing, which point to sharing’s networked or distributive potential. Wadbring and Ödmark (2016) argue that there is a difference between the news media’s deliberate distribution strategies, which are designed to attract, hold and measure audiences, and social sharing’s ‘circulatory’, participative logics. We disagree with that assessment. As we will argue, platforms have their own deliberate, quantified distribution logics which seek to reach as many as possible with the most popular content their algorithms identify.</p>\n\n<p>The second meaning for news sharing is reproductive, as online messaging collapses sharing’s two primary meanings of communication and distribution. In sharing news, we are simultaneously engaged in an interpersonal act of telling others about ourselves and publishing that information for an audience to read. Sharing in these reproductive senses is also enabled by what Meikle (2016: 26) calls the “intrinsic” capacity of computational technologies to produce perfect copies, which means we can give information without losing it and in a non-rivalrous manner, as the original is still around for others to consume. Despite the culture industries’ ongoing legal proceedings to assert their digital rights in music and film file sharing (Borschke 2017) and in news aggregation (Martin 2015), the capacity to make and exchange digital copies has established online media sharing as a commonplace, even virtuous act—especially where it challenges the constrictive licencing regimes of media and publishing conglomerates. Meikle quotes programmer and former Reddit co-owner Aaron Swartz arguing that “sharing isn’t immoral—it’s a moral imperative. Only those blinded by greed would refuse to let a friend make a copy” (Swartz 2015).</p>\n\n<p>This <em>reproductive potential</em> underpins the third and most potent reason for sharing’s expropriation by Silicon Valley—what John calls “its positive connotations of equality, selflessness and giving” (John 2017: 63). This beneficial sense of online sharing, he notes, accords neatly with origin mythologies of the internet as a democratic force, a network for free, transparent and non-hierarchical exchange of information. So, news sharing has an <em>egalitarian potential</em> that makes us feel good about commending coverage of events and ideas to others.</p>\n\n<p>Fourthly, a factor that John briefly alludes to and Meikle develops, is sharing as intimacy, and immediate personal exchange. We most often share online with our friends, family and workmates, people with whom we have some close, emotional or intellectual connections, and these bonds influence what we share and why we share it in networked circumstances. We share to build, affirm, maintain or repair relationships (Goh et al. 2017; Meikle 2016: 26). As we will see in Chap. 6, the users of an online news service who we surveyed primarily share news to show they care about a topic. They do it to express their feelings about a subject and sometimes to provoke a reaction from their friends or followers. This <em>affective potential</em> of sharing is key to understanding the topics people will respond to, and what types of stories might go viral on a social platform (Fig. 2.1).</p>\n\n<figure>\n\n  <p><img src=\"\" alt=\"\" /></p>\n\n  <figcaption>\n\n    <p>Fig. 2.1 The semantic associations of social media sharing. Image: Fiona R Martin</p>\n\n  </figcaption>\n  <figure>\n\n    <p>Over time, social media companies have recognised how to capitalise on these semantic potentials, with Facebook in particular promoting our information sharing as a “mechanism for improving human relations and making the world a better place” (John 2017: 65–66). As John critically observes, they also tend to use the term sharing to underplay the commodity exchanges they have with third-party advertisers or developers, with whom they also ‘share’ our personal information. This appropriation of the positive meanings for sharing to legitimate purely commercial and wholly privatised trade activity is meant to divert user attention from the potential privacy implications of this activity.</p>\n\n    <p>The corporate intention is to promote sharing as collectively oriented, authentic and socially productive. Reddit’s 2018 homepage states it is “home to thousands of communities, endless conversation, and authentic human connection”. Twitter too wants us to “spark a global conversation”. Facebook is “giving people the power to build community and bringing the world close together” (Zuckerberg 2017). In this focus on dialogue and community, the platforms are tapping into an old understanding of sharing as a means of intensifying social relations through personal and emotional disclosure. According to media studies researcher Jenny Kennedy (2016), social intensification is one of the three main discourses of sharing in literature on technologically networked exchange.2 She uses Wittel’s (2011) sociological analysis to argue that the exchange of immaterial goods, such as news, is ‘inherently social’, as it only serves to invite responses from those in the sharer’s network and to accelerate interaction (2016: 468). In short, we share to connect more and more often. Platforms certainly seek to intensify our connective impulses by engineering rituals of information sharing. They send us regular, repeated daily alerts—who is tweeting what, who you should follow, how many responses you have had. To ignore these is to deny a basic drive to sociality, or so they would have us believe.</p>\n\n    <p>2 The other two being sharing as an economic exchange, which embraces different types of research on how we allocate access to resources, and sharing as scaled distribution, which explores the extension and contestation of ownership in sharing systems.</p>\n\n    <p>The concept of social intensification also alludes to a marriage of old and new meanings for sharing online: the search for connection with like minds, and the increased speed and scale of messaging possible in networked systems. Social media analytics now measure ‘social velocity’ to understand the intensity of news sharing (and other social signals) as a measure of informational relevance. The proposition is that the faster and more often we share, like or favourite a breaking story, the more engaged we are with the topic, and the more important it is to the news media’s agenda. In 2016, Chartbeat subsidiary NewsWhip patented its social velocity algorithm, which it claimed at the time was used to identify breaking news and guide editorial decision making in 300 newsrooms worldwide (PR Web 2016).</p>\n\n    <p>This industrial interest in the gauging and valuing of social intensity speaks to the way in which our map of sharing’s semantic evolution is still incomplete. Our research suggests that social media sharing is interesting to journalists and editors for the way it signals the popular worth of stories, and magnifies their reproduction and distribution, supporting the so-called virality of highly shared content. Online we can not only share stories but with more people than we ever could face-to-face, and more quickly. Due to the exponential distribution effect enabled (and encouraged)\nby online networks (Kelly 1997) sharing can extend the reach and amplify the impact of false, shocking or scandalous news, with potentially significant effects. In India, for example, social media rumours and disinformation campaigns have recently resulted in riots and lynching (Tharoor 2018). In social media contexts, then, news sharing then not only distributes and reproduces, but <em>amplifies</em>.</p>\n\n    <p>Further, to share something immaterial is to seek wider recognition of that report, analysis, idea or opinion. We recommend (and hope) that our friends and followers attend to what we post. From a Bakhtinian perspective, we speak with anticipation of an audience and a response.3 In a social media system, we may add comments to a link or tag it so it is discoverable as part of larger discussions on a platform. We lift this item out of the flows and try to make it more <em>visible</em>. Our success of course is contingent on our playing the social game—sharing regularly, interacting lots with other users, sharing what is already popular, tagging posts with popular keywords, ‘optimising’ them for search-engine discoverability. And while Google argues that social signals do not influence a news story’s visibility in search rankings, they do influence, for example, a post’s Facebook ‘relevancy score’ (Oremus 2016). At Menlo Park, Facebook’s Silicon Valley headquarters, your past performance in signalling what’s important to you is one of the ‘hundreds’ of factors that play into the platform’s algorithmic prediction of what will then show up in your news feed, and be subject to on-sharing.</p>\n\n    <p>3 See Mikhail Bakhtin (1981) The Dialogic Imagination: four essays, (Ed. Michael Holquist), University of Texas Press, Austin.</p>\n\n    <p>So, in sharing news on social media systems we are not only signalling what we think about that material, but making it available for evaluation by others.</p>\n\n    <p>Sharing is a collaborative evaluative process: we amplify and draw attention to a story, our contacts register their assessment (thumbs up, smiley face, heart, downvote), the platform registers, collates and analyses our reactions, and shows us more of what we like. We are entangled in a process of promoting and assessing information, a commendary circuit, which feeds social media companies’ understanding of how audiences behave and express desire, and enables personalised targeting of its behavioural advertising (Fig. 2.2).</p>\n\n    <figure>\n      <p>News Feed Alert News Consumed Consumer Evaluation and Annotation News Sharing Social response Corporate algorithmic/human evaluation News Feed Story Placement Targeted Advertising</p>\n      <figcaption>\n        <p>Fig. 2.2 The commendary circuit of news sharing. Image: Fiona R Martin</p>\n      </figcaption>\n    </figure>\n\n    <p>In socially networked environments, news sharing accrues meaning through its annotation and iteration. Users will normally frame the link to, an excerpt or copy of, a story with either an iconic evaluation, a like or an emoji, a comment about it, or a discursive categorisation such as a hashtag or category tag. Each type of annotation has different effects in terms of the way that the news item is then indexed, and its relative value is measured by the social system involved. Comments give companies richer data about our opinions and attitudes, so Facebook users are urged to “Say something about it”, while Twitter gives us the option to retweet with a comment. These evaluative gestures set the agenda for any subsequent interpretation and discussion of that material within a social network, and provide the data for the algorithmic placement of those items in the news feeds of the users’ social network. Hashtags, which are now used on Twitter, Facebook and Instagram to organise information flows, also act to amplify the reach of any original share, by making the item visible to interested parties. Thus, it is difficult to interpret the impact of any one act of sharing without also analysing its diffusion, and the accumulation of these incremental social signals.</p>\n\n    <p>The other problem for news organisations trying to promote this commendary activity is that social media users seem to favour the entertaining over the newsworthy. In 2017, there were no news organisations on YouTube’s top ten most viewed and subscribed channels, none in Instagram’s most viewed stories list, and only four news articles and two research features in the top fifteen most shared Facebook posts. One study of the best performing WeChat accounts found 24% were ‘media’ companies, but another noted that only 24% of users posted trending news content on Moments (tagline: “Share your life with friends”), and only 16.6% wanted to view this type of content, compared to 66.8% who posted personal life records and 62.7% who wanted to see that material. Only Twitter and Reddit stand out as systems that sustain significant reposting of news journalism (Bastos 2015; Barthel et al. 2016). The Pew Research Center found that 78% of Reddit users relied on its threads for news, either via information from other users or links and excerpts on the site. What then do we know about why people might share news rather than celebrity shots, funny pet videos and political memes, and why they might gravitate to particular news platforms?</p>\n\n    <h2 id=\"why-we-share-news-online\">Why We Share News Online</h2>\n\n    <p>As many journalism texts will argue, news fulfils a basic human need to know about the actions of others, our place in the society and our capacity to act in the world. Exchanging news with others is fundamental to establishing and maintaining our social connections, to building and demonstrating social (and cultural) capital—and beyond those motivations is a way of projecting ourselves into the world, as self-determining individuals and agents of change. A recent and extensive survey of literature looking at news sharing motivations suggested they fell into three broad categories: we seek to attract attention, promote our reputation and gain social status; to socialise with others, discussing and developing ideas; and to share important information, not just altruistically but in order to process it better and in the hope that others will reciprocate (Kümpel et al. 2016).</p>\n\n    <p>While study participants are sometimes unwilling to characterise themselves as self-serving, research by boyd et al. (2010), Lee et al. (2012) and Ihm and Kim (2018) found that there are strong presentational and impression management aspects to news sharing. People will carefully choose what they share to construct an ideal self-image, one that will attract approval from friends and peers. Some are keen to build their followers, or to get a response from a more visible actor in the network. Those who are highly motivated by self-presentation are more likely to share news, and on several platforms, than those who are less motivated by attracting recognition and acknowledgement (Ihm and Kim 2018: 14). This is likely because the more people we share new, credible and arresting information with, the higher our social visibility and the potential for building influence. The need to develop a reputation as a valuable news source applies most strongly when people are building professional networks on platforms like Twitter or Linkedin. But self-presentation factors may not be as broadly important in news sharing as social motives (Picone et al. 2016).</p>\n\n    <p>danah boyd and her colleagues found that people retweeted news for a range of social outcomes: as “an act of friendship, loyalty or homage”, to validate others’ ideas, and to recognise less visible participants in a conversation (boyd et al. 2010). A widely cited New York Times study also found the primary motives for their readers to share news were social: to improve the lives of people they cared about and to build and maintain relationships (NYT 2011). They hoped to help others have good experiences (or to avoid bad ones), to save money and to give them information about products that they’d be interested in. Readers also derived enjoyment from giving people useful news, from the responses they received and the sense of connection they felt to the world. Eighty per cent used news sharing as a way of keeping in touch with others, while 73% wanted to connect with those who shared their interests.</p>\n\n    <p>The NYT study also found evidence of a strong informational motive for news sharing: three in four of their surveyed readers “process information more deeply, thoroughly and thoughtfully as a result of sharing it with others” and said responses from their network helped them to interpret its meaning (NYT 2011: 6). Importantly, the survey suggested politics was key, as people share news about causes they want to promote. This rationale is supported by many studies that show the centrality of social sharing in mobilising political activity (see Hermida 2016; Papacharissi and de Fatima Oliveira 2012; Halpern et al. 2017; Kahne and Bowyer 2018).</p>\n\n    <p>However, context is clearly central to why we share. In cases of natural disasters, we share to express our emotions and to establish how we can best act to respond to the crisis (Shaw et al. 2013). In therapeutic communities, where people gather to self-manage illness, they exchange news about research which might aid their treatment. Pedagogical groups share news about teaching, and which may be of use in the classroom (Swart et al. 2018). In an extensive survey of news sharing literature, researchers also found that people who most often share news on social media “perceive themselves as opinion leaders and tend to have lots of friends or followers”, consume a variety of news, especially on social media and follow news organisations (Kümpel et al. 2016: 5).</p>\n\n    <p>We certainly share news to influence others’ opinions, and intimate sharing is very effective in this sense. Both political and sociological studies have shown social recommendations have significant power to shape our belief in news, and the likelihood of us commending news to others. An American Press Institute study (Media Insight Project 2017) argues that Americans evaluate the trustworthiness of news based on who shares it, rather than where it comes from. If they trust the sharer, they are more likely to “recommend the news source to friends, follow the source on social media and sign up to news alerts from that source” (p. 1). This is particularly the case for young people. Boczkowski and his colleagues (Boczkowski et al. 2017) found that much of the news young people encounter is ‘incidental news’, seen in the process of following their social feed rather than accessed from a trusted brand as part of a deliberate news search strategy. In this environment, news appears as part of an undifferentiated flow of social and informational content, its value mediated by social recommendations.</p>\n\n    <p>As the fake news moment testifies, some people share news to deceive. In analysing the false stories that she saw circulating online during the 2016 Trump election campaign, digital journalism researcher Clair Wardle found seven different kinds of ‘fake news’—or mis-, dis- and mal-information, as she more accurately labels it (Wardle and Derakhshan 2017). There were items that made false connections or presented false context, as well as manipulated, misleading, imposter and fabricated content, and satire or parody (Wardle 2017). However, it is a moot point whether much of this deceptive material stemmed from ordinary individuals with unhappy intentions. Subsequent academic research suggests that politicised institutional and economic rationales have been driving the circulation of deliberately deceptive news—not in the least the notorious Internet Research Agency of St Petersburg, once Russia’s most prominent trolling operation, which a Guardian investigation (Hern et al. 2017) showed was posting supposed witness accounts from disasters and crime scenes that were picked up by UK and US publications including the Telegraph, BuzzFeed, the Daily Mail, the Huffington Post and the BBC.</p>\n\n    <p>Journalists being fooled by imposter accounts is one problem, but a larger concern, which Wardle and Derakhshan (2017) highlight in a report for the Council of Europe, is whether and how ordinary social media users can identify a hoax post. The templating of social media items and their random appearance in our feeds, means that a post from a legitimate news source can look similar to a story from a conspiracy site, and equally shareworthy. They cite political scientists Messing and Westwood’s (2014) observations that social media content is decontextualised, since the focus is the story content, not the source. The Reuters Institute found that less than half of those they surveyed in 2017 usually recognised the brand of news content when they access news from social platforms (Kalogeropoulos and Newman 2017). Thus, without rich indicators of quality or origin, people are left to their social networks to help them identify and critique what’s authentic and what’s not.</p>\n\n    <p>Unfortunately, misinformation is far more likely to be recirculated than the truth because it makes a claim that is unexpected and contradicts what people know about the world. Its novel, unusual nature makes it more interesting to share (Vosoughi et al. 2018). Researchers at MIT’s Laboratory for Social Machines found that the average ‘fake’ news item took 10 hours to spread to 1500 users on Twitter, whereas a verified story took 60 hours to reach the same number of people. They also found it travelled “farther … deeper, and more broadly than the truth”, partly because of the strong emotional reactions people had to it (Aral 2018). This points to an interesting aspect of virality that will be explored further in Chaps. 6 and 8—that news we feel strongly about is more likely to be shared.</p>\n\n    <p>Research from the last decade indicates that whether we share news will depend on how we feel about what we’re viewing. As Jonah Berger (2011) showed in a ground-breaking marketing psychology study, we are more likely to share information that stimulates or arouses us emotionally, whether we feel positively or negatively about that content. His work with Katherine Milkman (2012), which suggests that arousal increases the speed of news sharing, triggered a wave of studies that have explored how our emotional or affective states influence the sentiment of news we share, and the rate at which it spreads. As Chap. 6 indicates though, this research has provided some confusing and contradictory results.</p>\n\n    <p>For a start, it is not clear whether emotional reactions make us more inclined to share positive or negative stories. Decades of research on legacy news distribution has suggested journalists and consumers pay more attention to bad news than good (Galtung and Ruge 1965; Glasgow University Media Group 1980; Trussler and Soroka 2014), a trend reflected in news sharing on Twitter (Hansen et al. 2011). Yet Berger (2013) has argued that if we have a choice we prefer to pass on good news rather than bad, and this is supported by a recent study showing Twitter users share ‘overwhelming positive’ viral news (Al-Rawi 2017). Meanwhile, another group who set out to determine whether everyday bad news is likely to travel faster than good found that it did—at least as far as Brexit news on Twitter was concerned (Fang and Ben-Miled 2017).</p>\n\n    <p>It is clear, though, that sharing news is very much an affective pursuit, an immediate reaction to socially significant events, issues and actions. It invites others in our social network to identify with or dispute ideas and helps us to mobilise others around causes or campaigns. Zizi Papacharissi (2015) notes affective publics form around networked expressions of sentiment about news, supporting large-scale, pluralised responses to it, and connective—if not collective—action on it. Social sharing is a move away from the controlled, centralised representative agendas of analogue news, which offered limited means to publicly express our feelings, in our ways about important changes in our worlds. As Chap. 6 will argue, sharing news enables users to place themselves in the news matrix, integrating or contesting the concepts, symbols and ideologies of news narratives in ways particular to themselves and their cultural identities.</p>\n\n    <h2 id=\"how-we-share-news-online\">How We Share News Online</h2>\n\n    <p>Just as our motives for news sharing vary, we have many ways to do the sharing. People will use any networked communicative tool they read news on to flag or repost the URLs of content they find interesting or valuable, and there are many competing systems designed to allow us to do that. Much news sharing is done via email clients, from secure forums or using messaging apps like WeChat, Whatsapp, LINE or Facebook Messenger. The problem is that these channels, together with a range of mobile social apps, such as those belonging to Reddit, Facebook and Instagram, don’t generate traffic referral information in their message headers. That means measuring and understanding how people share news via these ‘dark social’ avenues is a difficult pursuit, and one that has become more pressing with the upswing in mobile traffic over desktop access to news websites (Breaux 2015).</p>\n\n    <p>Dark social is a phrase coined by former Atlantic magazine deputy editor Alexis Madrigal to describe the large volume of traffic coming to his site from sources that he couldn’t track. As Madrigal (2012) noted at the time, this problem exposed not only the range of sharing activities outside social platforms, but one of the great shifts that these platforms had visited on web media consumption—visibility:</p>\n\n    <blockquote>\n      <p>the social sites that arrived in the 2000s did not create the social web, but they did structure it. This is really, really significant. In large part, they made sharing on the Internet an act of publishing (!), with all the attendant changes that come with that switch. Publishing social interactions makes them more visible, searchable, and adds a lot of metadata to your simple link or photo post. (Madrigal 2012)</p>\n    </blockquote>\n\n    <p>Social media has become the main way that we share news online, and are encouraged by news companies to share, because it makes our activity visible and measurable. There are several ways in which Facebook, for example, makes what you share more apparent. It first gives your post a permanent URL, so it is discoverable by search engines. It adds traffic referral data to any link you post, say to a news story. When someone clicks on your linked item, the metadata tells the publisher’s server where your visit came from. Facebook’s software also crawls the web to capture data about that site, and to auto-populate a preview image for that link, excerpting the content and making the share more attention grabbing. This last technique is designed to capitalise on the likelihood that visual material such as photos and videos is more engaging than text alone. Imagery makes your posts more shareworthy. Finally, social media provides a range of automated commendary tools so we can evaluate others’ posts. As a whole, these algorithmic processes greatly elevate social media’s role in promoting media sharing, and produce quantifiable data about the scale of post creation and engagement. In enabling news publishers to collect, aggregate and measure social sharing metadata, or referral data about visits coming from platforms to news sites, social media companies have constructed themselves as key players in the digital distribution of news. As Madrigal also notes, social media platforms are only the latest in a long succession of computational systems for sharing information and news (see Fig. 2.1), each of which presented its own political and economic battles over how we could share information, what was and wasn’t proper use of the system. To get a more complete understanding of these struggles to define the authorised boundaries of news sharing, it is instructive to traverse some of that history, starting with first popular distributed online system, PLATO.</p>\n\n    <p>Launched in 1960, the US educational computing platform PLATO connected display terminals at the University of Illinois to dozens of networked mainframe computers. The platform’s development spawned many concepts now familiar in social computing including forums, message boards and chat rooms (Dear 2017) but also saw intense arguments over what was legitimate (coursework) and non-legitimate (gaming) use of the network (Latzko-Toth and Jones 2017). In a similar sense, France’s successful Minitel videotex system, which delivered news, directory, gaming, ticketing and billpaying services to ordinary citizens across the country, as part of a government sponsored roll-out, was scandalised by the growth of ‘messageries roses’, adult oriented chat rooms (Mailland and Driscoll 2017). US commercial service Prodigy offered news, email and bulletin board systems (BBS), but no chat, and alienated its subscriber base with over jealous censoring of content (Banks 2008). The UK’s centralised Prestel service was simply too expensive and visually uninteresting for most subscribers (Lean 2016).</p>\n\n    <p>The most successful news sharing systems were those that democratised the hosting and exchange of news, like BBS, which allowed any computer hobbyist with a PC and modem to publish or connect with specialised networks via the telephone network (Fig. 2.3).</p>\n\n    <p>Usenet, which was developed in 1979 at Duke University so that people could exchange public messages between connected computers, expanded this concept beyond the national. Over its first decade, Usenet gradually linked more and more research institutions, and became host to an extraordinary variety of specialist discussion or ‘newsgroups’. There were struggles about the structural hierarchies, naming and control of these groups, and debates about what constituted permissible speech, all part of Usenet’s evolution as a public access network. it is remarkable for being the first international scale decentralised news sharing initiative, with the user-led alt. newsgroups still being the largest section of the system. Sharing information with other systems however, like the defence research oriented ARPAnet was not easy, with technical and administrative hurdles to inter-mailing list communications (Hauben and Hauben 1997). In 1982, users dreamed of a World Net “that would tie all sorts of computers worldwide together” (ibid.), a vision that would be realised in 1993 with the World Wide Web, a graphical language, system of resource locators and open transmission software protocol that would enable any computer users to share information with another over existing telecommunications networks.</p>\n\n    <figure>\n      <figcaption>\n        <p>Fig. 2.3 Online news sharing systems development timeline, adapted from Weber (2017)</p>\n      </figcaption>\n    </figure>\n\n    <p>Internet historian Marc Weber (2017) argues that we lack a deep historical account of many of these early networks which might help us compare factors that are key to the effective distribution of information: the efficiency of different linking systems, of strategies to reduce the incidence of misinformation. However, it is likely that the one core reason we chose to share news via privatised, controlled social media rather than the more openly owned and structured World Wide Web is automation.</p>\n\n    <p>The advent of the World Wide Web in 1993 gave every PC and modem owner the possibility of being an author-publisher, and with a graphical focus that brought a new, if low-resolution visuality to news sharing. Websites were the new way to share your interests, even if much of the content still consisted of URL lists pointing to interesting sites and documents, and blogrolls of favoured writers. Yet, while the web lowered the barriers to sharing information across the globe, beyond the territoriality of analogue media, it still required technical skill, and an understanding of computing and coding beyond the reach of most people (let alone the billion or so worldwide living without electricity).</p>\n\n    <p>Automating the web publishing process made it more accessible to more people. Geocities, the US-based webhosting service founded in 1996, lowered barriers to creating web pages with templates and neighbourhood directories, and later drag and drop publishing tools. Together with Angelfire, Tripod and much later MySpace, Geocities gave users a base from which to learn the principles of html coding and media uploading, but it was templating that simplified, expedited and fully democratised online news sharing. Publishing templates defined the layout and features of many blogging services from Blogger and Live Journal to WordPress, and the first social media services, Six Degrees in 1997, Lunar Storm in 2000 and Friendster in 2000.</p>\n\n    <p>In 2006, Facebook further revolutionised the automation of news sharing by enabling websites to instal a share widget or button to their pages, which automated the posting of content to their accounts or contacts, making it more “more efficient by giving people a simple structure to do it” (Facebook 2006). In 2009, it then introduced its version of the like button, originally introduced by Friendfeed in 2007 before Facebook acquired that company. Liking enabled people to react to others’ sharing activities, but only positively. Zuckerberg recognised the value of this positive social reinforcement in encouraging interaction when he rejected calls for a ‘dislike’ button:</p>\n\n    <blockquote>\n      <p>The like button is really valuable because it’s a way for you to very quickly express a positive emotion or sentiment when someone puts themselves out there and shares something. Some people have asked for a dislike button because they want to be able to say, “That thing isn’t good.” That’s not something that we think is good. We’re not going to build that, and I don’t think there needs to be a voting mechanism on Facebook about whether posts are good or bad. I don’t think that’s socially very valuable or good for the community to help people share the important moments in their lives. (Zuckerberg in Lafferty 2016)</p>\n    </blockquote>\n\n    <p>Share and react buttons are purposely designed technological affordances of Facebook that make acting on our impulses to pass on and respond to online information easy, quick and ‘frictionless’ (to use a term favoured in Silicon Valley). They encourage affective responses, but most obviously constructive ones as these are more likely to generate.</p>\n\n    <p>Facebook’s resistance to automating critical signals is just one example of <em>affective regulation</em>, the process of controlling and directing expressive responses during social sharing. Another is Facebook’s introduction during 2017 of a set of five new emoji reaction buttons—Love, Haha, Wow, Sad and Angry—to supplement the Like. These options were ostensibly introduced in order to reduce incidences of reaction avoidance, where users felt it was inappropriate to ‘like’ content that made them upset. However, Facebook also indicated it considered the new reactions ‘stronger’ signals of relevance than the original like, raising questions about how it weighs them in calculating post visibility, and what affect users might have on news feeds by choosing different types of emotional response.</p>\n\n    <h2 id=\"how-users-affect-what-is-and-can-be-shared\">How Users Affect What Is, and Can Be, Shared</h2>\n\n    <p>Users have changed the way platforms can be used to share news. Software developer and designer Chris Messina (2007) came up with the idea to use keyword ‘channel tags’ as a way to signal that a group discussion was opening, but keeping the signal short enough to be embedded in the message. Hashtags are now used to organise and aggregate shared content on Facebook and Instagram, illustrating the normalisation of socio-technical standards for sharing online. Tumblr users also adopted the GIF or ‘graphics interchange format’ to animate their personal reactions to posted content, drawing on an eclectic store of appropriated pop culture motifs. This playful convention was later adopted by digital-born news publishers like BuzzFeed, and GIFs are now being used by legacy media to give visual interest to stories and to animate simple infographics.</p>\n\n    <p>How we share news and what we share depends not only on a platform’s affordances, but also its commendary culture. This becomes clear in contrasting Reddit and Twitter, the international platforms most clearly focused on news sharing. On Reddit, users take part in discussion groups, called subreddits, devoted to particular topics, and news sharing involves excerpting, explicating and debating, to some extent, stories from existing media sources. The top 5 subreddits in terms of subscribers were humour (funny) world news, and educational (AskReddit, todayilearned, science). Users come from the US, Canada and the UK primarily, and in the US are largely young, 18–29-year-old males (Pew Internet 2016).</p>\n\n    <p>Twitter, on the other hand, is more cosmopolitan and gender balanced, with 57% male users and 42% females in total, although this ratio varies by country. 79% of Twitter users come from outside the US, including China, its biggest national market, and India its third largest, while many countries including the UK and Japan are heavier users than North Americans. While its users are diverse, Twitter operates mainly as a professional and expert communications network, with news sharing done primarily through hyperlinks, due to the messaging constraints (280 characters per post, double that of pre-2017 tweets). More than half of the top most retweeted messages in 2017 were motivational, including three from Barack Obama, and two disaster appeals. While there is no recent global survey of account types on Twitter, an Australian study found that aspirational accounts were one of the top four concentrations, alongside teen culture (music and celebrity especially), sport and netizen/technologists (Bruns et al. 2017). That research also showed that user interests had changed over time, with technologists being a more important group in Twitter’s early years, while teens and sports enthusiasts being later adopters.</p>\n\n    <p>Each evaluative system then demonstrates different politics of participation and recognition, based on its measures of social visibility, status and approval. On Reddit, which is a community of pseudonymous users, the title, visual content, placement and timing of a post will determine its visibility and popularity more than on Twitter, where real identities are more common and visibility is defined by numbers of followers and popularity of posts. A Twitter user’s profile and status, or social authority, has a distinct positive effect on content sharing. In terms of approval, each Twitter post accumulates a register of signals, including retweets, favourites, replies, follower requests and the addition of the user to expert lists. By way of contrast, Reddit users, in a quasi-democratic gesture, use up and down-voting to register their approval or disapproval of posts. As pseudonymity and anonymity pose particular issues of recognition and behavioural control for Reddit, this service has developed an additional reputation strategy. Users accumulate ‘karma’ points (a Silicon Valley gesture towards the Buddhist concept of an individual’s actions influencing their future), which then determine the degree of freedom they have to post information, the degree of attention paid to moderating their contributions, and their status within the system when users have low karma.</p>\n\n    <p>The sharing of politically contentious or illegal material can trigger internal or external regulation to control sharing behaviours, a subject we look at more closely in Chaps. 3 and 8. While the major social media platforms subscribe to the North American protection of speech freedoms, they nevertheless are obliged to control what we share, and how we speak about it in order to prevent the circulation of hate and discrimination, misinformation, propaganda and criminal activity. As Tarleton Gillespie argues in his investigation of content moderation online:</p>\n\n    <blockquote>\n      <p>Platforms must, in some form or another, moderate: both to protect one user from another, or one group from its antagonists, and to remove the offensive, vile, or illegal—as well as to present their best face to new users, to their advertisers and partners, and to the public at large. (Gillespie 2018: 124–125)</p>\n    </blockquote>\n\n    <p>What we share and how it frames our relationship to the legal and political frameworks of nation state and platform revenues then affects what we care able to share. States, for example, routinely step in to regulate news sharing, showing they acknowledge its power to shape public opinions and political processes. Countries like China, Iran, Turkey, and Vietnam control to varying degrees both access to social media and activity on the platforms, usually to prevent political dissidence, and mobile internet access is sometimes restricted in areas dominated by religious and ethnic minorities. Governments will also restrict access to social sharing to further their own political agendas, as when the Ukrainian government blocked access to the popular Russian-owned VKontakte social platform on national security grounds (Kelly et al. 2017). As Chap. 3 will argue, these battles over access to news sharing systems represent a major front in struggles for freedom of expression and media freedom.</p>\n\n    <p>What we share is also, as this book argues, fundamentally reshaping the news business, its relationship to the social media platforms and its audiences, and the production of news. To understand its impact on news journalism, we’ll now examine some of the factors that have affected social news sharing’s integration into newsroom routines, and its part in influencing how news is constructed.</p>\n\n    <h2 id=\"reshaping-the-news-the-likeable-engine\">Reshaping the News: The Likeable Engine</h2>\n\n    <p>In 2009, Richard Sambrook, then director of the BBC Global News Division, told the Oxford Social Media conference that he thought “the impact of social media was overestimated in the short term and underestimated in long term” (Bunz 2009). Unlike most media predictions, his was remarkably accurate. The legacy news media have not wholesale adopted BuzzFeed’s social first approach to production, or, as Jonah Peretti’s organisation did early on, tied wages to evidence of social impact, clicks, likes and shares. Yet, nearly a decade later, social media organisations and their users have had a profound influence on how news is identified, selected, produced, distributed, promoted and discussed.</p>\n\n    <p>At peak social sharing, before Google reasserted its dominance and Facebook changed its algorithm to demote news in favour of personal information from people’s feeds, industry reports suggested anywhere from 7–50% of news traffic would come from social media referrals, depending on the site’s business model and the story topic (Mitchell and Jurkowitz 2014; Parse.ly 2015). Those digital-born companies based on social first publishing strategies like BuzzFeed and Upworthy had referrals at the top of that scale. At this time, social media metrics were a proxy for digital news consumption. This conception drove the uptake of social analytics which we document in the next chapter, and a greater editorial focus on monitoring what was being shared online and why. Although there is now a healthy scepticism about the importance of social media for news making, and a move for news organisations to refocus on a more diversified approach to distribution (Cornia and Sehl 2018), it is worth remembering how and why newsrooms sought to integrate social media during this period.</p>\n\n    <p>The growth of commendary culture and the rise in social referrals to news sites saw news organisations become tightly bound into the social media ecology, investing in branded social media channels, advertising and revenue deals, hiring specialised social media teams, adopting digital workflows and analytics systems and exploring occasional partnerships with the platforms (Cherubini and Nielsen 2016). As Bruns (2018) suggests, journalists were under pressure to boost sharing and increase audience trust in news product via personal social media engagement with audiences. In turn, these social branding activities led to power inequities in the way some individual reporters could command public attention and mobilise audiences, and engendered a lesser hold by employers over the opinions and job mobility of those employees (Bruns, 222–223). So, news sharing affected production investment and innovation agendas, while destabilising (to some extent) existing labour models and relations.</p>\n\n    <p>From 2010 onwards, we were attempting to measure news sharing, first internal sharing of news across co-owned titles in Australia (Dwyer and Martin 2010) and then, together with our Share Wars colleagues, user-led sharing internationally. For the latter project, we used Share Wars’ Likeable Engine, a homegrown news analytics tool. For four years from 2011, Likeable tracked social news sharing trends using data from 140–160 major news sites worldwide and information feeds from Facebook’s and Twitter’s public application programming interfaces (APIs). Like other later commercial analytics applications like Chartbeat, Newswhip or Parse. ly, or the Guardian’s in-house developed Ophan dashboard, Likeable enabled editors and journalists to monitor changing patterns of social media news distribution and consumption (including their own relative performance), to track trending stories and to strategise live editorial decision making in competitive online news markets.</p>\n\n    <p>Likeable was created by the Share Wars group, employees of one of Australia’s top-rated online news sites Nine.com.au (formerly ninemsn). Share Wars was a start-up built by former news editor Hal Crawford, former Windows 8 editor Andrew Hunter and former Network Development Director Domagoj Filipovic. They sought to use its analytics capacity to discover “what makes stories shareable and how this new distribution force will change workflows, roles and resourcing” (Crawford et al. 2015). The project marked a period where company executives started to take an interest in the uses of computational intelligence in audience research via their creative agency Mi9, and as Nine Entertainment split with Microsoft Network. The founders, who have all moved on to senior executive positions in large news media organisations and Facebook, maintain a blog of the project findings at http://share-wars.com/.</p>\n\n    <p>The Likeable Engine had four core components:</p>\n\n    <ol>\n      <li>News Feed Processor (extracting story URIs from news websites)</li>\n      <li>Database (storing story data and matching it with social media data)</li>\n      <li>Like Processor (extracting commendary data stats from FB and Twitter)</li>\n      <li>Live reporting Interface (graphing the mashed data)</li>\n    </ol>\n\n    <p>Its architecture demonstrated the classic public/private nature of the social media ecology. It relied on public internet carriage and open standards web content (news sites), accessed public (but privatised) commendary information, formatted by proprietary social platform feeds, using bespoke scraping and querying algorithms and stored it using open source database software (Fig. 2.4).</p>\n\n    <p>Until its decommissioning in 2016, the Likeable Engine scraped new news story URIs linked from its sample news homepages every 11–20 minutes and then queried Facebook and Twitter APIs for the share counts of those stories. It requested these counts hourly for the first 24 hours, then less frequently for 5 days. Where possible URIs were canonicalised to their Facebook IDs, to avoid the same story being differently counted on its mobile or web domain. The interface then displayed the top ten stories graphed over 22 hours, and listed the top 100 with their share counts. Users could filter by region (AU, UK, US), by date and by satire (with an eye already to post-truth news).</p>\n\n    <figure>\n      <figcaption>\n        <p>Fig. 2.4 Likeable Engine information model. Image: Fiona R Martin AU14</p>\n      </figcaption>\n    </figure>\n\n    <p>The journalists interviewed in the early days of our project used Likeable as a complement to other paid analytics services such as Spike, Chartbeat, BuzzSumo, Omniture, and the in-built or native analytics provided by RedditInsight, YouTube and Google Analytics. Likeable users were primarily the ninemsn editorial staff, although users in several other workplaces, including Fairfax Media (which has since been taken over by Nine Entertainment), adopted the app. Interviews with five online journalists who used Likeable in two commercial newsrooms suggested that it played a useful and important role in their editorial routine when it was released, but had been overtaken by newer, more sophisticated analytics packages which they preferred.</p>\n\n    <p>The major insight from the interviews was the way in which analytics had become an important tool in editorial decision making. Interviewees said that they used analytics daily, and often throughout the day, to filter content, identify story ideas and trigger coverage, and to determine news scheduling and placement. Likeable was often consulted in the morning to identify story ideas based on what social media news was getting most user attention, where and when, to indicate whether journalists might need to produce reaction stories or explore new angles such as local perspectives. Two users regularly looked for ‘pre-viral’ stories or videos that were performing well on smaller news sites, and which had yet to be shared internationally. One sought news which “isn’t prominently placed on the original website, or from sites that are not part of the normal news cycle”.</p>\n\n    <p>Sharing analytics were also used to monitor the performance of competitor coverage of similar material. One news editor noted that Likeable helped him:</p>\n\n    <blockquote>\n      <p>… to monitor sharing across the day and look at what’s sharing well comparatively—also to think about domestically what might get a push from posting on social media.</p>\n    </blockquote>\n\n    <p>Analytics also emphasised the temporal flows of sharing, with trends suggesting times when stories might be posted so they “move better” and are shared more. One user noted he always looked at the top videos and what time they were posted “live versus the number of shares they get”.</p>\n\n    <p>Overall, the interviews suggested social analytics had five distinct uses in editorial decision making. They were regarded as:</p>\n\n    <ul>\n      <li>an evaluative system—to identify stories and topics that were sharing well, or which might go viral, and which stories should stay on the homepage</li>\n      <li>a news filter—to identify what topics were getting most attention, where and when</li>\n      <li>an alert system—to signal new stories that are sharing well on smaller sites, or the need for reaction stories to viral content</li>\n      <li>a scheduling device—to suggest times when stories might be posted to attract more shares</li>\n      <li>an audience filter—to indicate which on which publications and platform certain stories were being most often shared.</li>\n    </ul>\n\n    <p>These interviews also suggested how traditional new values are being reinterpreted relative to social sharing metrics. A story’s timeliness, which was once understood relative to the publication’s production lead time, is now also determined by whether and how long it has had a presence on social media. The need to ensure optimal news sharing is also determining whether stories are posted to social media or the World Wide Web first. The audience relevance of a story is being assessed by where its sharing is initiated (from the article page or via social platforms) and on which platform/s people are posting about it. Controversial or polarising topics tend to be highly shared, suggesting conflict remains an important news value, but exploiting that interest demands journalists include provocative keywords in story straps, excerpts and headlines to attract searches and shares. Shareable items as a whole sit on either end of the impact spectrum: they must be either novel and pre-viral, or ubiquitous, that is, viral and otherwise newsworthy. These evolving ideas about what makes news significant shows just how subtly and comprehensively social sharing analytics are influencing journalists’ normative interpretations and applications of news values.</p>\n\n    <p>While we didn’t directly ask journalists if they understood how these figures were compiled, they did voice a range of concerns about the transparency of the metadata capture from Twitter and Facebook, particularly how analytics tools weighted views against shares or retweets, and how they should interpret differences in sharing relative to publication readership or audience size. One user requested better indexing of Australian news sites. The majority wanted better graphing of sharing in real time, one indicating that Omniture provided the best model.</p>\n\n    <p>These journalists’ interest in tracking the rate at which news was shared was consistent with the news media’s obsession with speed in news production and distribution. It also suggests that social virality is a central factor in the adoption of social analytics services such as Chartbeat, and a rationale quite distinct to measuring user attention to stories or journalistic engagement.</p>\n\n    <p>Our research was never intended to fully map journalistic responses to social news sharing—far better covered in reports like Caitlin Petre’s <em>Traffic Factories</em> (2015) or Fedrica Cherubini and Rasmus Kleis Nielsen’s <em>Editorial Analytics</em> (2016)—but rather, we wished to trigger a critical enquiry into how social media news sharing is analysed, measured and valued. The next chapter takes up this story, particularly questions about the clarity and reliability of the data captured.</p>\n\n    <p>My newsroom encounter also highlighted the diversity of players in the social sharing ecology, with analytics companies based in Ireland, Israel and the US, and a distinct gap in the tracking of non-Western platforms like WeChat, Sina Weibo, LINE and Kakaotalk from our Asian region. This was a reminder that much news sharing knowledge still reflects the late twentieth-century cultural hegemony of transnational Western news domination. The deficiencies of this focus are obvious with the rise of BRICS countries. It was encouraging then to see, in December 2016, that Newswhip recorded IndiaTimes.com as the most shared site internationally on Facebook, part of a shift to it acknowledging greater diversity in the publishers who were achieving high share counts (Corcoran 2017). In an increasingly globalised communications world, we need to understand how our news sharing is shaped by different local, national and extra-national forces, and how commendary cultures might emerge in response to particular economic, political and social contexts.</p>\n\n    <p>The rise and fall of Likeable is one small part of that story. Likeable’s shutdown was partly the result of transnational competition, from analytics start-ups with more venture funding, more viable business models and better designed products, and partly due to Facebook and Twitter’s full commodification of what had previously been ‘freemium’ data subscription models. Likeable originally sampled data from the pre-2015 Facebook Developer API, which provided free access to separate counts of shares, likes, comments, and on-page comments—some of which were no longer available from the amended public API. It also drew on Twitter’s public data feed. This had been narrowing for years as the company slowly withdrew users’ permissions to access certain types of information, but it still provided tweet count data which was valuable to app developers and marketing services. Even if the free APIs had been somewhat unstable and subject to sudden change, they still enabled experimentation and innovation in the analytics space among a raft of small technology companies. The closure/commodification of both streams in 2015 undermined Likeable’s already limited functionality, and as it didn’t have a viable subscriber base, it was not a commercial proposition for Nine to buy that data from the platforms. Yet, Likeable had already fulfilled its ideological promise, introducing a group of journalists to the ground zero of analytics use, its incorporation into their news routines and the alteration of their production norms.</p>\n\n    <h2 id=\"commending-the-news\">Commending the News</h2>\n\n    <p>Social media news sharing has become a powerful new force in digital news production, in its variety of commendary forms and practices, its everyday adoption, scale of interaction and integration with news media production, distribution and consumption. Platform designs and their affordances operate to enable audiences to make visible the news they value as shareworthy, to marshal attention to their judgements, and to enlist others in responding to their reactions. Their evaluations, when tracked, analysed and represented, affect not only platform advertising, but journalism practice. Commendary culture influences the reporting and editorial practices of journalists in ways that are both obvious and also difficult to quantify, especially given tech companies’ relative lack of accountability for their calculus.</p>\n\n    <p>Social media platforms have, by promising informational and communicative agency, colonised the lifeworld of their users, and captured their commendary motivations and agency in ways and at a scale that have the potential to transform.</p>\n\n    <p>Read in light of this analysis, online sharing means much more than redistribution, communication or disclosure. It now encompasses the amplification of ideas, the speed and scale of reproduction, the ritual and affective nature of our acts and the differing levels of access we provide to our shared material. It also presages an era of machine learning and automated news discovery and redistribution.</p>\n\n    <p>There is little doubt that social sharing is reconfiguring news journalism, even where social referrals constitute a small percentage of traffic to a</p>\n\n    <p>news site. That is because social media users now routinely influence what gets seen by others, and what is valued in the news selection process, in a circular and self-reinforcing logic, just by redistributing and evaluating news stories. Their sharing metrics feed the editorial algorithms of the platforms, influencing how news is made visible and prominent and generating the metrics that newsrooms will monitor for signs they are producing what the audience wants. It is this commendary dynamic, quite apart from the gravitation of advertising revenue to the platforms, that is reshaping the future of news and how it will reach audiences.</p>\n\n    <h2 id=\"references\">References</h2>\n  </figure>\n</figure>\n",
 "content_hash": "sha256:66f11214c03d64e5defdc8b612e70faf29f63d314aa06cf5beb8f0a1cb21eede",
 "media": [],
 "changelog": [
  {
   "version": 1,
   "at": "2021-07-26T03:18:00Z",
   "note": null
  }
 ],
 "transclusions": []
}