Where There Is Social Media There Is Politics

Forthcoming in Routledge Companion to Social Media and Politics, 2016, (Eds.) Bruns A., Skogerbo E., Christensen C., Larsson O.A. and Enli G.S., NYC, NY: Routledge Where there is Social Media there is Politics Karine Nahon Abstract This chapter discusses the role of politics in social media – the power interplays among actors on social media as they attempt to promote their interests and values. It argues that social media cannot exist without some kind of political involvement: where there is social media there is politics; neutrality is the exception rather than the norm in social media. Next, it discusses the different manifestations this involvement and suggests classifying the social media politics into politics of architecture (platforms and networks structure) and politics of dynamics (networks structure, information flows, and curated flows). Power is exercised in each dimension in three key modes: influencing decisions, setting the agenda, and shaping stakeholder preferences and norms. Introduction The recent general elections in Israel, held on March 17, 2015, were won by the right. When examining empirically the content created by politicians on social media, it was clear that right-wing politicians exhibited higher levels of activity in terms of the number of posts, likes, and shares. However, users who were supporters of the center or left-wing parties would subsequently complain that they had been certain of a center-left victory, as indicated by their Facebook feed. It was full with posts, videos and images attacking the incumbent 2 right-wing Prime Minister Benjamin Netanyahu, who eventually won, and supporting the idea of replacing the government. The Facebook algorithm was largely responsible for the gap between their hopes and illusions and the electoral reality. Facebook presents users with only a small fraction of the information flows created by their friends (Constine, 2014). If I had 100 friends on Facebook, and they all posted at the same time, Facebook would show me only a few and I would not even know that the remainder posted as well. This reduced feed is even more bounded as Facebook prioritizes homophilous contents or those which one is more likely to agree (Pariser, 2012). These two practices of the Facebook platform are an example of how the self-selection power of users can be skewed. This gap between electoral preferences as reflected in social media and the actual preferences of Israeli society, in this case, is a distinct product of social media politics. “No idea is more provocative in controversies about technology and society than the notion that technical things have political qualities. At issue is the claim that the machines… can embody specific forms of power and authority” (Winner, 1996, 19). Politics of social media refers to the power interplays among actors on social media platforms, as they attempt to promote their interests and values. The concept of social media politics remains elusive despite the recent wealth of case and empirical studies on related topics, such as the bias of algorithms (Gillespie, 2010, 2014); power law of networks (Barabasi, 2003; Nahon, Hemsley, Walker, & Hussain, 2011); mediators control of information flows (Barzilai-Nahon, 2008; Lievrouw, 2009; Shaw & Hill, 2014) and user attention (Wihbey, 2014); agenda-setting in networks (Wallsten, 2007; Woodly, 2008); and the politics of protocols (Elmer, 2010), search engines (Halavais, 2008; Segev, 2010), and technology in general (Introna, 2006; Tufekci, 2014). These diverse topics have something in common: they are all different manifestations of the politics of social media. 3 Social media are the collection of web- and mobile-based platforms where individuals and groups interact. They include blogs such as Wordpress, update streams such as Twitter, general social networks such as Facebook, image-sharing platforms such as Flickr, location platforms such as Swarm, social news forums such as Reddit, business networks such as LinkedIn, and curation platform such as Pinterest. Given its unique affordances and rules, each platform is conducive to particular power dynamics. The politics of social media may have the power to affect the behaviors, preferences and value systems of individuals and groups according to the intentions of those wielding it. This may have significant consequences for the sort of information people receive, potentially shifting the gravity centers of meaning-making power. Researchers have observed some of these changes, for example through the skews and biases of information flows. However, they may largely remain obscure. The scope of this concept is broad, and this chapter does not pretend to provide a comprehensive typology. Rather, it represents an interdisciplinary attempt to enrich the conversation in the field while suggesting, defining, and classifying a number of ways to better understand the building blocks of the politics of social media. This chapter makes the normative claim that social media cannot exist without some kind of political involvement or bias. In social media, neutrality is the exception rather than the rule. This is followed by a description of the different manifestations of social media politics, a review of empirical evidence and discussion of several examples of this powerful phenomenon. As the literature on these topics is scattered, this chapter attempts to classify and organize the different types of political manifestations of social media. 4 The Neutrality Myth Shattered: Power Modes in Social Media Information technologies and social media in particular are not neutral artifacts but significantly political and social spaces. Power relations are fundamental to any society, whether mediated off- or online. Wherever there are people and social relationships there are power relationships (Castells, 2009). Politics does not reside in a vacuum, but in a social locus where actors (potentially) exercise their power. The first question we need to ask therefore is, are social media social? If so, to what extent? Trottier and Fuchs (2014) answer the first question in the affirmative, and identify three forms of sociality which determine the extent to which social media are social: cognition, communication, and cooperation or coproduction. Individuals have certain cognitive features that they use to interact with others, “so that shared spaces of interaction are created. In some cases, these spaces are used not just for communication but also for the coproduction of novel qualities of overall social systems and for community building” (p. 6). The special appeal of social media resides in their ability to not only host but also facilitate and enhance social interactions. This is the source of their impact, but at the same time also of their flux and complexity, as so many holders have a stake in them. Every decision about how technology is designed, how information is produced, shared, distributed or accessed, involves various stakeholders jostling for different normative positions. This dynamic, political process leads perforce to a struggle over hegemony of certain actors (individuals, institutions, groups or networks) over others, influencing their values and behaviors. How these issues are resolved will, thus, determine which people, under what circumstances, can do what on Facebook, Twitter or WhatsApp. The politics of social media, like any other technology, produces, reproduces, reinforces, and shifts power and privilege among designers and users, but also – increasingly – among non-users. Just like in the Roman 5 forum, the outcomes of the politics of social media promote the interests and values of the powerful. True enough, social media have empowered users (especially non-professionals) with readymade tools that enable them to share and distribute information, create viral events, enrich content through metadata, locate people with similar interests, develop applications, collaborate to produce knowledge, and build on others’ work and donations to create new things much more easily than before. This empowerment was accompanied by the illusion of social media being neutral, egalitarian, objective and democratic. In fact, however, the basic elements of social media – their architecture (platforms and networks structure) and dynamics (networks structure1, information flows and curated flows) – are political, non-neutral and non-democratic in design, practices, and policies. These elements are inherently biased by the values of particular stakeholders, which in and through social media regulate the others’ behavior in line with those values. Since politics is the exercise of power in the resolution of issues – which for our purposes may arise in or be addressed through social media – we need to define power in social media. I have discussed the notion of power in networks in depth in some of my earlier writing (Barzilai-Nahon, 2008; Nahon, 2011), and since this is not the main focus of this chapter, I will briefly present the three main power modes critical to the understanding of the phenomenon in question, and will later use them to inform some of its instances.2 The first power mode has to do with the capacity to influence the decisions of other social actors. This is aligned with the ideas of political scientists in the 1950s and ‘60s, such as Robert Dahl (1957) and Nelson Polsby (1963), who argued that decision-making analysis 1 As elaborated further below, the politics of networks structure can occur both on the architecture level and on the dynamics level. 2 For an extensive discussion of social power in general, see Lukes (2005). For a discussion of power in networks, see Castells (2009), and a special volume dedicated to network theory and power in the International Journal of Communications (Vol. 5, 2011). 6 would be the best way to determine which individuals and groups have more power in social life, and that decisions involve direct, that is actual and observable, conflict. For example, in the context of social media, Liu et al. (2011) found in a qualitative study that 36% of content remained shared with the default Facebook privacy settings. They also found that in the majority of cases there was a gap between the desirable privacy settings and the actual controls users decided to apply. In most cases, users exposed content to more users than expected, aligned with the Facebook’s vested interest to exercise power over its users, or more precisely influence their decisions regarding what content to expose, and to whom, in service of its business model. The second major mode of power is the shaping of and control over the political agenda, which determines how social media are designed and operate, and the how potential issues are kept out of the political process and public spheres. The conceptualization of the second mode of power is informed by political scientists of the 1960s, ‘70s and ‘80s such as Peter Bachrach and Morton Baratz (1962). Any satisfactory analysis of power thus involves not only examining social media’s influence on decisions (the first mode of power), but also examining non-decisions (suppressing challenges to the status quo or adding new issues to an agenda) as decisions. For example, Twitter restricts tweets to 140 characters. This design has had major ramifications on the content flowing across the service. When users are limited to 140 characters, their posts must be short, laconic, and simplistic, if not outright blunt. It is no coincidence that Twitter is mainly used for live event updates. It has been purposely structured this way by its designers, imitating SMS practices and consequently appropriating this user behavior as a tool for sharing activities in the immediate present (Sagolla, 2009). By constraining the agenda (to 140 characters), Twitter has privileged a particular type of content (real-time posts) over other content, such as complex and nuanced arguments. 7 The third mode of power focuses on actions and inactions aimed at shaping and influencing one’s perceptions, cognitions, and preferences (latent or manifest). This is done for example by securing acceptance of the status quo since no alternatives appear to exist, or because it is seen (but actually shaped) as “natural”, unchangeable, or favorable. This conceptualization is well aligned with Lukes (2005) and to some extent also with Foucault (1977; 1980) who argued that power is the ability to shape the mind and construct meaning to phenomena. While this mode of power is difficult to observe, it has the strongest impact among the three, as the changes occur within the social actor (person or group) affected by it. It involves a profound transformation of value systems, making Social Actor A believe and choose to act in a way that reinforces the system’s bias, thereby promoting the interests of Actor B at her own expense, usually in the form of compliance. For example, Bond et al. (2012) studied the effect of the iVote button added by Facebook in order to encourage voting in the 2010 US congressional elections on the actual behavior of 61 million users. The study showed that the button directly influenced the “political self-expression, information seeking and real world voting behavior of millions of people. Furthermore, the messages not only influenced the users who received them but also the users’ friends, and friends of friends” (Bond et al., 2012, p. 295). Critical voices raised the concern that the use of the iVote button by Facebook was not transparent, and could be exploited to create social pressure on particular groups and exclude others (e.g. by providing the iVote button only to people identified by Facebook as Democrat supporters). Be that as it may, this study exemplified the ability of social media politics to change the preferences of people who were reluctant to vote, and mobilize them to vote. One way to enrich the debate on the spectrum of social media politics is by classifying it into social media architecture and the dynamics (see Figure 1). The three modes of power are exercised in each one of these dimensions by influencing decisions, setting the 8 agenda, or shaping stakeholder preferences. The overlap between the dimensions is minimal, as the politics of dynamics focuses directly on content, while the politics of architecture focuses on technology and only indirectly on content. Both, however, aim at changing people’s behavior, preferences, and values. Figure 1: Politics of Social Media Dimensions Platforms Politics of Architecture Networks Structure ructure of Networks Networks Structure Politics of Dynamics Information Flows Curated Flows Politics of Architecture The architecture of social media – the design of the actual components that make up social media (hardware, software, operating and network systems) and their interrelations – is ultimately based on code written by developers. This code is constructed hierarchically: similarly to any other human language, where words come together to form a sentence, and many sentences form a text that expresses an idea. The lines of code form the algorithm, or set of rules for making the technology “do something”. The algorithm is at the heart of any automated or semi-automated technological process in social media: from making the 9 hardware react to users pressing on a button to unfriending someone with a click. More important for our current purposes, the architecting of social media is a conscious, willful, non-neutral act by various stakeholders, usually designers or developers, but also users. The struggle over who gets to architect the platforms and affordances, and how they are architected, is one of the key manifestations of the power struggles and political arrangements in social media. Technocrats tend to argue that because technology is based on algorithms devoid of human interference, it is able to construct consistently neutral and non-discriminatory processes. However, by the very fact that humans design it, every technology is inherently political, involving values and interests cast in the image of its architects and subsequently shaped by its users. Architecting social media is a conscious act of exercising power in its three modes: influencing decisions, setting the agenda, and shaping preferences of stakeholders around questions of affordances and use. Through the design of infrastructure architects influence the decisions of users regarding their privacy, what to share, how to write, and how to behave on that platform. Through the design of infrastructure architects determine the rules and boundaries of users’ speech, behavior on their platforms: what to write, what types of photos to upload or not, what types of videos to share. And through the design of infrastructure architects not only change the decisions and boundaries of behavior, but also shape the preferences and values of users according to their own interests. One example of controversy manifesting a power struggle between users and the platform is the issue of nude photos. Facebook has a policy of removing such photos. However, users have complained that this policy also censors breastfeeding, nudes in art, naked mannequins, and kisses of same-sex individuals. In January 2015, a group of mothers protested online by uploading breastfeeding images. Consequently, in March 2015 Facebook 10 announced it would no longer remove such images, as they did not violate its rules on nudity. The updated policy states that: We restrict the display of nudity because some audiences within our global community may be sensitive to this type of content.... In order to treat people fairly and respond to reports quickly, it is essential that we have policies in place that our global teams can apply uniformly and easily when reviewing content…. We remove photographs of people displaying genitals or focusing in on fully exposed buttocks. We also restrict some images of female breasts if they include the nipple, but we always allow photos of women actively engaged in breastfeeding or showing breasts with post-mastectomy scarring. We also allow photographs of paintings, sculptures, and other art that depicts nude figures (“Facebook: Community Standard Page,” n.d.) Studies on how the architect's role introduces valence into technology, and social media in particular, have flourished in the last decade. These include analyses of censorship policies of Facebook and YouTube and their changes over time, and their impact on user behaviors (Gillespie, 2010; 2014); as well as studies on biases in Google content provision and ranking for different localities (Halavais, 2008; Segev, 2010). Gillespie (2014) articulates six dimensions of the political valence of algorithms, which he uses as a heuristic for considering the scope of the “politics of algorithm”: (1) the choices (of those who architect the algorithms) behind the platforms’ inclusion/exclusion rules; (2) the collection of information not necessary for the algorithm to operate; (3) the level of obscurity of what is relevant; (4) the way the algorithm’s technical character is positioned as an assurance of impartiality; (5) the technology’s appropriation for purposes of political contest; and (6) how the algorithmic presentation shapes the public’s sense of itself. Gillespie argues that the algorithm, presented as objective results of queries, shapes users’ practices and serves as a legitimization apparatus for the production of biased knowledge. 11 While it is important to focus on the platforms’ architects as those who determine the values of social media affordances, as Gillespie suggests, the power struggles ecosystem involves more than just a unilateral, omnipotent type of actor. It involves many other actors: users and non-user individuals, groups, and ephemeral networks of people and institutions. Each of these stakeholders seeks hegemony by pushing their own value sets and interests in the context of the various debates arising in or addressed by social media. Winner suggests that one way for an artifact to contain political properties are “instances in which the invention, design, or arrangement of a specific technical device or system becomes a way of settling an issue in the affairs of a particular community” (1986, 2). How these issues are resolved and by whom will determine the extent to which social media will reproduce power structures and biases, or enforce new structures of power in an attempt to regulate user behavior. When social media users appropriate (or attempt to appropriate) a technology by altering the algorithm to their purposes or by using the same algorithm for purposes other than intended – this is a political act. This is when a latent power struggle occurs between the original architects and the users. Nevertheless, platforms and architects have an inherent power advantage when it comes to determining affordances on their platforms. They can decide to change in the design unilaterally, arbitrarily and again – non-neutrally. Users can invoke a power struggle, but in a manner somewhat reminiscent of labor struggles, this would require them to unite in order to create a critical mass that can counter the values imposed by the architects. Two examples illustrate this politics of platforms: one has to do with the decisions of several leading platforms to censor any explicit illustrations of the beheadings carried out by ISIS. These guidelines came after the beheading of American journalist James Foley in 12 August 2014. On August 20, the CEO of Twitter, Dick Costolo, tweeted: “We have been and are actively suspending accounts at we discover them related to this graphic imagery” (Costolo, 2014) . Consequently, Twitter changed its policies to allow family members to request the removal of content depicting a deceased user. Unlike Twitter, YouTube did not feel required to make any specific changes, as its community guidelines already included specific rules against content that incites violence or depicts violence with the intent of causing shock. This example shows however, that platforms do not hesitate dictate their values on critical questions such as the boundaries of freedom of expression. Another example for politics of social media is when Facebook introduced the “Year in Review” app by the end of 2014. This application invited members to watch and share the important moments of their life over the past year with the default caption “It’s been a great year! Thanks for being a part of it”. Soon users began complaining on the inadvertent algorithmic cruelty (Meyer, 2014), which reminded events they did not want to remember, such as death and divorce. Here, Facebook exemplified the third power mode by proactively impacting the memory and awareness of people, sometimes against their choice. Networks Structure: Between Politics of Architecture and Politics of Dynamics The structure of networks encompasses two main aspects. First, a conceptualization derived from the social sciences denoting the rules, practices and arrangements through which the behavior of people is regulated in networks (Bourdieu, 1977; Durkheim, 1982; Foucault, 1978, 1990; Giddens, 1986; Weber, 1946). These rules can be social rules determined by a group of people, but can be also be created by algorithms. Both interpretations refer to regulating users’ behavior in networks, and both are related to particular manifestations of politics. For example, McKelvey (2010) studied the conflict around two types of algorithms: 13 quality-of-service and end-to-end algorithms. Each type offers a different solution for the network neutrality issue: the former prioritizes preventing network congestion while the latter focuses on providing equality between communication modalities. Accordingly, each promotes different values and the dynamics of promoting these values is a manifestation of the politics inherent in the networks’ structure. A second aspect of the network structure refers to “the typology of interconnected nodes” (Castells, 2009), identified by “the observed set of ties linking the members of a population” (Watts, 2004, 48) in their social networks. This definition complicates our discussion of politics, as social media offers different types of structures. Bruns and Moe (2014) classify these structures on Twitter into three types: meso (follower-followee networks), macro (hashtagged exchanges) and micro (@reply conversations). More generally, there are two important types of networks’ structures relevant to our discussions. The first is a more permanent and stable model of network structure directly related to the platform’s architecture. For example, the network of Twitter followers or Facebook friends of a single social media account. While their number changes, it is a slow change and the boundaries of the network are clearly identified. There is little room for power dynamics here as the network structure is determined to a large extent by the platform designers. An example for the politics of network structure at the architecture level is the Internet backbone, which serves as the physical basis for the social media operation, and refers to the core routers that connect large networks, including network access points that control traffic between countries. The competition for joining the exclusive group of backward network providers is a political, not just a business one, as it affects the scope of control these institutions have. 14 Ephemeral types of network structures are the shapes and patterns we see in the links connecting people in social networks as information flows on topical issues (e.g., a conversation around a particular hashtag). Network structures like that are constituted dynamically around a topic and their boundaries and members are dynamic. Importantly, hashtags were not originally a Twitter design feature. Norms around their use emerged out of the collective practices of users. The use of hashtags later spread to other platforms, and they are now commonplace. The hashtag is part of the network structure in that it functions as a classifier that allows other people to follow specific conversations or topic. It evolves as information flows, and dissolves as that flow fades. Politics of networks structure with ephemeral nature should be classified as politics of dynamics. In summer 2014, Israel launched a military operation in the Gaza Strip against the Hamas Rule, operation protective edge. The event was tweeted using different hashtags like #IsraelUnderFire and #GazaUnderAttack, representing competing narratives. This exemplified politics of topical networks structure in an attempt to capture the attention of users around the world and influence their awareness. Politics of Dynamics Most of the literature on the politics of social media focuses on instances in which algorithm writers and platform providers introduce values into social media components, or on the way architects shape policies and standards. However, the politics of social media is not just about the architecture. It is also about forms of power, which operate as the dynamics of interactions between social actors evolve. These relationship dynamics are revealed to us, as researchers, through the information as it flows or is curated in social media. 15 Politics of Information Flows In the information age, the ability to control the flows of information is a significant source of power. The politics of information flows refers mainly to the conflict around how information, the most critical resource in social media, is shared and distributed among users. Studies have repeatedly demonstrated the formation of skewed information flows in social media, which result in unequal distribution of attention, and unequal impact on behavior and preferences (See for example, Nahon & Hemsley, 2013; Wihbey, 2014). Not every skewed information flow is a result of a political intervention. For example, homophilous patterns found in conversations in social media platforms may form by users independently of any intervention by their platforms. However, collective patterns of behavior (manifested by the clustering of information flows) rarely evolve without any political intervention, let alone political implications. In this section I focus solely on patterns that are driven or exploited by social actors, which I consider as political dynamics. A systematic review of the politics of information flows should include a discussion of (1) mediators or gatekeepers, the actors who control of information flows, and (2) clustering effects that are the product of information flow politics. (1) Mediators. Network gatekeepers have a tremendous impact on information flows: by choosing which information can or cannot pass, by connecting networks or clusters to one another, or more generally by regulating the movement of information as it flows. They can impact the chances of one video getting millions of views, while millions of other videos will receive only few. Network gatekeepers (people, collectives, or institutions) are those with the discretion to control information as it flows in and among networks. However, their power is not absolute and their impact depends to a large extent on the gated – those subjected to their 16 gatekeeping – and on the power dynamics with other network gatekeepers (Barzilai-Nahon, 2008; 2009; Nahon, 2011). Network gatekeepers are social actors that control information as it flows, so by definition they exercise power and are therefore political actors. “Actors in this system are articulated by complex and evolving power relations based upon adaptation and interdependence. They create, tap, or steer information flows in ways that suit their goals and in ways that modify, enable, or disable others’ agency, across and between a range of older and newer media settings” (Chadwick, 2013, 157). Therefore, a major power struggle in social media is over the number and identity of gatekeepers or mediators. Technological improvements have immensely enhanced the individual user’s ability to both produce and disseminate data. Despite this ability, however, true control of information flows still lies in the hands of a small number of mediators. The huge amount of information produced every second, as well as the need to create, share, and read content, require the user to rely on their services. They help users in all their activities in social media, from filtering excess information through connectivity with others to producing new content. We rely on Google to find what we are looking for, on social media opinion leaders to keep us posted, or on Facebook and Twitter to show us the posts uploaded by our friends. By definition, however – otherwise this mediation service would be of little practical use – Facebook, for example, does not show us all of our friends’ posts, but only those it selects. In return for this service, it gets to control the agenda of the information transferred from one subscriber to another. The struggle over the number and identity of gatekeepers or mediators is a struggle for controlling the agenda of the information conveyed and transferred from one person to another. While traditional gatekeeping focuses mainly on selection (e.g. by newspaper editors), network gatekeepers have many additional information control mechanisms. The power of 17 network gatekeepers does not necessarily reside in their ability to stop or filter information as it is transferred. On the contrary, it is concealed in their ability to link networks together, allowing information to travel far and fast, and to connect people to information and ideas. Attracting users’ attention is the name of the game for network gatekeepers (Wihbey, 2014). Content will spread if people know it is available to be spread, and mediators bring content to the attention of those who follow them. They become network power hubs. (2) Clustering Effects. Skewed clustered information flows can occur for a number of collective behavior effects, including power-law and follow-the-herd tendencies, homophilous tendencies and polarization. Research has demonstrated that linking entities in social networks follows a power-law distribution, where a few elites receive the attention of many and thus have a disproportionate amount of influence (Adamic & Glance, 2005; Drezner & Farrell, 2008; Karpf, 2008; Nahon et al., 2011; Wallsten, 2011). Scientists have shown that the structure of networks plays an important role in how, and to what degree, information spreads. For example, Barabási uses mathematical models to show that many social networks are scale-free, which means that the number of connections between individuals follows a power-law distribution (a few nodes have many connections while most have relatively few), supporting the idea that the attention of the masses is concentrated on a few influential actors. It turns out that a power-law distribution of attention or linking is a fairly normal social pattern evident both online and offline. Of course, capturing the attention of others may later translate into the ability to influence them (Nahon & Hemsley, 2013). Figure 2 illustrates this increasingly uneven distribution through the growing market share of the top four search engines from 2002 to 2015. From 2010 they have captured the attention of more than 98% of users. While the identity of the four top has changed with time (AOL and MSN were replaced by Baidu and Bing after 2010), the main question we need to 18 address is what this implies in terms of the variety and value bias of the information we are exposed to. These leading search engines exercise all three forms of power. Specifically, through their responses to billions of queries they control the personal agendas of every one of us: from how to find a particular plumber to what type of news to read. What information to exclude or include and with what priority are clearly manifestations of power over information flows and political decisions. The power-law tendency is also evident in other types of social media: in the blogosphere with dominant of blogs such as Huffington Post, in micro-blogging with Twitter, or in general social media with th e dominance of Facebook – making the social media landscape significantly political. Figure 2: Four Top Search Engines’ Market Share 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Other Search Engines Four Major Search Engines *Source: Alexa, Netmarketshare, Search Engine Watch The power law is just one dominant information flow tendency with political implications in social media. Empirical studies have also demonstrated that social networks 19 also induce homophily, fragmentation, and polarization (Benkler, 2006; Sunstein, 2009). In particular, homophily, the tendency of people with similar attributes to associate with each other more frequently than they associate with others (Lazarsfeld & Merton, 1954; McPherson, Smith-Lovin, & Cook, 2001), has long been recognized as a factor in linking behavior. More recently, it has been shown to be statistically confounded with influence (Shalizi & Thomas, 2011), meaning that statistics cannot be used to differentiate between similar behavior due to homophilous linking or due to the influence of Actor A over B. Scholars who have attempted to quantitatively distinguish homophily from personal influence (e.g. Aral, Muchnik, & Sundararajan, 2009; Centola, González-Avella, Eguíluz, & San Miguel, 2007) have been refuted by prominent statisticians who show that the two processes are generically confounded (Shalizi & Thomas, 2011). Thus, a complex relationship exists between homophily and influence. Homophilous links in social media arise because people interact with similar others. As they interact over time they co-influence each other and become more similar. Thus, over time, homophily changes the group structure, in a process indistinguishable from social influence (Centola et al., 2007). Homophily is fundamentally a mechanism of selection, but at the same time it is also a mechanism of (albeit latent) influence at the individual and group levels. It is induced by social structure and, in turn, influences those structures in what Centola et al. (2007) refer to as a co-evolutionary model. Going back to our example at the beginning of the chapter on the Israeli elections, users’ homophilous tendency was exploited and intensified by Facebook, resulting in a more skewed clustering of information flow, and ultimately in an illusory electoral reality. 20 Politics of Curated Flows “Data and data sets are not objective; they are creations of human design. We give numbers their voice, draw inferences from them, and define their meaning through our interpretations. Hidden biases in both the collection and analysis stages present considerable risks” (Crawford, 2013). The vastness of information in social media is created and shared by users, not the platforms themselves, and continues to explode on a daily basis. These big data are the basis of social media platforms’ business model. Many stakeholders engage with this information: users, platforms, third-party companies, governments, researchers and more. Politics also occurs after information flows, at the curation phase. “To curate, historically, has meant to take charge of or organize, to pull together, sift through, select for presentation, to heal and to preserve. Traditionally reserved for those who worked with physical materials in museum or library settings, curation today has evolved to apply to what we are all doing online” (Mihailidis & Cohen, 2013). For the purposes of the current discussion, I would like to borrow the term “curated flows” from Thorson and Wells (2015), which refers to curation in the broadest sense: to select and organize, to filter abundance into a collection of manageable size, to search, reframe and remix, or in short manage information flows after they have flowed in, particularly, social media. Being a participatory space, social media empower users to perform functions previously reserved to professional curators, such as archiving, annotating, appropriating and recirculating real-time information (Jenkins, Purushotma, Weigel, Clinton, & Robison, 2009). In the social media and big data era, there are suddenly many stakeholders who curate content, raising many different information issues: What is the most appropriate way to harvest information flows? What are the ethical considerations regarding privacy that need to be addressed when archiving public data? Who is responsible if a post is taken out of context? Who can access this data? The resolution of these and other emerging issues is a 21 political act, given that, interpretations and behaviors that rely on curated flows are inherently biased as they depend on what and how we collect from these information flows, how we clean the noise, how we understand the data, and most importantly, what our power motivations are. For example, interpreting the impact politicians have on social media will depend on multiple decisions: the hashtags or keywords we collect, the languages, the platforms, the technical constraints (such as the API), and the way we understand the context. Each of these will require a political decision, no matter whether it is manually or automatically operated. Someone behind the scenes will have to decide about how to curate flows, and this decision will impact on others. Politicization of curation occurs in small amounts of data and all the more so in big data. Expecting curation of flows to be a neutral, objective and accurate process in big data because they are too big for humans to handle them directly is a myth (boyd & Crawford, 2012). In another key example, Crawford and Gillespie (2014) show how the flagging mechanism for reporting offensive content on social media platforms “serves both as a solution to the problem of curating massive collections of user-generated content and as a rhetorical justification for platform owners when they decide to remove content”. In practice, it is a political mechanism of negotiating contentious public issues between users, groups, moderators and platforms which attempt to promote certain values by reconciling “their ability to directly know or quantify community values with their practical and rhetorical value” (p.3) Power struggles in Wikipedia will serve as our final example for a political struggle around curated flows. Its stated mission is to provide a free encyclopedia that people all over the world can use and contribute to. However, multiple studies have shown that it is rife with 22 struggles around determining why and how content is included that have, for the most part, taken place behind the scenes and far from the public eye. These dynamics occur among editors who represent different value systems. In January 2006, a Wikipedia entry item was created to document the controversy surrounding the publication of a cartoon depicting the prophet Muhammad by a Danish broadsheet called Jyllands-Posten, which in return sparked a controversy around whether the original cartoons should be published, or made available via thumbnails or links. When editors involved in the article’s initial creation decided to republish a large thumbnail version of the original cartoons at the top of the article, many Wikipedia readers and editors objected to this as unnecessarily inflammatory. After a power struggle among editor groups, those in favor of keeping the images on the top won the day, claiming that a consensus decision had been reached (Morgan, Mason, & Nahon, 2011). Conclusion This chapter discussed and demonstrated the pervasiveness of the role played by politics in social media. Social media politics represents conscious and frequent acts whereby multiple actors exercise power to wrestle for competing claims on issues relevant to the governance and use of social media, but also issues with political ramifications outside of the social media realm. Lately there has been a an important discourse raising growing concerns that most decisions related to social media algorithms are not only non-neutral, but heavily biased politically, leading to the conclusion that “Cyberspace never was – and never could be – independent from the governing institutions, economic structures, and culture and social worlds that gave rise to it”(Kreiss, 2014, p. 133). This chapter attempts to add a further refinement to this discourse deconstructing the general and opaque notion of algorithm and 23 focusing on the locus of the political process: whether it occurs while information flows or curated, or while the basic elements of a social media platform are architected (see Figure 3). Manifestations of politics can be identified in social media architecture (on the platform and networks structure levels) and dynamics (on the networks structure, information flows and curated flows levels). In all instances, the product is the same: a constant attempt to regulate human practices and norms, and moreover, to transform or reinforce the behavior, preferences and values of individuals and groups according to the worldview and interests of those in power. Figure 3: Politics of Social Media Relationships Politics of Dynamics Structure of Networks Information Flows Curated Flows Politics of Architecture Information Structure of Networks Infrastructure Behavior & Preferences Platforms The politics of platform architecture is different than the dynamic politics of information flows. Whereas the former mainly involves infrastructure as a mechanism to regulate behavior, the latter mainly involves the content of information flows. The legitimacy of information to flow is derived mainly from users who share content. In contrast, platforms receive their operational legitimacy from the false belief in procedural justice, the myth that platforms are neutral. 24 For example, there is a difference between the politics around allowing to post nude photos on a particular platform, and that of making it go viral or preventing it from being shared at all. It is a subtle but important difference. The politics of information flows entails a greater number of stakeholders, and is more inclusive in terms of the power at play, since it usually does not reside in a single platform. Deconstructing the politics of social media also allows the discourse to move from examining at specific dominant actors such as platforms and governments to a more nuanced evaluation of interrelations that also involve users and non-users. However, identifying the political manifestations is only a small fraction of the story. Most of the political processes occur without any serious public discussion about the values at stake. Questions such as the right to be forgotten by Google; what governments may collect about me on social media; or whether nude photos should be censored inevitably receive different answers from different people. What holds for one community doesn’t hold for another sharing the same network. This messy ecosystem creates a political vacuum in which platforms gets to play the police (sometimes against their will) while users look for creative solutions to circumvent policing. Political struggles and actions are determined without thorough scrutiny by all relevant stakeholders, rather than just the powerful. Where there is social media there is politics. 25 Acknowledgements I would like to express deep gratitude to Ami Asher for helping me with the language editing of this chapter. Also, I thank Shawn Walker and Alison Wohlers for providing comments on earlier drafts of the chapter. References Adamic, L. A., & Glance, N. (2005). The political blogosphere and the 2004 U.S. election: divided they blog. In Proceedings of the 3rd international workshop on Link discovery (pp. 36–43). New York, NY, USA: ACM. http://doi.org/10.1145/1134271.1134277 Aral, S., Muchnik, L., & Sundararajan, A. (2009). Distinguishing influence-based contagion from homophily-driven diffusion in dynamic networks. Proceedings of the National Academy of Sciences, 106(51), 21544–21549. http://doi.org/10.1073/pnas.0908800106 Bachrach, P., & Baratz, M. S. (1962). Two Faces of Power. The American Political Science Review, 56(4), 947–952. http://doi.org/10.2307/1952796 Barabasi, A. L. (2003). Linked. Plume New York. Barzilai-Nahon, K. (2008). Toward a theory of network gatekeeping: A framework for exploring information control. Journal of the American Society for Information Science and Technology, 59(9), 1493–1512. Barzilai-Nahon, K. (2009). Gatekeeping: A Critical Review. Annual Review of Information Science and Technology (ARIST), 43, 433–479. Benkler, Y. (2006). The wealth of networks: how social production transforms markets and freedom. New Haven: Yale University Press. Bond, R. M., Fariss, C. J., Jones, J. J., Kramer, A. D. I., Marlow, C., Settle, J. E., & Fowler, J. H. (2012). A 61-million-person experiment in social influence and political mobilization. Nature, 489(7415), 295–298. http://doi.org/10.1038/nature11421 Bourdieu, P. (1977). Outline of a Theory of Practice. (R. Nice, Trans.). Cambridge University Press. boyd, danah, & Crawford, K. (2012). Critical Questions for Big Data. Information, Communication & Society, 15(5), 662–679. http://doi.org/10.1080/1369118X.2012.678878 Bruns, A., & Moe, H. (2014). Structural layers of communication on Twitter. In K. Weller, A. Bruns, J. Burgess, M. Mahrt, & C. Puschmann (Eds.), Twitter and Society (Vol. 89, pp. 15–28). New York: Peter Lang. Retrieved from http://eprints.qut.edu.au/66324/ Castells, M. (2009). Communication power. Oxford University Press, USA. 26 Centola, D., González-Avella, J. C., Eguíluz, V. M., & San Miguel, M. (2007). Homophily, Cultural Drift, and the Co-Evolution of Cultural Groups. Journal of Conflict Resolution, 51(6), 905 –929. http://doi.org/10.1177/0022002707307632 Chadwick, A. (2013). The Hybrid Media System: Politics and Power. Oxford University Press. Constine, J. (2014). Why Is Facebook Page Reach Decreasing? More Competition And Limited Attention. Retrieved from http://social.techcrunch.com/2014/04/03/the-filtered-feed-problem/ Costolo, D. (2014, August 20). We have been and are actively suspending accounts at we discover them related to this graphic imagery. Retrieved from https://twitter.com/dickc/status/502005459067625473. Crawford, K. (2013, April). The Hidden Biases in Big Data. Harvard Business Review. Retrieved from https://hbr.org/2013/04/the-hidden-biases-in-big-data Crawford, K., & Gillespie, T. (2014). What is a flag for? Social media reporting tools and the vocabulary of complaint. New Media & Society, 1461444814543163. http://doi.org/10.1177/1461444814543163 Dahl, R. A. (1957). The concept of power. Behavioral Science, 2(3), 201–215. http://doi.org/10.1002/bs.3830020303 Drezner, D., & Farrell, H. (2008). The power and politics of blogs. Public Choice, 134(12), 15–30. http://doi.org/10.1007/s11127-007-9198-1 Durkheim, E. (1982). Rules of Sociological Method. Free Press. Elmer, G. (2010). Exclusionary Rules? The Politics of Protocols. In A. Chadwick & P. N. Howard (Eds.), Routledge Handbook of Internet Politics (pp. 376–384). Taylor & Francis. Facebook: Community Standard Page. (n.d.). Retrieved from https://www.facebook.com/communitystandards Foucault, M. (1978). Discipline & Punish: The Birth of the Prison (2nd Edition). Vintage. Foucault, M. (1990). The History of Sexuality. (R. Hurley, Trans.) (Fifth or Later Edition). Vintage. Giddens, A. (1986). The Constitution of Society: Outline of the Theory of Structuration. University of California Press. Gillespie, T. (2010). The politics of “platforms.” New Media & Society, 12(3), 347–364. http://doi.org/10.1177/1461444809342738 Gillespie, T. (2014). The Relevance of Algorithms. In T. Gillespie, P. Boczkowski, & K. Foot (Eds.), Media Technologies (pp. 167–194). Cambridge, MA: MIT Press. Retrieved from https://www.academia.edu/2257984/The_Relevance_of_Algorithms Halavais, A. (2008). Search Engine Society. Polity. Introna, L. D. (2006). Maintaining the reversibility of foldings: Making the ethics (politics) of information technology visible. Ethics and Information Technology, 9(1), 11–25. http://doi.org/10.1007/s10676-006-9133-z 27 Jenkins, H., Purushotma, R., Weigel, M., Clinton, K., & Robison, A. J. (2009). Confronting the Challenges of Participatory Culture: Media Education for the 21st Century. Cambridge, MA: The MIT Press. Karpf, D. (2008). Measuring Influence in the Political Blogosphere: Who is Winning and How Can We Tell?. George Washigton Univeristy’s Institute for Politics, Democracy and the Internet. Retrieved from http://www.the4dgroup.com/BAI/articles/PoliTechArticle.pdf Kreiss, D. (2014). A Vision of and for the Networked World: John Perry Barlow’s A Declaration of the Independence of Cyberspace at Twenty. In J. ‫אל‬tt & N. Strange (Eds.), Media Independence: Working with Freedom or Working for Free? (pp. 117–138). Routledge. Lazarsfeld, P., & Merton, R. (1954). Friendship as a social process: a substantive and methodological analysis. In M. Berger (Ed.), Freedom and Control in Modern Societies (Vol. 18, pp. 18 – 66). New York: Van Nostrand. Lievrouw, L. A. (2009). New Media, Mediation, and Communication Study. Information, Communication & Society, 12(3), 303–325. http://doi.org/10.1080/13691180802660651 Liu, Y., Gummadi, K. P., Krishnamurthy, B., & Mislove, A. (2011). Analyzing Facebook Privacy Settings: User Expectations vs. Reality. In Proceedings of the 2011 ACM SIGCOMM Conference on Internet Measurement Conference (pp. 61–70). New York, NY, USA: ACM. http://doi.org/10.1145/2068816.2068823 Lukes, S. (2005). Power: A Radical View (2nd ed.). Palgrave Macmillan. McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27, 415–444. Meyer, E. (2014, December 24). Inadvertent Algorithmic Cruelty. Retrieved from http://meyerweb.com/eric/thoughts/2014/12/24/inadvertent-algorithmiccruelty/ Mihailidis, P., & Cohen, J. N. (2013). Exploring curation as a core competency in digital and media literacy education. Journal of Interactive Media in Education, 2013(1), Art–2. Morgan, J. T., Mason, R. M., & Nahon, K. (2011). Lifting the veil: the expression of values in online communities. In Proceedings of the 2011 iConference (pp. 8–15). ACM. Retrieved from http://dl.acm.org/citation.cfm?id=1940763 Nahon, K. (2011). Network Fuzziness of Inclusion-Exclusion [forthcoming]. International Journal of Communication. Nahon, K., & Hemsley, J. (2013). Going Viral (1 edition). s.l.: Polity. Nahon, K., Hemsley, J., Walker, S., & Hussain, M. (2011). Fifteen Minutes of Fame: The Power of Blogs in the Lifecycle of Viral Political Information. Policy & Internet, 3(1), 6–33. http://doi.org/10.2202/1944-2866.1108 Pariser, E. (2012). The Filter Bubble: How the New Personalized Web Is Changing What We Read and How We Think (Reprint edition). New York, N.Y.: Penguin Books. Polsby, N. W. (1963). Community Power and Political Theory. New Haven, CN: Yale University Press. 28 Sagolla, D. (2009). 140 Characters: A Style Guide for the Short Form (1st ed.). Wiley. Segev, E. (2010). Google and the Digital Divide: The Bias of Online Knowledge (1 edition). Oxford: Chandos Publishing. Shalizi, C. R., & Thomas, A. C. (2011). Homophily and contagion are generically confounded in observational social network studies. Sociological Methods & Research, 40(2), 211. Shaw, A., & Hill, B. M. (2014). Laboratories of Oligarchy? How the Iron Law Extends to Peer Production. Journal of Communication, 64(2), 215–238. http://doi.org/10.1111/jcom.12082 Sunstein, C. R. (2009). Republic.com 2.0. Princeton, N.J.; Woodstock: Princeton University Press. Thorson, K., & Wells, C. (2015). How Gatekeeping Still Matters: Understanding Media Effects in an Era of Curated Flows. In P. V. Tim & H. Froncois (Eds.), Gatekeeping in Transition (pp. 25–44). Retrieved from https://wordery.com/gatekeeping-intransition-tim-p-vos-9780415731614 Trottier, D., & Fuchs, C. (2014). Theorising Social Media, Politics and the State. In Social Media, Politics and the State (pp. 3–38). New York: Routledge. Retrieved from http://www.bokus.com/bok/9780415749091/social-media-politics-and-thestate/ Tufekci, Z. (2014). Engineering the public: Big data, surveillance and computational politics. First Monday, 19(7). Retrieved from http://www.firstmonday.dk/ojs/index.php/fm/article/view/4901 Wallsten, K. (2007). Agenda Setting and the Blogosphere: An Analysis of the Relationship between Mainstream Media and Political Blogs. Review of Policy Research, 24(6), 567–587. http://doi.org/10.1111/j.1541-1338.2007.00300.x Wallsten, K. (2011). Beyond Agenda Setting: The Role of Political Blogs as Sources in Newspaper Coverage of Government. In 2011 44th Hawaii International Conference on System Sciences (HICSS) (pp. 1–10). http://doi.org/10.1109/HICSS.2011.80 Watts, D. J. (2004). Six Degrees: The Science of a Connected Age. W. W. Norton & Company. Weber, M. (1946). From Max Weber: Essays in Sociology. (H. H. Gerth & C. W. Mills, Eds. & Trans.). Oxford University Press. Wihbey, J. P. (2014). The Challenges of Democratizing News and Information: Examining Data on Social Media, Viral Patterns and Digital Influence. Shorenstein Center on Media, Politics and Public Policy Discussion Paper Series. Retrieved from http://dash.harvard.edu/handle/1/12872220 Winner, L. (1996). Who will we be in cyberspace? The Information Society, 12(1), 63–72. Woodly, D. (2008). New competencies in democratic communication? Blogs, agenda setting and political participation. Public Choice, 134(1), 109–123.