Public sphere/Draft
From Driscollwiki
Goals
This paper will:
- discuss "capacity to support the emergence of networked publics" as a characteristic that varies among different social network sites
- consider the potential role of networked publics in users' choice of SNS, specifically attending to the theme of "migration" from MySpace to Facebook
- compare the specific technological features of MySpace and Facebook that might support (or foreclose) the emergence of networked publics and counter publics
- Identify three characteristics of Facebook's software that enable the emergence of networked publics and counter-publics: feeds, threads, and ambiguous privacy settings.
- discuss the ethical costs associated with these features
Abstract
Introduction
Coverage of the uprisings in Egypt, Tunisian, Bahrain, and Yemen, often referred to collectively as "Arab Spring", were often accompanied with speculation regarding the role of "social media" to organize demonstrations and circumvent state censorship. In spite of its application to these very specific events, the term "social media" itself remains remarkable unspecific. In any moment, it may refer to practices as diverse as sharing a home video on YouTube, sending text messages from a mobile phone, breaking a headline on Twitter, or creating a call to action on Facebook. Anticipating the events of Arab Spring, Hashemi-Najafabadi's analysis of internet adoption among Muslim societies suggests that social network sites (SNS) like Facebook provide a platform for the formation of revolutionary counter-publics. Research on the use of SNS by young people in the U.S. affirms this possibility by establishing a relationship between the common features of social network sites and the emergence of networked publics. Less common in the existing literature, however, is attention to the specific technological features that distinguish one social network site from another. To what extent do different implementations of the same SNS features affect a platform's capacity to support (or foreclose) the emergence of networked publics and counterpublics?
Like the journalists' use of "social media", scholars curious about the possibilities of online publics have had a bad habit of thinking about the internet as a coherent single communication technology that might be meaningfully compared to earlier forms such as broadcast television or amateur radio. Communication scholars would rightfully balk at a category of analysis so broad that it contained desktop publishing, drive-in movies, and FedEx Overnight Delivery but Dahlberg's review of the scholarship on internet communication and deliberative democracy employs a category of "two-way, semi-decentralized communication technologies" inclusive of "Web publishing, electronic bulletin boards, e-mail lists, and chat rooms" (59). Hashemi-Najafabadi speaks directly to this bad habit by encouraging readers to think of the internet as a series of technologies and practices that continue to evolve over time. As conventionally "offline" services such as banking, telephony, and television are increasingly carried out using internet protocols, more narrowly defined categories (e.g. synchronous/asynchronous, anonymous/authenticated, location-aware/location-agnostic) may enable more meaningful comparison among available sites, softwares, and services.
In its singular form, "the internet" is a useful abstraction for an unbounded network of networks joined together by a set of shared protocols. Common protocols does not extend upwards into user-facing interfaces, however, and the practices and platforms that make up users' everyday experience of the internet are so not easily compared. Network software and services differ according to genre (email, chatroom, messageboard, search engine), platform (mobile, web, desktop) and implementation (YouTube, Dailymotion, Vimeo, blip.tv), none of which is necessarily exclusive to any of the others.
Frustrated that online political discourse seems to lead to polarization and away from critical-rationality, some scholars dream of designing ideal communicative spaces on the web. Though these efforts are instructive on their own, their widespread adoption is limited by a lack of inquiry regarding the publics that already exist. "Social network sites", in particular, represent a highly visible shift in the everyday practices of internet users. Services like Facebook replicate many pre-existing genres of online communication (email, chat, public messaging, and media sharing) within a centralized, authenticated socio-technical architecture. Whereas previous research has tended to focus very narrowly on the content of specific interactions among users or very broadly on the political economy of commercial service providers, the current project addresses MySpace and Facebook - two widely used SNS in the U.S. - as software. Although each site offers a unique implementation of the core SNS features, they also vary considerably in their circulation of discursive material, their arrangement of public message, and the "privacy" settings they make available to users.
Social network sites
The leading definition of web-based social network sites (SNS), crafted by boyd & Ellison, identifies three basic activities available to users: to construct a public or semi-public profile within the platform, to articulate a list of other users with whom they share a connection, and to view and traverse this network of connections (2007). These general criteria appear in a variety of sites and services including interest-based sites like Yelp or ESPN, identity-based sites like BlackPlanet or MiGente, media-sharing sites like Flickr or YouTube, and general-purpose sites like MySpace and Facebook. One reason for the widespread deployment of these core SNS features is that the social meaning of SNS activity is highly flexible and may be adapted to a wide variety of contexts. In many cases, including ESPN and QQ, social network features were added to sites that were initially developed for other purposes.
Not all social network sites enable encounters with strangers or stranger discourse. Most, including MySpace and Facebook, require "bi-directional confirmation" of friendship (boyd & Ellison, 2007) and research suggests that SNS are used primarily to connect with existing "offline" social networks (Haythornthwaite, 2005). Variation in access to user data is a key dimension on which SNS sites distinguish themselves. Access to profiles is managed differently in various sites and is not always limited to friends. Public MySpace profiles, for example, are accessible to non-users and included in search engine results. As with all features, these characteristics are subject to change at any time.
At the outset of their definition, boyd & Ellison urged readers to adopt the term "social network sites" and discouraged the usage of "social networking sites". Beyond clarifying ambiguous terminology, this distinction calls attention to the gap between the social practice of networking and the articulation of one's social network on a website. It is easy to confuse the features of a communication technology from the social meanings that users attach to them. For example, the ability to establish a connection with another user ("Friending") is a basic SNS feature implemented by software developers and exposed to users through a web interface but the social meaning of this feature varies according to context. A single user might deploy the same Friending feature for entirely different purposes over time: "networking with my boss", "flirting with the new letter carrier", "re-connecting with an old friend", or "showing forgiveness to my sister".
Unlike desktop or mobile software that executes on a user's local machine, social network sites are web-based services that run on a server and are accessed by users across the internet through client software such as a web browser. This arrangement contributes to a productive platform instability as developers regularly update features in response to user demand, usage patterns, or internal corporate strategy. It also presents thorny methodological challenges to the curious researcher. Web-based services do not have distinct "versions" or milestones. Constantly updated, the experience of using the service is utterly ephemeral, cannot be recreated, and is difficult to accurately archive. Since they cannot simply install an old version of MySpace to understand what users were seeing in 2006, researchers today must rely on scholarly and popular accounts of using the service. The second methodological challenge concerns the interrelationship of developers and users. As users adapt services to their social needs, developers may respond by altering the software architecture to match emergence user practices. In cases like the use of the @ and # symbols on Twitter, user innovations contribute materially and immediately to the development of the platform. Researchers may find it difficult at times to ascertain whether a given implementation was the result of developer initiative or user demand.
Profiles
boyd & Ellison's definition of social network sites places the construction of a user's profile at the center of his or her experience of the site. Each user can have only one profile on a given service and it serves both as a kind of home page and as an avatar representing him or her within the site (boyd, 2007, 6). Upon creation of a new account on any SNS, the user is typically prompted to begin creating a profile by filling out a short questionnaire. The prompts in the questionnaire vary according to the designers' intended uses of site. A profile on Yelp might ask for "Favorite Food" while the general-purpose MySpace includes more open-ended prompts such as "About Me".[1] Unlike a traditional home page, the degree to which a user can personalize his or her profile beyond the basic questionnaire depends on the features of the specific site. While many invite users to upload one or more photos, others enable users to customize the layout, alter the color scheme, add background images, or embed rich media applications.
boyd describes the creation of a profile as an "initiation rite" after which the user is officially a member of the site (2007, 10). From that point forward, the profile is a locus of individual expression wherein the user combines text, images, and sounds into a multi-media collage to be read and interpreted by others. Drawing on Jenny Sundén, boyd compares the MySpace profile to "a form of digital body where individuals must write themselves into being" (2007, 13). Writing proliferates in and around the profile in a quite literal sense as well. Once established, the profile not only serves as an expressive artifact but also as the means through which a user may speak or be addressed.
= Public messaging
In addition to the core features of SNS discussed above, most sites implement public and private messaging features. These messaging systems exist within a broader ecology of messaging that includes email, SMS, instant messaging, and web forums. There is overlap among the features of these various systems that may undermine their usefulness and add noise and confusion to users' experience. Some sites attempt to mitigate these negative effects by implement gateways among different platforms. For example, users may opt to receive email or SMS notifications regarding activity taking place within a social network site. In most cases, SNS messaging systems function more or less that same as email, IM, or web-based messageboards with the added context of one's articulated social network.
Of particular interest to the present research are implementations of public messaging that take advantage of their context. On some sites, public messages may be inscribed directly onto a user's profile page. Nearly identical implementations of this feature across sites belie the diversity of uses and meanings that emerge among various user populations. On LinkedIn, public messages are titled "Recommendations"; on Friendster, they are "Testimonials"; on MySpace, "Comments". Each case implies a certain tone and style but users are free to negotiate the extent to which they will comply.
Some architectures seem to encourage dialogic forms of public messaging while others are more strictly one-sided (boyd, 2007, 6). For example, when a Facebook user reads a message posted to her "Wall", she is offered a link to respond on the writer's Wall. Through this simple "Wall-to-Wall" mechanism, Facebook users to bring private exchanges into the semi-public space of their articulated social network (boyd, 2007, 7). Although Facebook implements features that make public dialogue easier, dialogic interactions are possible in any site with public messaging.
Thelwall & Wilkinson conducted one of the few large scale studies attending to the practice of dialogic public messaging in social network sites (2010). They found that 78% of active MySpace users in the US and UK received at least one public comment and the typical user received public comments from a third of their Friends (398). In spite of the wide use of comments, most appeared to be one-way as fewer than one-third of users engaged in dialogic activity (395). Although the study was designed to assess the use of public comments for the circulation of gossip, they found that public dialogs typically unfolded sporadically over the course of days and may have been used primarily like a "greeting card" to "stay in touch" (394-5).
Thewall & Wilkinson's study is valuable because it calls attention to the distinction between the features of MySpace and the actual meanings that users attach to the use of those features. The authors report that two-thirds of the profiles in their initial sample were set to "Friends-only" and could not be included in the study. They note that although users might be more likely to share gossip on a profile that is set to "Friends-only", there is not feature implemented in MySpace to indicate whether a given profile is set to public or private (395). Similarly, the publicity of a comment on a "Friends-only" profile is contingent on the growth of that user's network. What seems "private" on a profile with 100 Friends today may feel quite "public" tomorrow should that user's network balloon to 10,000 Friends.
It is possible that some of the comments collected in Thewall & Wilkinson's study were actually part of conversations involving more than two participants. The architecture of public comments on a MySpace profile offers no mechanism for determining or demarcating participants. Instead, comments are arranged in a flat, chronological list of messages. In the absence of a "Wall-to-Wall" feature, Thewall & Wilkinson presumably had to analyze the profiles of each commenting user for evidence of a back-and-forth exchange. If public conversations with three or more participants were taking place on any of the profiles in their sample, it would have required close reading of an exponentially larger number of comments to identify them.
The implementation of features on a social network site mark a point of contact among researchers, developers, and users. The challenge that MySpace's public messaging system poses for researchers mirrors the experience of its users. Just as the researcher would struggle to identify all of the participants in a public conversation, so would participants find it difficult to keep track of a unfolding conversation. This is not to say that group conversation is impossible in MySpace but that it requires extra work on the part of users to manage and track.
Networked publics and network imaginary
Deliberative democrats tend to analyze communication platforms in terms of their ability to encourage or limit argumentation. The commercialization (and popularization) of the internet in the early 1990s seemed to offer a promising new platform for argumentation. In comparison to contemporary mass media, internet-based communication was "two-way, relatively low cost, semi-decentralized and trans-national" (Dahlberg, 50). The limitations of public messaging on MySpace indicate that this may have been an over-broad characterization of internet-based communication. Evaluating individual sites, softwares, and services requires a closer look at the specific implementation of features proven to support (or foreclose) public discourse.
A traditional view of the public sphere, in the Habermassian sense, invokes a space of rational deliberation over disputed validity claims aimed at reaching understanding and agreement (Dahlberg, 49). The publics found on social network sites, as defined by boyd & Ellison, tend not to be oriented toward reaching consensus. Instead, these publics emerge as a result of the circulation of discourse among overlapping social networks constituted by individuals who may or may not be previously acquainted. Public rationality and opinion are formed through the practices of deliberation: making claims, providing evidence, listening, and changing one's mind. In an ideal case, participants find themselves transformed from "privately-oriented selves into publicly-oriented 'citizens'" through the inter-subjective activity of engaging in public discourse with strangers (Dahlberg, 50).
"Networked public" is a term used by boyd & Marwick to describe the "space constructed through networked technologies" (7). Their understanding of the term "public" draws less from the goal-oriented tradition of Habermas or Rawls than it does from the understanding of audiences and peers in Anderson and Arendt. They describe the emergence of "imagined communities" within sites like Facebook and Twitter "as a result of the intersection of people, technology, and practice" (7). According to this understanding, a social network site is both a platform within which a networked public might emerge and a networked public itself. Evidence of this latter understanding is found in colloquial sayings like "Twitter was going crazy" or "everyone on Facebook was talking about it".
Four properties of social network sites distinguish the circulation of discourse in networked publics from face-to-face interactions in public places (boyd, 2007, 3). First, the conventional communicative features of social network sites are asynchronous and require persistence. Unlike the ephemeral speech acts of face-to-face interaction, a public comment on a MySpace profile will lingers until either the speaker or the addressed chooses to delete it. In addition to its persistence, this public comment is replicable. Another user may create a local copy of the comment be either copy-pasting or making a screenshot. Even if the original comment is deleted, there is a possibility that copies will continue to circulate within or without the original context to which it was posted. The third unique characteristic of networked publics is their searchability. If the social network site itself does not implement a search feature, user data may still be located using an outside search engine like Google. Finally, because of these three characteristics, it is impossible for a user to determine exactly who will encounter the trace artifacts of any given act. Even seemingly "private" contexts such as one-on-one messaging may have invisible audiences in the future.
boyd & Marwick's use of "networked public" calls to mind Warner's discussion of "publics" and "counter-publics" (2002). Similarly departing from the consensus-based models of earlier theorists, Warner describes publics as self-organized spaces of discourse that emerge amongstrangers paying attention to a text or set of related texts. The combination of persistence and imagined audiences in SNS enables a mode of public speech that is at once "personal and impersonal". Speakers address an audience that is immediately known but the trace artifacts of their speech potentially circulate among audiences that could not have been known from the start. These lingering artifacts provide the textual material around which publics (and counter-publics) can gather. In the same sense that users "write themselves into being" through the construction of individually-expressive profiles, networked publics "write themselves into being" through their use of public messaging systems with unstable boundaries.
Although the lack of certainty in public messaging systems may be productive of publics, it also confounds conventional strategies of address. It is easy to imagine a situation in which lack of certainty has a chilling effect on public speech and public messaging system go largely unused. In a series of interviews with Twitter users, boyd & Marwick find that the imagined audiences among Twitter users are neither the conventional "writer's audience" - a mental image of who is being written to - nor the "broadcast audience" - a group of people gathered around, engaging with, and interpreting media artifacts. This "networked audience" combines elements of both into a group that "contains familiar faces" but cannot be fully enumerated or identified (16).
In some respects, the networked audiences of SNS, as described by boyd & Marwick, are more like the counterpublics described by Warner or the participatory cultures theorized by Jenkins (2005) than the word "audience" might suggest. A networked audience is defined by its practices. It has a clear means to respond and speak back; its relationship to a text is always potentially dialogic. Audience members are explicitly linked within a network that enables them to also speak directly to one another. In particularly active networked audience, the role of the original speaker quickly recedes as "audience members take turns creating and producing content" (16). Whereas the capacity for a counterpublic to "act" is often immaterial, discourse is persistently inscribed in mediated spaces, leaving behind a trail of trace artifacts. The aggregation of these traces by semi-autonomous software agents can result in quite material effects such as the reorganization of Google search results.
Contrasting the capacity of MySpace and FB to support networked publics
A lack of specificity in discussions of "social media" continues to allow unproductive comparisons of social network sites based on their common features rather than their distinguishing characteristics. Apocryphal histories of the last decade describing a neat succession of social network sites reflects this tendency. Take, for example, the following passage from a recent feature article in Wired magazine:
"The early leader in social networking, Friendster, had terrible technology. The advertising-heavy MySpace, which dethroned Friendster, was a bit like Yahoo Answers: chaotic and low-rent, prone to spammers and scams. Yet both were immensely popular - until Facebook came along and figured out how to do social networking right." (Rivlin, 2011, 178)
What is "social networking"? And what does it mean to "do it right"? Rivlin's brief narrative history indicates that the technology must not be "terrible", "chaotic", "low rent", or "prone to spammers and scams" but these colorful descriptions tell us little about the specific practices that make up "social networking" or the features that are required to carry them out.
According to the teenagers interviewed by boyd, spatial and social isolation due to suburban sprawl and over-scheduling drove many teens to adopt sites like MySpace for the purpose of "socializing" outside of high school (2007). Instead of hanging out in shared geographic spaces like the mall, the town park, or a Dunkin Donuts parking lot, these teens connected with one another in the mediated spaces of social network sites. In offline spaces, socializing is easily observed: people gather and talk as they eat food, window shop, dance, play sports and video games, and watch TV. How is something analogous realized with the specific features of MySpace and Facebook?
Socializing is a set of activities that can bring together a public. The shared experience of a media artifact such as a TV show opens a space for discourse. Should commentary turn into discussion and debate, a public is formed out of the everyday practices of "hanging out". If teenagers were able to effectively socialize on MySpace and Facebook, then these sites should also support the emergence of a networked public.
A very brief history of MySpace and Facebook[2]
MySpace was launched in 2003 as a competitor to early social network sites like Friendster, Xanga, and Asian Avenue. Among its early adopters were a significant number musicians and event promoters from the LA area. MySpace representatives reached out to those users and prioritized the development of features to meet the needs of indie musicians. As musicians came on board, so did their audiences, leading boyd to characterize their early growth as "bands-and-fans".
The responsiveness of MySpace developers to the demands of its users distinguished the site from Friendster, which had recently frustrated many early adopters by disallowing fun and productive "fake" accounts (Marwick, 2005). In one case, this responsiveness came in the form of benign neglect rather than proactive development. Users discovered that they could alter the appearance of their MySpace profiles far beyond what was intended by the site's developers by pasting snippets of code into the fields on the profile questionnaire. Conventionally, web developers disallow this sort of manipulation of the platform because it can introduce significant security issues. MySpace, however, tacitly allowed the practice to continue and a rich culture of code swapping and peer learning soon emerged (Perkel, in press). Profile customization, the result of user innovation, later became one of the key features distinguishing MySpace from Facebook.
The visibility of MySpace increased significantly after it was purchased by News Corp. in 2005. While this increased visibility attracted many more users to the site, it changed the social context within which users had grown accustomed. For early adopters, profiles marked "public" likely began to feel increasingly exposed as large numbers of new users and new eyeballs were drawn to the site. A moral panic regarding sexual predators on the site emerged later that year, although research indicates that the risks were exaggerated (Marwick, 2008; Schrock & boyd, 2009). The same benign neglect that earlier enabled a productive culture of profile customization was soon exploited by less well intentioned users and spam began to proliferate across the network.
Facebook, meanwhile, launched in 2004 as a social network available exclusively to Harvard University students. Over the next three years, the site slowly opened up; first to other Ivy League institutions, then gradually to all college students, next to selected high schools and corporations, and finally to all internet users in September 2006. In contrast to the colorful profiles, lengthy blog posts, and focus on music found on MySpace, Facebook's initial feature set was quite limited. Users were only allowed a handful of profile photos and could not customize any other aspect of their profile pages. Where Facebook lacked expressive capacity, however, it excelled in stability and an "aura" of exclusivity (boyd, forthcoming, 23).
Comparison of their implementation of the core SNS features
The combination of malware and moral panic gave some the impression that MySpace was an unsafe space. boyd's interviews with teenagers of the time reveals that some parents began to monitor their children's use of the site while others forbid them children joining altogether (boyd, forthcoming). Facebook, with its Ivy League roots and "clean" appearance offered a safer-seeming alternative social network sites. Although MySpace continued to grow, a migration narrative (like the one quoted above) emerged suggesting that users "left" MySpace to join Facebook. Although not entirely contradicting this narrative, boyd's research suggests that a large number of teenagers found themselves simultaneously maintaining a presence in both sites.
In 2006, MySpace and Facebook both met the minimum definition of a social network site as set forth by boyd & Ellison. On each site, new users filled out a short questionnaire to establish a profile and used similar "Friending" features to establish an articulated social network within the site. Today, Facebook is a significantly larger, more widely used site than MySpace. To understand why users with accounts in both sites ultimately chose to focus their attention on Facebook requires closer analysis of the distinguishing features of each site. Specifically, what features drew users back to Facebook more frequently than MySpace in spite of its sparser layouts and limited capacity for building expressive profile pages?
Three key features distinguish Facebook from MySpace and continue to characterize user interaction in the site. First, in the same year that Facebook opened to its doors to the entire internet, it shifted from a profile- to a feed-based paradigm. Second, public messaging in Facebook is arranged into threads that circulate throughout the network. And, third, the "privacy" settings in Facebook are complex, unstable, and widely misunderstood by users - toward unexpectedly productive ends.
News Feed
As discussed earlier, the implementation of profile pages on MySpace provided a highly customizable means for expressing group affiliation, personal values, and aesthetic preferences. Accessible to the entire internet through a custom URL (e.g. http://myspace.com/my_home_page) or a Google search, MySpace profiles were largely indistinguishable from the personal home pages described by Papacharissi (2002). Users might spend hours assembling the perfect blend of images, songs, and videos on their profiles but for all this expressive possibility, the MySpace profile is a starkly asymmetrical medium for interpersonal communication; a high-bandwidth channel extends outward from the profile with very little opportunity for reciprocity. In contrast, a Facebook profile is a much more symmetrical medium; minimally expressive in form but providing numerous opportunities to spark dialogic interaction.
In the four years since boyd & Ellison crafted their milestone definition of social network sites, the role of profiles in Facebook has become increasingly marginal. Rather than log in periodically to maintain an expressive homepage-like profile, users monitor - and in some cases comment on - a stream of artifacts generated by the activities of others throughout the day. Facebook profiles, in this arrangement, are reduced to mere portals for the latest public messages, photos, and other shared media associated with a user. Whereas a MySpace profile might be updated in successive "versions", the Facebook profile reflects the underlying architecture of Facebook as a web service - it is dynamically generated according to the current state of the network and liable to change at any time. In this stream-based paradigm, it is no longer necessary to traverse one's social network to stay abreast of other users' activities on the site. All of this information is available via the News Feed.
When the News Feed was implemented in 2006, it proved stressful for many users (boyd, 2008). Although it did not change the public/private status of any material on Facebook, it changed the extent to which data circulates through the network. A user may have been comfortable with a piece of information sitting deep in their profile page, assuming that only a very few people would read closely enough to encounter it, but the stakes are raised when every change to a profile generates a widely-circulated news item. The anxiety produced by the introduction of the News Feed reveals the interdependence of publicity, circulation, and imagined audience with the specific implementation of features as software.
Feeds are not unique to social network sites. In fact, they were implemented first in media-sharing sites designed to maximize the circulation of user data.[3] The basic activities of social network sites do not seem immediately feed-like. The innovative dimension of Facebook's News Feed was to generate trace artifacts from moment-to-moment user activity. Actions that were previously noticed by only a small subset of one's network such as adding a Friend or updating one's profile photo, were elevated into newsworthy events. By adding these activities to the News Feed, the Facebook platform effectively transforms non-verbal social interactions into discursive events. For example, when two users connect, a declarative statement appears in the News Feed notifying everyone in both of their networks that says "Suzy and José are now friends."
The shift from a profile- to stream-based paradigm, draws attention away from profiles and users as the primary loci of activity. Instead, users attend to a stream of trace artifacts generated by the social activities of their Friends in the network. On one hand, this stream offers a constant reminder of the presence of other users in the network. Even if two Friends never interact directly, they may be aware of one another through the shared experience of the News Feed. On the other hand, the software underlying the News Feed transforms all activity into discourse. Combined with thread-like public messaging system, each item in the News Feed is a provocation, a potential starting point for dialogic activity. Responses to these provocations are themselves transformed into tiny discursive shards and circulated throughout the network, leading to a powerfully self-reflexive discursive space of the kind theorized by Warner in 2002.
Threads
In one of the interviews quoted by boyd in her article on the perceived migration of young people from MySpace to Facebook, a teenager recalls gathering with the "honors kids" to "discuss" a homework assignment on Facebook (boyd, forthcoming, 27). Neither MySpace nor Facebook featured real-time multi-user chat functionality at the time of the interview. How did the available public messaging features of Facebook allow these kids to "discuss" a homework assignment? Could a similar interaction have been possible in MySpace? What did the "discussion" look like on Facebook?
The "thread" is a recurring technology for carrying out public discourse in online environments. It is an abstract form that has been implemented in countless platforms across the history of networked personal computing. The thread is a techne that situates the communicative acts of unbounded groups of users in a history, temporality, location, and future. Communication systems lacking a specific architecture for managing threads are either shortly abandoned or driven to change through the creativity of their users.[4]
Threads are neither especially new nor unique to computer-mediated communication systems. Thread-like interactions are evident in print media as well as many early computer-mediated communication systems. The form of the thread is as evident in letters to the editor, notes passed around in the backs of classrooms, and serial pamphleteering as it is in email, listservs, bulletin boards, newsgroups, forums, guestbooks, blogs, and social-networking sites.
Today, threads are so thoroughly domesticated that their implementation and use persists with scarcely as much consideration as the QWERTY keyboard or the "http" at the start of a URL. With a few notable exceptions, the thread is rarely a locus of innovation in new communication systems. For example, uploading a video to YouTube is elegant and easy to learn but participating in its commenting system is confusing and clunky. When threads are not implemented haphazardly or as an afterthought, they may be outsourced to a third-party such as Disqus, a decision that trades heterogeneity, innovation, and end user privacy for stability. In spite of this ubiquity, threads proliferate within Facebook and are all but absent in MySpace.
Core features of all threads
All threads begin with a speech act. Unlike a chat room in which space precedes speech, the space of a thread is created in the moment that the "first post" is made. The content of a "first post" varies according to platform but its role in the unfolding thread is consistent. The "first post" is always both an invitation and provocation for readers to respond. As the number of replies grow, the first post also acts as an anchor for the thread. Even in cases where subsequent posts go "off-topic", traces of the first post frequently linger in the form of quoted text, a URL slug, or a thread title.
Posts that take up the initial provocation have many names: "replies", "responses", "posts", "comments", etc. Each represents a discrete speech act that contributes some discursive material to the unfolding thread. Like the first post, these subsequent messages in the thread also have the capacity to provoke a response from lurking readers. In the abstract, there is no upper bound to the number of replies that might accumulate in a single thread. Although the first post marks a definite beginning, threads lack any definite end.[5]
Threads are also open in terms of participation. Even if a specific audience is named in the first post, the future participants cannot be known from the outset. New participants may join at at time, and existing participants are not obligated to return. As a result of this ambiguity, the audience imagined by later participants may be significantly different from that of the first poster.
Threads are emblematic of the asynchronous communication that typically takes place on social network sites. Because of their lack of closure, it is difficult to simply capture their temporal character. Given a sufficiently large, active audience, a thread may accumulate dozens of replies within a few minutes of its first post. And due to its open-endedness, the thread might also fall dormant shortly thereafter ... only to be revived in two years and subject to another flurry of activity.
The experience of reading a thread is also one of significant temporal elasticity. On one hand, it is possible for participants to watch a thread unfold in near real-time by continuously reloading their web browsers. On the other hand, new participants may arrive to a thread months or years after the creation of its first post. For these late arrivals, the thread's temporal history may be difficult to ascertain. In some cases, it may be erased altogether, leaving the reader with the false impression that all posts occur in contemporaneity.
Variation in implementation and use
Not all implementations of the thread enable a networked public to emerge according to boyd, Marwick, and Warner's criteria. The extent to which structure, participation, and temporality are open-ended depends on the thread's specific technological circumstances. These fundamental characteristics are modified, extended, and, in some cases, limited, by the techno-material constraints that arise during the process of implementation. For scholars analyzing a specific instance of online discourse, separating the generic characteristics of a thread from the features of its unique implementation is helpful for uncovering details about its underlying platform.
Although threads appear throughout the history of asynchronous computer-mediated communication, today's most visible examples are on the web. These systems fall under three broad categories and threads emerge differently in each. On those sites and services organized around the display of digital artifacts (e.g. Flickr for photos, YouTube for videos, or New York Times Online for news articles), threads unfold in response to digital objects that may or may not be scheduled and arranged by a centralized authority. In systems that place the shared interest of users at their center (e.g. a messageboard for people suffering from carpal tunnel syndrome), first posts are more likely to be generated sporadically by users in a variety of media. Finally, on services designed to encourage interaction among users like Facebook, threads unfold in the wake of everyday user activities. The degree to which any of these threads constitutes a public is affected first by the structural elements of the platform including the arrangement of URLs, visual presentation of replies, and opportunity for users to revise or redact their contributions.
Structural
Permalinks
"Permalinks" are unique URLs that point to specific resources within a dynamic system on the web. As the name implies, the purpose of a permalink is to provide participants with a reliable pathway back to the thread far into the future. In general, sites organized around artifacts (YouTube, NYT) assign permalinks only to the individual artifacts. To access the related thread, users have to start at the beginning and scroll through the entire history. Discussion forums like Reddit, however, assign a unique permalink to every post in a thread so that users can jump directly to one or another reply. The collected permalinks across a system provide a map determining the total possible points of entry into the discussions unfolding within.
Permalinks are critical technologies for the circulation and proliferation of threads across the web. Participants who wish to publicize their contributions outside the bounds of the immediate system can copy-paste these URLs into IMs, emails, tweets, and Facebook updates. As traces spread throughout the participants' networks, new audiences are implicitly invited to find their way into the thread and make a contribution. In such cases, the first post in a thread will lose some of its capacity to frame the discourse of the thread as users enter later in the temporal sequence. Furthermore, posts and updates meant to publicize the thread may in turn spark new threads on these complementary systems.
Visual arragement
Various implementations of the thread also structure replies in different logical, temporal, and visual arrangements. In the simplest case, the first post will be followed by an arbitrary number of response posts in no particular order. Much more common, of course, is a reply structure that maintains either a flat chronological sequence or a topical tree with chronological brances. Historically, highly decentralized messaging systems such as FidoNet or USENET have not maintained strict chronologies because messages from closer nodes arrive before more distant nodes regardless of the time that they are composed. In most web-based systems, however, messages are stored in a single database and presented to the reader in a strict chronological sequence.
As threads grow beyond one or two replies, it is often difficult to determine the audience that each new reply is intended to address. In a flat chronological sequence, readers must reply on linguistic heuristics such as quotations or direct address ("I disagree, UserX, ..."). In the absence of such clues, replies are likely presumed to address the entire public of the thread. Tree-like structures offer additional visual information by identifying branches with extra-textual elements such as indenting, arrows, or shifts in color.
TODO insert screencap of Gawker and Reddit indenting
As topics diverge and narrow in very active discussions, the branches of a thread tree may reveal emergent "sub-threads". It is possible for a sub-threads to achieve sufficient independence from its trunk that it constitutes a public all its own. As with most features, discussion trees are not uniformly implemented across systems. While the comprehensive permalinking system within a site like Reddit encourages lengthly sub-threads to emerge and circulate, Gawker constrains the growth of branches (and their publics) by limiting the "depth" of replies to one or two "levels".
TODO insert screencap of Slashdot
Threads with very large numbers of participants present a significant challenge to readers and interface designers. As the number of replies rises into the hundreds or thousands, it may be unreasonable to assume that a participant has read the entire thread that precedes them. In these cases, responses may be filtered from view according to a series of heuristic filters. Slashdot filters high-volume threads with the help of volunteer moderators, Digg relies on a simple thumbs-up or thumbs-down voting system, and Facebook displays only the most recent contributions. Critically, none of these systems permanently obscures any contributions and the curious participant is always given the option to remove filters to explore the thread in full.[6]
Revisions and redactions
Individual contributions to a thread are not always etched in stone. Messageboards based on the phpBB software platform, for example, may enable users to revise posts at any time after they are made. This can create historical discontinuities that befuddle newcomers to a thread. It is not hard to imagine that the expression of a controversial political opinion will inspire a series of angry responses in a forum that does not usually discuss political issues. Should the original poster redact his or her comment, the subsequent responses may fall into incoherence for future readers. This simple example underlines the interdependence of all posts within a thread.
Thread continuity may also be ruptured in systems with strong-handed moderation. Wikipedia administrators occasionally "lock" discussion pages if they believe that the interaction is in violation of community principles. This exercise of authority is possible only at points of interchange within a communication network. Moderation is quite final in highly-centralized systems such as Wikipedia but in distributed networks like USENET, moderation occurs at end nodes and may not be experienced equally by all participants in a thread.
Whereas synchronous systems such as an IRC channel or an AOL chatroom display a list of active users, there is no definite number of active participants at any given point of time in the life of a thread. This ambiguity is mitigated in some implementations by restricting access to certain groups of potential participants. Goal-oriented systems often limit user activity according to reputation. On the question-and-answer sites in the StackExchange network, for example, a novice user cannot start a new threads until after her contributions to existing threads have been positively reviewed by more senior users on the site.
There are variety of approaches to managing the identity of users in public messaging sites but they tend to vary along two axes: persistence and authentication. At one extreme, Neighborgoods strongly encourages the use of persistent, authenticated identity based on location and a small donation. At the other extreme, 4chan discourages the use of persistent identities and hundreds of different users will post to the same thread using under the name "Anonymous." Facebook accounts are for individual persons only and the site requires users to list their "real" first and last names.[7] MySpace is much less strict and users often create profiles using pseudonyms.
Threads within MySpace and Facebook
Neither MySpace nor Facebook implements a particularly dense map of permalinks. Aside from the custom URLs pointing to user profile pages, MySpace generates permalinks only for individual blog posts. Links to material within Facebook are unstable due to changes in privacy and publicity settings. What was accessible one month ago may not longer be accessible today. The lack of permalinks in these sites reflects a conscious effort to keep user attention facing inward. Rather than allow links to circulate freely on the web, they wish to direct users into the network through a limited number of gateways.
Although the implementation of threads in Facebook is not particularly sophisticated, it is omnipresent within the site. Every photo, Wall comment, status update, and News Feed item acts as a potential first post within the site. Each time a response is posted, the thread jumps up to the top of participants' News Feeds. As a result, threads both proliferate and circulate widely throughout Facebook.
MySpace implements comment threads in only two places on a user's profile: the blog and the public comments. In both cases, there are no technical means for tracking the history of a given topic nor for keeping track of an on-going thread. As reflected in the results of Thelwall & Wilkinson's study, public messaging in MySpace tends to be limited to sporadic one-sided commenting rather than on-going discussion.
Ambiguous privacy settings
One of the fundamental differences among social network sites is the extent to which user data is exposed to fellow users and the internet at large. In some cases, the implementation of privacy features is the key competitive advantage of choosing one SNS over another. In contrast to claims made by Facebook founder Mark Zuckerberg in December of 2009, a survey of 18 and 19 year old Facebook users found that many feel strongly about privacy. In a series of interviews concerning attitudes toward privacy, Marwick & boyd found that many managed their publicity through a complex system of social and technical means (2011).
MySpace and Facebook both require bi-directional confirmation to establish a social network link but similarities between their implementation of privacy features ends there. While MySpace appears to offer less fine-tuned tools for regulating access to user data, its simple system largely works as advertised. Facebook users, by contrast, have endured an unstable array of privacy settings leading to unpredictable behaviors including accidental publicity. Privacy settings impact the capacity of these sites to support the emergence of publics because they regulate both the circulation of discourse and the size and character of imagined audiences.
On MySpace
Privacy on MySpace is effectively a binary switch. Profiles are either private ("friends-only") or fully public and indexed by Google. Setting one's profile to private reduces likelihood of unwanted viewers but limits the possibility for serendipitous encounters with semi-strangers and weak tie acquaintances. Keeping a public profile and promiscuously Friending other users opens up access to a larger social network with in the site. As one's network grows, the landscape of MySpace quite literally enlarges as new pages and blog posts become accessible.
Teens interviewed by Marwick and boyd believed MySpace to be implicitly teens-only and that adults should not be accessing the site (16). Many parents did not share this expectation and accidentally transgressed the social norm established by their children. The blunt privacy instruments implemented by MySpace did not match the more nuanced expectations of teen users resulting in feelings of violation and unintentional exposure. In some cases, teenagers report that adult supervision had a chilling effect on their willingness to engage in public discourse on MySpace.
On Facebook
Intitially, privacy on Facebook was implemented according to Network, a loosely-defined grouping of users that has since been abandoned (boyd & Hargittai, 2010). Networks could be regional or institutional, organized around an employer, city, or university. During its earlier, more exclusive phase of development, Facebook verified users' network affiliation by the domain name in their email address. In other words, only users with @usc.edu addresses could join the University of Southern California Network and share data with other USC students.
As users without academic affiliations joined the site, Facebook developers implemented a matrix of privacy settings arranged around categories of data: jobs, photos, relationship status, etc. For each category, users establish a boundary indicating which subset of their social network should be able to see the material: No one, Friends, Friends-of-friends, Network, Apps, Everyone. The meaning of these groupings is not self-evidence, however. Does "everyone" mean "everyone on Facebook" or "everyone in the internet"? Can I choose only the "friends-of" some friends?
Over time, the privacy features became unwieldy and lead to increasingly confusing settings. On top of the growing matrix of privacy options, users were given the opportunity to create their own categories which would be integrated into the matrix. Facebook revised the settings several times, including an effort to finally simplify them in 2010 (Zuckerberg, 2010). In a survey of college students, boyd & Hargittai found that active Facebook users tended to adjust their privacy settings quite often and that nearly every user in their survey had altered the settings at least once in the previous year (2010). Facebook's internal data indicates that as many as 65% of users leave the default settings in place, making their profile, network, and posts publicly accessible (Perez, 2007).
Each time Facebook introduced new features, the options were set to the most permissive, most broadly shared setting by default. In spite of evidence that active Facebook users often change their privacy settings, teenaged users interviewed by boyd & Marwick were "not convinced that Facebook's privacy settings [...] actually help them control how information flows" (2011, 19). It is difficult to ascertain the extent to which users' privacy settings actually match what they believe them to be. The productive dimension of this confusing (or is it deceptive?) implementation of privacy controls is that many Facebook users may be acting as though they are addressing a smaller audience than is actually listening, reading, and watching.
Productive dimension of ambiguous privacy controls
Privacy is assumed in face-to-face situations because of physical limits. The circulation of an unrecorded spoken syllable is limited to the number of listeners paying attention. In such a situation, widespread sharing requires extra effort but in online spaces, it's often easier to share than not to share (boyd & Marwick, 2011, 11). By inscribing one's speech into a networked digital environment, it takes on the characteristics of a networked public: persistence, replicability, searchability, and availability to invisible audiences. boyd & Marwick dub these spaces "public-by-default" (26).
The dimensions of publicity are unstable in public-by-default environments. As the network expands, conditions change but traces of old actions persist (boyd, 2007, 16). A speech act intended to reach one imagined audience lingers and may resurface amid a new one. Although this may be embarrassing for the individual speaker, it may also provoke the emergence of a rich public discourse. The emergence of networked publics (and counterpublics) depends on the circulation of speech acts beyond immediately apprehendable audiences. For better or worse, the lack of coherent privacy settings of Facebook yields a context in which users feel confident to speak as if they are addressing an audience they know personally while the trace artifacts of their discourse circulates to a much wider population of strangers and semi-strangers.
Privacy outside of the platform
Privacy control is not strictly limited to the affordances of a particular platform. In the absence of adequate privacy features, users strive to develop their own social and discursive protocols for refining or transgressing the available privacy structures. Obfuscation, "security through obscurity", and deception through the use of false names, secret profiles, and code words enabled teenagers in boyd's study to evade parental oversight on MySpace (2007, 16). As social network sites are domesticated, users may also establish a multi-channel ecology in which they segregate their discursive activities by platform. In other words, users may strategically restrict certain topics to certain technological contexts. boyd & Marwick's study of Twitter users suggests that this holistic approach may already be widespread among early adopters and highly visible users (2011).
Two of the most powerful tools for circumvention public/private regulations are hardly new: copy-paste and "save as..." Once digital, the circulation of speech acts is remarkably durable. Although the Twitter platform enables users to "retweet" the messages of others into their own network, tweets are just as often extracted from their original context and inserted into email, Facebook, MySpace, blogs, etc. (Marwick & boyd, 4). Porousness among platforms is also achieved through the implementation of interoperable APIs. By linking otherwise distinct services through these alternative interfaces, a message posted in one site may be propagated among several other social network sites. This distribution of messages among services can at once unify and fracture the experience of a social network site as a stable imagined audience. On one hand, users are able to manage their participation in a variety of sites even if they only rarely log in to each one directly. But on the other, the potential for rich discourse is undermined as responses to a single update may be scattered across a variety of inaccessible web services.
Implications and Next Steps
Despite their common features as social network sites, MySpace and Facebook vary significantly as foundations for the emergence of networked publics and counter-publics. The feed-based paradigm, threaded messaging, and unstable implementation of privacy settings on Facebook enable far greater proliferation and wider circulation of discourse than MySpace. Finding this much variation among two frequently compared sites in a preliminary analysis suggests that more work needs to be done developing a robust framework for evaluating the discursive capacity of social network sites. Such a framework would also enable scholars to return to older projects with a renewed sense of technical specificity.
Return to migration
One immediate need is a revision of migration narratives that are quickly calcifying in popular and scholarly accounts of early social network sites like Friendster, MySpace, and Friendster. While early adopters (represented in the earlier quote by Rivlin) offer a techno-centric narrative in which older sites are abandoned because of technical instability, boyd produced a widely-read paper comparing the adoption of Facebook by affluent college-bound teenagers to the "white flight" de-urbanization phenomenon in many U.S. cities during the 1970s. boyd's narrative attends to a recurring concern when talking about online spaces: the reproduction of conventional structures of power. Dahlberg points out that while dominant discourses are present in online spaces, they are likely introduced by the "(offline) subject positioning(s) of participants" rather than specific features of the architecture (56). Early adopters, like those who jump at the first chance to join Facebook, seem very likely to reproduce the normative structures of power as their early adopter status suggests that they are beneficiaries of the status quo.
That MySpace and Facebook are comparable services is taken for granted in discourses of social network migration. These discourses do not take into account the features that distinguish sites from one another. For example, while MySpace fails to generate the volume of discourse found on Facebook, it offers a range of features for musicians and their fans that has never been reproduced in Facebook. Furthermore, because many MySpace profiles are publicly accessible, former users may continue to enjoy the music features of MySpace without actively maintaining an account on the site. How might attention to these features complicate migration narratives and offer new ways to think about the coexistence of MySpace and Facebook?
One participant in boyd's study says that MySpace is "boring" and another reports that he gradually stopped checking MySpace because most of his friends "don't check it that often" (boyd, forthcoming, 1-16). What role does the proliferation of trace artifacts in the Facebook News Feed make Facebook feel less boring and more active than MySpace? To what extent does a feed-based paradigm contribute to a feeling of copresence in a social-network site that is not possible in the profile-based paradigm?
boyd also spoke with teenagers who explained their preference for Facebook as a matter of taste. For many teenagers in her study, MySpace was their first experience with social network sites. How do teenagers' taste-based judgements of social network sites compare with other areas of teen taste? In the years since they adopted Facebook, have their tastes been further refined? Are they now even more discriminating consumers of social network software and services?
Some of the teens in boyd's study deployed racially-coded language in their discussions of taste. Using terms like "hip-hop", "rap", "ghetto", and "bling" to describe the aesthetics of the site, they disparaged MySpace using stereotypes of Black culture. Facebook may appeal to tastes rooted in more affluent White cultures but to what extent might we rethink the comparison of MySpace to "hip-hop"? S. Craig Watkins describes hip-hop as a rich culture in which young people are rewarded for technical sophistication, skills, and mastery. The highly customizable architecture of MySpace may actually reflect the most admirable aspects of hip-hop culture.
Finally, narratives of migration suggest that social network sites vie for user attention in a zero-sum economy. boyd & Marwick's engagement with Twitter users indicates that users are simultaneously paying attention to a variety of sites and services. Knowing this, migration narratives are incomplete without a more narrow definition of what it means for a user to "leave", "abandon", "quit", "move", "migrate", or "flee" from a social network site?[8] Do users delete their old accounts or simply log in less and less often? If the latter, what impact do their aging profiles have on the sense of copresence and vitality in the space? In a feed-based paradigm, news items would appear less frequently but how is absence and abandonment felt in a profile-based paradigm?
Ethical costs and considerations
A second area of future inquiry regards the ethical dimension of these observations. While unintended publicity may have productive effects on the emergence of networked publics, it is clearly a sub-optimal means to achieving this goal. Given a choice, would SNS users prefer tighter control on the circulation of their materials or the experience of participating in a diverse networked public? Anecdotal evidence (the persistence of Facebook in spite of its many privacy gaffs) suggests that users have a high tolerance for unintended publicity. Where are the thresholds and to what extent should scholars concerned with online discourse attempt to identify them in advance of commercial organizations transgressing them?
Additionally, further research is needed to enumerate the cost of managing one's presence in social network sites like Facebook. In all of the examples offered by boyd & Marwick, users expend considerable energy establishing and maintaining systems and boundaries to achieve their preferred degree of privacy and publicity. Some users maintain a complex set of lists to segregate their Facebook users, others regularly "deactivate" their accounts, and still others delete all of their contributions shortly after they post them to the site (18-20). These solutions are effective in the short term but remain contingent on a particular socio-technical arrangement. They cannot scale as networks grow over time and require constant management and upkeep.
Return to Arab Spring example
Lest the capacity of Facebook to support the emergence of networked publics suggest that it is likewise an ideal platform for deliberative democracy, Hashemi-Najafabadi reminds us that the same platforms that enable young Muslims to transgress regulations regarding dress and speech also enable Islamist groups to strengthen and circulate the ideology underlying those regulations (7). Likewise, the same tools that enable popular progressive movements in the U.S. also undergird the proliferation of popular conservative movements. This plurality of publics reflects the "agonistic internet-public sphere position" outlined by Dahlberg. Rather than pursue consensus, the agonistic approach "emphasizes political struggle and conflict as central to democracy" (55). The presence of Islamist, Tea Party, and radically democratic publics within a single social network offers the opportunity for productive encounters with the discourse of strangers. It is in these moments of conflict that participants test the boundaries of dominant discourse and test new discursive possibilities.
Footnotes
- ↑ boyd & Ellison note that this practice is a legacy of the early influence of online dating on the development of social network features.
- ↑ This entire section is informed by boyd & Ellison, 2007.
- ↑ Flickr's "photostream" is a particularly visible example of such a feed.
- ↑ The canonical example of user-driven innovation leading to platform implementation of the thread is the semantic use of # and @ on Twitter.
- ↑ In practice, of course, limits may be imposed by the idiosyncrasies of disk size, available memory, programming language, or moderator preference.
- ↑ The only exception is in sites or services with strict content restrictions. Online newspapers, for example, may censor comments deemed offensive by moderators.
- ↑ This policy is documented here: http://www.facebook.com/help/?faq=12964&tq
- ↑ In 2005, artist Cory Arcangel highlighted the ambiguity of this question by committing "Friendster suicide" in a public performance. In a blog post anticipating the event, he instructed remote audience members to "keep reloading your browser window on Friendster, I think I will simply disappear from your friend list" (Arcangel, 2005).
References
Arcangel, C. (2005). "Friendster Suicide LIVE IN PERSON â Dec 2005". Cory Arcangel's Internet Portfolio Website and Portal. As observed on May 22, 2011 at http://www.coryarcangel.com/2005/12/friendster-suicide-live-in-person-dec-2005/
danah boyd (Forthcoming). "White Flight in Networked Publics? How Race and Class Shaped American Teen Engagement with MySpace and Facebook." In Digital Race Anthology (Eds. Lisa Nakamura and Peter Chow-White). Routledge.
boyd, danah. (2007) âWhy Youth (Heart) Social Network Sites: The Role of Networked Publics in Teenage Social Life.â MacArthur Foundation Series on Digital Learning â Youth, Identity, and Digital Media Volume (ed. David Buckingham). Cambridge, MA: MIT Press.
boyd, danah. 2008. âFacebookâs Privacy Trainwreck: Exposure, invasion, and social convergence.â Convergence: The International Journal of Research into New Media Technologies 14 (1): 13â20.
boyd, danah. 2010. "Making Sense of Privacy and Publicity." SXSW. Austin, Texas, March 13.
boyd, d. m. & Ellison, N. B. (2007). "Social network sites: Definition, History, and Scholarship". Journal of Computer-Mediated Communication, .
boyd, d. & Hargittai, E. (2010). "Facebook privacy settings: Who cares?". First Monday, 15(8), August 2. Retrieved from: http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3086/2589
boyd, d. & Marwick, A. (2011). "Social privacy in networked publics: Teens' attitudes, practices, and strategies". Work-in-progress paper for discussion at the Privacy Law Scholar's Conference, June 2.
Dahlberg, L. (2007). "The Internet, deliberative democracy, and power: Radicalizing the public sphere". International Journal of Media and Cultural Politics, 3(1), 47-64. doi:10.1386/macp.3.1.47/1
Hargittai, E. (2007). "Whose space? Differences among users and non-users of social network sites". Journal of Computer-Mediated Communication, 13(1), article 14. http://jcmc.indiana.edu/vol13/issue1/hargittai.html
Hashemi-Najafabadi, (2010). "Has the information revolution in Muslim societies created new publics?". Muslim World Journal of Human Rights, 7(1), 4.
Haythornthwaite, C. (2005). "Social networks and internet connectivity effects". Information, Communication, and Society, 8(2), pp. 125-147.
Jenkins, H. (2005). Convergence culture. New York: New York University Press.
Marwick, A. ââIâm a Lot More Interesting than a Friendster Profileââ: Identity Presentation, Authenticity and Power in Social Networking Services.â To be presented at Association of Internet Researchers Conference (AOIR 2005). Chicago, IL, USA, 5-9 October 2005.
Marwick, A. (2008). "To catch a predator? The MySpace moral panic." First Monday, 13(6): article 3.
Marwick, A. & boyd, d. (2010). "I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse, and the Imagined Audience". New Media & Society, July 7. doi: 10.1177/1461444810365313
Papacharissi Z (2002) The presentation of self in virtual life: Characteristics of personal home pages. Journalism and Mass Communication Quarterly 79(3): 643â60.
Perez, S. (2007). "Facebook Brags: 35% Adjusted their Privacy Settings". ReadWriteWeb, February 1. Retrieved from: http://www.readwriteweb.com/archives/facebook_brags_35_adjusted_their_privacy_settings.php
Perkel, D. (in press). Copy and paste literacy? Literacy practices in the production of a MySpace profile. In K. Drotner, H. S. Jensen, & K. Schroeder (Eds.), Informal Learning and Digital Media: Constructions, Contexts, and Consequences. Newcastle, UK: Cambridge Scholars Press.
Rivlin, G. (2011). "The Big Question". Wired, May, 19(5), pp. 174-182.
Schrock, A. and boyd, d. (2009). "Online threads to youth: Solicitation, harassment, and problematic content." Enhancing child safety & online technologies (eds. John Palfrey, Dena Sacco, danah boyd). Carolina Academic Press.
Taylor, C. (2002). "Modern social imaginaries". Public Culture, 14(1), Winter, pp. 91-124.
Thelwall, M. & Wilkinson, D. (2010). "Public dialogs in social network sites: What is their purpose?". Journal of the American Society for Information Science and Technology, 61(2), 392-404.
Warner, M. (2002). "Publics and Counterpublics". Public Culture, 14(1), Winter, pp. 49-90.
Mark Zuckerberg, 2009. âAn open letter from Facebook founder Mark Zuckerberg,â Facebook Blog (1 December), at http://blog.facebook.com/blog.php?post=190423927130, accessed 9 July 2010.
Mark Zuckerberg, 2010. âMaking controls simple,â Facebook Blog (26 May), at http://blog.facebook.com/blog.php?post=391922327130, accessed 9 July 2010.
