“I’m a Lot More Interesting than a Friendster Profile”: Identity Presentation, Authenticity and Power in Social Networking Services Alice Marwick New York University Department of Culture and Communication alice.marwick@nyu.edu Association Of Internet Researchers 6.0: Internet Generations October 5-9, 2005 Chicago, IL Marwick 2005 Power in Social Networking Services Abstract: The rise in popularity of social networking services (SNS) in the last few years is not unproblematic. The structure and underlying philosophy of social networking services presents two problems regarding user self-presentation of identity. First, the fixity of profiles creates conflict in user self-presentation strategies. SNS privilege a single identity presentation as both “authentic” and “real”, which diminishes user agency. I present a typology of user presentation strategies on Friendster, Orkut, and MySpace that discusses how users navigate this fixity in a variety of ways, and how successful the application architecture is in encouraging a particular type of presentation. Second, I draw from social networking theory to discuss how the presentation of social networks is decontextualized both in terms of relational ties and larger social structures. The social and cultural power inherently embedded in networks is made invisible, replaced with both structural regulation and power (by the application) and political and cultural assumptions in how identity is presented. Additionally, the structure of both application and profiles encourages framing oneself as a consumer and commodifying complicated relationships as social capital. 1. Introduction In the last three years, social networking services (SNSs)1 have rapidly grown in popularity among internet users worldwide. Friendster, the original social networking application, was founded in 2002, currently has a user base of seventeen million people,2 and subsequently spawned a small industry. Some of these websites, such as MySpace and Thefacebook, have surpassed Friendster in popularity by targeting particular populations and incorporating new features on a regular basis. Others, like Microsoft’s Wallop, Rojo, and Yahoo!360, combine social networking with additional sociable functionality like RSS feeds, SNS can be used to stand for either “Social Networking Site” or “Social Networking Service.” This statistic is according to the Friendster corporation. The homepage of Friendster claims 17 million as of May 25, 2005. Friendster, Inc. Friendster. 2005, <http://www.friendster.com/> (25 May, 2005). However, Friendster has recently been overtaken by competitors such as MySpace and Thefacebook; the CEO of Friendster quit on May 25 th, and MySpace now claims 9 million unique visitors a month and has 7 th highest number of page views a month. See “MySpace Ranks Seventh in List of Domains with Highest Number of Pageviews.” WWWCoder. 14 March, 2005. <http://www.wwwcoder.com/main/parentid/472/site/4544/266/default.aspx> (3 May 2005). Like most internet statistics that are provided by proprietary sources, these should be viewed as questionable without independent corroborating data available. 1 2 2 Marwick 2005 Power in Social Networking Services blogs and picture sharing. Overall, there are several hundred SNSs online today, including Tribe.net, Jobster, Dogster and Orkut.3 All of these applications are generally based on a common idea drawn from social networking analysis: that publicly articulated social networks have utility. That is, enabling actors to codify, map and view the relational ties between themselves and others can have useful and positive consequences. SNSs are designed specifically to facilitate user interaction for a variety of goals, mainly dating, business networking, and promotion. However, I maintain that the current generation of social networking software is problematic in several ways, particularly in the types of self-presentation privileged within the applications. The types of self-presentation strategies that the applications allow are directly influenced by the sites’ commercial purposes rather than user needs. As a result, users deploy a variety of strategies in order to increase the utility of the networks and circumvent these commercial-driven assumptions. Nancy Baym notes in her study of Usenet participants that people’s ability to shape their identity is influenced by the medium.4 The structure and underlying philosophy of social networking services presents two problems regarding user self-presentation of identity. First, the fixity of profiles creates conflict in user self-presentation strategies. SNSs privilege a single identity presentation as both “authentic” and “real”, which diminishes user agency. I present a typology of user presentation strategies on Friendster, Orkut, and MySpace that discusses how users navigate this fixity in a variety of ways, and how successful the application architecture is in encouraging a particular type of presentation. Meskill, J. “Home of the Social Networking Services Meta List.” The Social Software Weblog. 14 February 2005, <http://socialsoftware.weblogsinc.com/entry/9817137581524458/> (30 June 2005). 4 Baym, N. Tune In, Log On: Soaps, Fandom, and Online Community. Thousand Oaks, CA: Sage Publications, 2000. 3 3 Marwick 2005 Power in Social Networking Services Second, I will draw from social networking theory to discuss how the presentation of social networks is decontextualized both in terms of relational ties and larger social structures. The social and cultural power inherently embedded in networks is made invisible, replaced with both structural regulation and power (by the application) and political and cultural assumptions in how identity is presented. Additionally, the structure of both application and profiles encourages framing oneself as a consumer and commodifying complicated relationships as social capital. It is important to keep in mind that SNSs are an immature technology. Amin and Thrift write about the “invisibility” threshold of a technology, or the point where it is used without thinking.5 For example, when changing the television channel, people do not think “I am going to pick up and use the remote control which will change the channel using infrared transmission”; rather, they are aware only of the function, changing channels. The technology that facilitates this process has become invisible. In contrast, users of social networking sites are generally focused on the use of the applications, rather than their utility. It remains to be seen whether SNSs will outlive their novelty and become a useful part of internet users’ social structures. 2. The Basics of Social Networking Services A social networking service allows users to publicly articulate and map their relationships between people, organizations, and groups. Although there are differences between the various social networking applications, they tend to have a basic structure in common. A new user begins by creating an account, filling out a profile, searching for other users, and adding people to his or her list of friends. Once people have established a network of “friends”, they become 5 Amin, A and Thrift, N. Cities: Reimagining the Urban. (Cambridge, Massachusetts: Polity, 2002), 58. 4 Marwick 2005 Power in Social Networking Services connected to a larger network of friends-of-friends. Depending on the SNS, users can browse through the profiles of other individuals (some sites restrict your ability to browse to profiles in your “extended network”, or people separated from you by a predefined number of “degrees”). To find specific people, users can search profiles by name or email address or browse the network by a particular set of criteria (high school name, single people under the age of 30 in Memphis, people who like David Bowie). Within the website, users can send messages to each other, chat, post on bulletin boards, and write “testimonials” for their friends. Most SNSs include community features that allow users to converse about shared activities or interests, and others have incorporated weblogs, journals and photo sharing into their feature set. The first, and most popular, SNS is Friendster, which maintains a widespread user base despite architectural shortcomings such as slow load times and frequent outages, and lack of community functionality. MySpace, the second most popular social networking website, claims nine million unique users6 and, while adhering to the general Friendster model, has avoided replicating some of its most obvious problems. It encourages communities and promotional profiles, is faster and more reliable, and has a sleeker, cleaner design. It also allows for a great deal of user customization in profile creation, including adding additional pictures, changing the background and font color and size, and embedding songs and movies. Orkut launched in January 2004 with a great deal of buzz in the early adopter technology community. Like MySpace, it is faster and more usable than Friendster, and it incorporates a variety of community features. However, Orkut is run by Google, the exceedingly successful search engine, which has the resources to make continual improvements to the site, while Friendster remained essentially functionally identical in its first two years of existence. “MySpace Ranks Seventh in List of Domains with Highest Number of Pageviews.” WWWCoder. 14 March, 2005. <http://www.wwwcoder.com/main/parentid/472/site/4544/266/default.aspx> (3 May 2005). See the disclaimer in the second footnote about depending on proprietary statistics for accurate information. 6 5 Marwick 2005 Power in Social Networking Services Before discussing identity presentation, I would like to situate SNS within social network analysis (SNA), which allows researchers to study micro-level social patterns by relating them to macro-level social theories.7 Garton, Haythornthwaite, and Wellman define a social network as “a set of people (or organizations or other social entities) connected by a set of social relationships, such as friendship, co-working or information exchange.”8 Social network analysis involves mapping and measuring relationships between network nodes, or people, entities and groups, to examine information flow across ties. SNA allows researchers to view networks both visually and mathematically9 (Krebs, 2004) to predict information flow, friendship networks, and behavioral patterns. Researchers who use network analysis to study human behavior generally assume that the ways in which actors, or nodes, behave, is dependent on their relationships (ties) and social patterns (structures). When using network analysis to study, for example, historical events or movements, cultural, political and normative structures are generally set aside in favor of looking closely at mapped networks.10 For example, Granovetter uses social network analysis to predict how two people, each connected to a third by a “strong tie”, will behave towards each other. Note that the “strength” of a tie is determined through a set of criteria that includes the amount of time two people spend together, their intensity of emotion, amount of intimacy and “reciprocal services which characterize the tie”11 Take Kristy, an actor, or node, in a social network, and her two closest friends, Claudia and Mary Anne. Granovetter makes the argument that Claudia and Mary Anne 7 Granovetter, M. “The Strength of Weak Ties.” The American Journal of Sociology 78 no. 6 (1973): 1360-1380. Garton, L., Haythornthwaite, C., and Wellman, B. “Studying online social networks.” Journal of ComputerMediated Communication 3 no.1 (1997), <http://www.ascusc.org/jcmc/vol3/issue1/garton.html> (18 February 2003), 2. 9 Krebs, V. “How to do social network analysis.” Orgnet.com. 2004, < http://www.orgnet.com/sna.html> (18 February 2004). 10 Emirbayer, M., and Goodwin, J. “Network Analysis, Culture, and the Problem of Agency.” American Journal of Sociology 99 (1994): 1411-1454. 11 Granovetter, 1971, 1361. 8 6 Marwick 2005 Power in Social Networking Services will, at least, have a weak tie to each other through their mutually strong ties to Kristy. It is very likely that Claudia and Mary Anne will have a strong tie to each other, as an adversarial relationship between the two would strain the two women’s strong tie with Kristy. Thus, the most important factor in predicting Claudia and Mary Anne’s behavior toward each other is their role in the network that connects them to Kristy. At this point it is useful to distinguish “community” from a “social network”. The key element in the social network is the network itself, which may be made up of many communities linked together, or disparate elements that are linked by a single weak tie. Communities, on the other hand, imply a group of people linked by some shared interest or commonality. Although communities may be social networks, social networks are not communities. With this in mind, how does social network analysis approach online interaction? Can it include virtual communities or actors whose primary relationships may be conducted online? Garton, Haythornthwaite, and Wellman argue that social network analysis can easily be applied to social interactions that take place online. For example, SNA can be used to study patterns of information flow, particularly forms that are highly effective online (viral or memetic information flow, for example). Second, the authors posit that online networks are highly useful for establishing weak ties, since the social overhead associated with contacting weak ties is lower online than that in real life. In this way, online social networks may actually be easier to bridge than “real-life” social networks, allowing information to be transferred to larger, broader groups of actors. Finally, online social networks allow people to interact who are separated by physical spatiality or distance, and, may in some cases allow people to bridge social hierarchies 7 Marwick 2005 Power in Social Networking Services as well. Not only is social network analysis useful for looking at online interactions, online interactions are particularly suited for establishing and using social networks.12 3. Identity Presentation Identity presentation within social networking applications takes place primarily within highly structured, multi-modal user profiles. In order to fully understand this presentation, I will use Irving Goffman’s theories of “front stage” and “backstage” identity performance to discuss different parts of the application.13 “Front stage” performances, in Goffman’s analysis, consist of scenarios in which a face is presented publicly, such as a waiter working in a restaurant waiting on customers. “Back stage” performances, on the other hand, take place in private spaces reserved for group members, such as the restaurant kitchen. Students might present a “front stage” identity in class, but present “backstage” while hanging out with other students afterwards at happy hour. In SNSs, front stage performance of identity takes place through profiles, while additional identity information may be conveyed through private messages, emails, or personal meetings. However, because this information is “backstage”, it is not available to the casual observer or researcher. Information about the user’s identity can also be gleaned contextually from the other member’s of the user’s publicly articulated network, but this is dependent both on the information that the other members of the network make public, and how the observer reads the network. Hence, user self-presentation is limited to profile construction.14 12 Garton, L., Haythornthwaite, C., and Wellman, B., 1997. Goffman, E. The Presentation of Self in Everyday Life. (New York: Anchor Books, 1959), 22-30, 111-140. 14 Note that in this section I am drawing examples primarily from Friendster, MySpace, Orkut, and Thefacebook. These services were chosen primarily based on popularity; however, there are hundreds of social networking services and it is outside the scope of this thesis to analyze each one. Although the great majority of social networking services follow the same structure, it is likely that there will be slight differences across applications; these differences are expected and should not contradict the conclusions I come to in this chapter. 13 8 Marwick 2005 Power in Social Networking Services The specifics of profiles differ slightly across applications, but they follow a generally codified structure with three parts: text, pictures and testimonials. The textual aspect consists of profile information that is written by the user and chosen according to the fields provided by the site, generally name, age, geographical location, likes and dislikes, “About Me” and “Who I’d Like to Meet”. The second part, typically labeled a “Photo Album”, allows users to upload and display digital pictures. Finally, testimonials, originally conceived of as a reputation system, are short messages written by the user’s “friends” that appear on the user’s profile, and are in practice a generally open space for varied commentary by others. Looking at the most popular social networking services, MySpace allows users the most customization, as users can configure their page by changing colors, images, and fonts. Some users choose to embed audio and video clips. Friendster recently launched a feature that allows users to choose from a variety of pre-set color schemes for their profile, including “Acid Wash”, “Bad Attitude” and “Marshmallow Peeps”. In the last year or so, most of the SNSs have added features such as blogging and RSS feeds; currently, these are primarily used by early adopters. It will be interesting to see whether these increased outlets for self-presentation influence user strategies. Overall, however, all social networking sites limit the user to an identity presentation that is both highly pre-structured and singular. This is problematic on three levels. First, the rigid profile structure encourages the user to present him or herself in a way that is partly constructed by the application, not the user. Whereas the agency of a person to self-represent is limited in face-to-face communication by social context, power structures, and so forth, there are still a variety of flexible presentation strategies available. For example, accent, body language, speech patterns, linguistic choice and appearance are all user-configurable facets of self-presentation. Social networking sites limit 9 Marwick 2005 Power in Social Networking Services identity presentation to a singular, fixed profile, and most services do not provide users with configuration or customization options to choose their own particular representation strategies.15 MySpace allows customization options to a point, but still only allots one profile per user and provides a set series of fields. Secondly, the way that profiles are structured is not neutral; rather, power is embedded throughout the applications in a variety of ways. Generally, the user is portrayed not as a citizen, but as a consumer. All three applications encourage people to define themselves through the entertainment products they consume: music, movies, books, and television shows. Although both Orkut and Thefacebook include political ideas, both sites define “politics” as simply ranking oneself on a spectrum of political identity (very liberal, liberal, moderate, conservative, and very conservative). Not only are users treated as consumers, they are encouraged to consume others in a concept of networking that privileges social capital over friendship or community building. “Networking”, in business terms, is a goal-oriented process in which one’s social circle is constantly expanded in order to connect with as many people as possible, in order to gain business advantages. Many professional organizations have networking evenings where members can quickly meet-and-greet a variety of people in their field who may be able to connect them to useful information or resources, or vice versa. Networking ideology commodifies relational ties and encourages amassing as many contacts as possible without deepening connections between actors in order to “bridge” disparate networks. It is this aspect of social networking that social networking applications are attempting to capture and apply across 15 See Orkut, Friendster, and Thefacebook for examples. For more specifics on profile fixity, see boyd, d. “Friendster and Publicly Articulated Social Networks.” Conference on Human Factors and Computing Systems (CHI 2004). Vienna: ACM, April 24-29, 2004. 10 Marwick 2005 Power in Social Networking Services a larger set of social phenomena, including business, but extending to friendship, romantic relationships, and community activities. Third, SNSs inherently exclude certain segments of the world population. For instance, the majority of sites are American applications that attract primarily US users.16 Orkut has a very high percentage of South American, Middle Eastern and Asian users, presumably due to its ties to Google, a site that is localized for international use and boasts an enormous international user base. Additionally, since all SNSs require internet access, their user base is inherently limited to a certain segment of the population, cutting non-Internet users out of the network completely. We could apply this criticism to all internet applications, but I think it is particularly egregious in social networking as the utility of the network is diminished as non-internet enabled individuals are excluded completely. Even if these people may be fully integrated into an offline social network, such as place of business or group of friends, once that network is shifted to the online sphere, they are excluded. Finally, none of the most popular social network sites have a profile field for racial or ethnic identity. Although this means users have agency over how they wish to represent their ethnic background, it may set up whiteness as normative, given that within American culture, white ethnic identity is usually privileged as “normal”. Beth Kolko writes that when race is erased from the online landscape, it is usually replaced by a presumption of whiteness.17 None of these cultural assumptions are addressed or discussed within social networking applications. 4. Embedded Power 16 Thefacebook limits its user base to particular colleges. Right now the great majority of the colleges on the network are American, although there are a few universities in the U.K. and Canada. It is likely that more colleges outside the US will be added in the future. 17 Kolko, B. E. “Erasing @race: Going White in the (Inter)Face.” In Race in Cyberspace, ed. Kolko, B. E., Nakamura, L., and Rodman, G. B, 213-232. New York: Routledge, 2000. 11 Marwick 2005 Power in Social Networking Services Social networking applications do not take into account critiques of social networking analysis that claim it decontextualizes relationships and social actions from their cultural, discursive and normative frameworks. Emirbaya and Goodwin, in their 1994 critique, suggest that social networking analysis ignores both human agency and external influences on social networks, such as culture, political discourse and societal norms. While human relationships are inherently fraught with Foucault’s “capillaries of power” which map to larger political, social and cultural norms, these are removed from the publicly articulated networks drawn within SNS. Emirbaya and Goodwin maintain that these “discursive frameworks and cultural maps”18 are equally as powerful as social networks, and, indeed, limit or enable the actions of individual people as much as the network itself does. Assuming that social networks are discretely explanatory for human behavior, then, ignores not only the influence of systemic power relations related to gender, sexuality, race, class etc. on behavior, but also how the subject’s own ability for empirical action is influenced by the larger interrelated context in which he or she is situated . Social networking applications remove these “webs of power” while simultaneously exposing identity self-presentation and relational ties, with the result of removing value and signification from the network. Discursive frameworks are always temporally and spatially located, meaning that what can be said, and how it is said, is highly context-specific. Removing relational commentary from its context diminishes, then, the embedded meaning within the relationship. Additionally, assuming that all relationships and social ties are equal ignores the reality of power differentials, particularly systemic oppressions like racism, sexism, classism, and homophobia. Emirbayer, M., and Goodwin, J. “Network Analysis, Culture, and the Problem of Agency.” American Journal of Sociology 99 (1994): 1427. 18 12 Marwick 2005 Power in Social Networking Services While it would be very difficult, and probably unpleasant, to design an application that maps micro-level power relations, subcultural identifications, or systems of oppression, it might still be worthwhile to take into account that power is embedded in social networks, this power is not equal, and that these power differentials play out on human actors. Danah boyd relates an anecdote where neo-Nazis used Friendster to identify and harass people of color.19 In the threedegree model, the fact that these white power advocates and these particular people of color were related through mutual friends would predict a successful social relationship. This, however, was of secondary importance to the cultural factors – racism – that defined their interactions. Ignoring socially contextualized power relations does not mean that their impact will not be felt by users in the networks. Although software designers may not want to replicate these embedded power structures, it seems important that there is acknowledge that they affect relationships. 5. The Problem of Authenticity Social networking sites overall presume that each user has a single “authentic” identity that can be presented accurately. Friendster, for example, strongly discourages any profiles that they deem to be “inauthentic”, and subsequently purged a great many of them from the site in 2003.20 Likewise, Thefacebook requires users to provide a working, current *.edu email address to verify that the user is a student at a particular university. But what is the authentic? Presumably, it is a truthful and “real” picture of oneself, with accurate information as to one’s name, age, sexuality, hometown, likes, and dislikes. Other social applications do not limit their users in the same way; online personal sites, for example, allow each user to create multiple 19 Boyd, danah. "Revenge of the User: Lessons from Creator/User Battles." Talk at Emerging Tech 2004. San Diego, California: February 11, 2004. 20 Mieszkowski, K. “Faking Out Friendster”. Salon.com. 14 August 2003, <http://www.salon.com/tech/feature/2003/08/14/fakesters/print.html> (7 May 2005). 13 Marwick 2005 Power in Social Networking Services profiles so they can present themselves in different ways. SNSs, in contrast, allow one profile per email address, and limit the user’s ability to change it based on audience. This fixity can be quite difficult to navigate when one is used to representing him or herself in multiple ways. This becomes most problematic when considering a completely exposed, articulated public network. The persona one presents for a family member or professional colleague differs greatly from the way one’s identity operates in, for example, a romantic context. Consider the concept of passing, discussed in the first chapter. Passing assumes that one chooses to reveal certain aspects of his/her identity at a particular time, often for reasons of safety. An openly gay man, for example, may use Friendster to look for possible romantic partners. However, as open as he may be in his personal life, he may not be completely “out” in his professional life, or to his parents or hometown contacts. Revealing his queerness becomes a strategic move, depending on context. Social networking applications remove this option, resulting in a lack of agency that can have real-world implications (losing a job, a friend, or parental approval). Cyberculture studies is rife with the concept of “identity play”, the idea that expressing identity online is liberatory, revealing, and somehow unique to the internet realm. Juliet Davis further critiques this view of technology as a utopian playground in which identity can be freely played with and transgressed. She writes “this idea follows the form-as-substance myth: i.e., that we can simply leave our history and cultural trappings behind and take on a new nature merely by taking on its visual form; that virtual reality provides the phenomenological experience necessary to approximate constructions of those identities in socially resonant ways.” 21 She is discussing virtual reality, but her ideas apply to social networks as well. Identities do not change 21 Davis, J. Myths Of Embodiment And Gender In Electronic Culture. 2004, <http://www.julietdavis.com/VC/paper.html> (1 June 2004). 14 Marwick 2005 Power in Social Networking Services just because they are expressed through software; they are still subject to the same power relations and problems that they are in the real world. Limiting each person to a single profile and treating each in the same way ignores the fact that people live in a contextualized world, with implications to match. Instead of idealized sites where power is non-existent and all people are treated equally, a new type of power is re-inscribed within technology: the problem of visibility and policing, and the fact that the technology is not at all neutral. 6. Enter the Users Social networking applications are rich sites of user presentation strategies. Morningstar and Farmer, in their study of Habitat, an early graphical multi-user application, point out that internet users often interact with sites in different ways than those intended by the site creators.22 For example, danah boyd examines the use of fake profiles on Friendster, and how these “Fakesters” allow users to transgress the site’s structural boundaries and engage in identity play.23 Additionally, users may choose particular linguistic styles (irony, sarcasm, satire) when constructing their profiles to compensate for the rigid structure of the profile format, or add digital pictures that represent themselves in strategic ways. Other people identify themselves within SN applications by using aliases that would only be known by people already within their social network. Others create “safe profiles”, where publicly exposed information is carefully chosen to avoid possible problems. This variety of resistance strategies shows that users are creatively and actively engaging with social networking services in multiple ways. Boyd discusses how users creatively circumvented Friendster’s lack of community features with the advent of “Fakesters”. She identifies three types of Fakesters, which are Morningstar, C., and Farmer, F. The Lessons of Lucasfilm’s Habitat. In Cyberspace: First Steps, M. L. Benedikt ed. Cambridge: MIT Press, 1991: 273-302. 23 boyd, d. “Friendster and Publicly Articulated Social Networks.” Conference on Human Factors and Computing Systems (CHI 2004). Vienna: ACM, April 24-29, 2004. 22 15 Marwick 2005 Power in Social Networking Services generally Friendster profiles constructed by users to represent “cultural characters as reference points” (Wonder Woman, George Bush), places or collectives that represent communities (Wellesley College, MIT), and ostensibly “authentic” profiles that are meant to be perceived as real.24 When a user becomes friends with a Fakester, he or she is connected to a larger group of people who share that Fakester in common. Fakesters additionally allow certain users to perform as different identities rather than being confined to their singular, fixed profile. However, the founders of Friendster believed that the Fakester concept diluted the utility of the site, and removed many accounts, to widespread user protest. Boyd points out that this attempt to “configure the user” has not been particularly successful. Other SNS are less concerned with Fakesters and it is likely that the community features built into Tribe, MySpace and Orkut will diminish the necessity for them to function as community bridges. However, it will be interesting to see how the idea of performed, “inauthentic” identity adoption plays out in social networking applications in the future. These strategies are successful in varying degrees, but I mention them in order to make room for the agency of the subject when discussing social networking software as a series of texts. I am interested in how users engage with sites and actively create certain selfpresentations. Despite the structural limitations, SNS users actively construct and perform their identities through a series of choices that they make from the “bottom-up” rather than “top down”; in other words, although SNS construct the framework of identity presentation, the users do have broad linguistic choice when considering the discursive styles they wish to use in identity performance. The multi-modality of user profiles, including pictures, text and testimonials, also allows a certain room for flexibility within SNS applications. For example, users may choose to write their profile using a sarcastic, ironic style, while selecting digital 24 boyd, 2004, 3. 16 Marwick 2005 Power in Social Networking Services pictures that convey a similarly satirical tone. In this way, the “authenticity” of a singular profile which is inherently “inauthentic” may be called into question. Finally, the process by which people interact with cultural products is always reciprocal. To argue that the highly structured nature of the applications denies the users all agency would be mistaken. 7. A Typology Of Social Networking Service Profiles In order to take a closer look at the choices users make while constructing their identities, I constructed a typology of SNS profiles. The goals for this exercise were to determine roughly how many users chose to create authentic profiles rather than Fakester profiles, to look at a few strategies used by users in profile construction, and to see if these choices and strategies differed by application. I chose three SNS to look at: Friendster, the largest; MySpace, which allows users the greatest amount of profile customization; and Orkut, the newest and most innovative. I randomly selected 30 profiles from each application and analyzed the pictures, text, and testimonials on each. I noted the strategies used by the user and categorized each profile accordingly. It is important to note that I am using a small sample and my findings are not, therefore, generalizable to the SNS population as a whole; however, they are a starting point to provide researchers and interested parties with a taste of how users work with profile structures within social networking services. Generally, the great majority of users were American, most between the ages of 18 and 30. Almost all the users on Friendster and MySpace were American, while Orkut had the largest amount of international users. Ethnicity and race varied, with white, Asian-American and Hispanic users having the most visible presence. 17 Marwick 2005 Power in Social Networking Services I divided profiles into two general categories, Authentic or Fakester, by determining whether the profile depicted a “real” person, or celebrity, place, thing, community or so forth. I then further divided the “real”, or Authentic, profiles by self-presentation strategy, whether the identity was presented straightforwardly or whether the user engaged in satire, parody, irony or sarcasm while presenting information (Authentic Ironic). The Fakester profiles differed so wildly that it was hard to categorize them accurately; generally, they portrayed characters, communities, or things. SNS Friendster MySpace Orkut Authentic 20 23 29 Fakester 10 7 1 Table 1: Authentic vs. Fakester Profiles by SNS SNS Friendster MySpace Orkut Authentic 15 16 29 Authentic / Ironic 5 7 0 Table 2: Authentic vs. Authentic Ironic by SNS Authentic An Authentic profile is one in which the user includes legitimizing personal information and characteristics such as their “real name” and “location” to further the perception of “authenticity”. The great majority of profiles (72 out of 90, or 80%) fall into this category, since most users create profiles in ways intended by the application. These users enter their (presumably) real names, pictures, and identifying information, and do not attempt to “play” fictional characters, celebrities, things, groups, or communities. It is important to mention that 18 Marwick 2005 Power in Social Networking Services “authenticity” is a problematic concept, and, like identity, varies based on context.25 However, the users who present in this “authentic” manner are presumably making self-presentation decisions based on their assumptions of context and audience. Furthermore, it is likely that the majority of users will not run into problems within social networking applications. However, with the rise of mobile social software and other liminal online/offline applications, convergence between online and offline social structures is likely to increase; similarly, “structures of conflict” (or applications that attempt to “configure the user”26 in manners that are personally or socially problematic) will multiply as this process continues. Authentic Ironic A subgroup of the Authentic profiles, the Authentic Ironic profile is one in which a user is generally performing as themselves, but uses sarcasm, irony, or satire as a modifying strategy. For example, the user may use a funny picture of a celebrity as their default photograph, identify themselves with a pseudonym, or state that they are a hundred years old. However, the user is still “themselves”. In other words, their friends will know to look for them under this profile, and usually their testimonials will bear this out by including their real name. Most Authentic Ironic profiles include a mix of “authentic” and “ironic” information, for example using a name like “GreenpotBluepot” but including accurate pictures, or stating one’s occupation as “sheepherder” while maintaining that their location is Manhattan. Users may create an Authentic Ironic profile for a variety of reasons, including amusing their friends, trying to act cool (if such profiles are normative within their larger social structure), for fun, or in order to mask their true information from people they may not want viewing it (their parents, for example). 25 For an excellent and in-depth discussion of authenticity, see Grazian, D. Blue Chicago. Chicago: University of Chicago Press, 2003. 26 Danah boyd’s terminology. 19 Marwick 2005 Power in Social Networking Services Fakesters I categorized all non-Authentic profiles as Fakesters. Into this category fell celebrities, objects, places, activities, and obscure in-jokes. This category had the most variation in profiles, ranging from Friendster profiles that served as substitute communities (such as a college dorm) to promotional profiles for dance club nights or magazines. Friendster had the largest amount of Fakesters, with a full third of the profiles I looked at categorized as fake. MySpace had seven Fakesters, but only one of those was a celebrity, with four promotional profiles and two group profiles. Orkut had only a single Fakester, the 2004 Democratic presidential candidate John Kerry. Fakester Baylor Texas Hold ‘Em Whitehouse Goat The F.O.C. Kit DeLuca Dalai Lama Roy Keane Pratt Institute Scaffolding The Cure Type Place, Community Activity In-Joke Thing In-Joke Fictional Character Celebrity Celebrity Place, Community Celebrity Table 3: Fakester profiles on Friendster, By Type The wide variety of Fakesters on Friendster, as shown in Table 3, shows the different functions that they serve the user base, probably due to Friendster’s lack of community features. Some Fakesters are solely in-jokes created by a group of friends, such as the F.O.C, which appears to be a gang of friends of some kind, and is filled with information that would probably only make sense to people who know the group in question. Others are fictional characters, such as Kit DeLuca, a supporting character in the 1990 romantic comedy Pretty Woman. Her profile is 20 Marwick 2005 Power in Social Networking Services filled out with information gathered from the movie, such as songs from the soundtrack as her favorite movie. Still others are designed to bring people with a common interest together, such as Texas Hold Em, a card game, and The Cure, a band. The Cure’s profile reads “Cure Fans Unite!” rather than pretending to be a profile written by the members of the band themselves. Finally, there are community Fakesters based on common places; both of the ones I looked at were college-related, in keeping with the general demographic of social networking services. Differences by application I found that user presentation strategies differed by application. Friendster had by far the most Fakesters. This again is probably due both to the lack of community features on the site and the extremely rigid structure of the Friendster profile structure, which allows for very little customization.27 MySpace, which allows a great deal of customization as well as community features, had fewer Fakesters, and of those Fakesters most were community-based or promotional rather than celebrities or objects. Orkut, however, had only one Fakester. This is probably due to the user base and community features of Orkut rather than customization, since Orkut has a very rigid profile structure. MySpace and Friendster had about the same amount of Authentic Ironic profiles, which is probably due to the similarity of their respective user base. Orkut had no ironic presentations, which seems surprising, but not when the users are considered. At the time this typology was constructed, Orkut was an invite-only community primarily patronized by academics, early adopters, bloggers and technology workers, who tend to be older and more interested in the networking and research implications of social networking rather than socializing, getting dates, or appearing “cool” to their peers. Additionally, the international nature of Orkut may mean I 27 Note that Friendster has since integrated a larger variety of customization options for its users. 21 Marwick 2005 Power in Social Networking Services missed a certain amount of irony since many of the profiles I looked at were written in languages other than English. Humor does not tend to localize well. Overall, I found that the way that users construct their identities in SNS is at least somewhat encouraged by the applications, as shown by the variance in self-presentation strategies by application. Since each application views the “authentic” differently, one’s identity, then, is inscribed within the software, rather than being inherently tied to one’s “self” or “body”. Additionally, the social context of the application itself may change the way people choose to present their identities. The young, hip MySpace network may encourage a cool, ironic selfpresentation, while the older and more academic Orkut environment leans towards authenticity and serious discussion. Erving Goffman wrote that the two primary factors influencing self-presentation choices and strategies were context and audience.28 As previously discussed, people change their dramaturgical performance based on who they are interacting with, and the context (environment, social structure) in which the interaction takes place. These factors are equally applicable to online environments, and, in fact, help to explain why users might choose to create a type of profile that either obfuscates their “authentic” identity or presents but a single facet of what is a truly multifaceted personality. Within social networking services, the audience is not immediately apparent. Users construct even an “authentic” profile based on their assumptions about the audience, which might include people who they already know within the network, or their friends who are active internet users. More savvy users understand that there is no way to determine who might see their profile, and so creating a fake or deliberately abstruse profile may be a way to circumvent potential conflicts with audience. Since most social networking sites lack the functionality that allows users to manage multiple profiles or vary information based on 28 Goffman, 239-240. 22 Marwick 2005 Power in Social Networking Services the viewer, Fakesters and authentic/ironic profiles are creative ways to manage necessary multiplicity in self-presentation. For example, using a nickname that is familiar only to one’s core group of friends prevents the user from being found by people outside that group. We can understand context in social networking sites in three different ways. First, the architecture or construction of the application has a great deal of influence over the selfpresentation strategies available for the user. MySpace, for example, has advanced customization features which allow for a much wider variety of self-presentation options than do Thefacebook or Friendster. As a result, MySpace profiles differ in terms of embedded multimedia, additional functionality (animated cursors or JavaScript calls), supplemental images, and the like. Similarly, Friendster’s early lack of community features led some users to create Fakester profiles in response. Secondly, different applications have different instrumental uses and thereby encourage particular types of singular self-representation. SNS that are specifically designed for job searching, such as LinkedIn, Ryze, or Jobster, encourage (explicitly or implicitly) highly professional self-representation, while users of Thefacebook, which is targeted towards college students, tend to create playful, humorous profiles. Thirdly, locating the user within a particular network of other users can provide clues to their self-presentation strategies. Browsing a social networking service, one can see commonalities between groups of friends: some post primarily sexy pictures of themselves, while other groups create deliberately abstruse profiles that function as social capital within their peer group or contain specific symbolic markers. Self-presentation strategies are influenced by both the application’s structure and the influence of one’s social group. 8. Conclusion 23 Marwick 2005 Power in Social Networking Services The first generation of social networking applications, while providing a certain degree of utility for dating and job searching, encourage a commodified, fixed, singular view of identity presentation that limits their usefulness for network mapping and relationship building. Each SNS privileges “authentic” user identity presentation, which assumes a normative, decontextualized concept of self. However, users navigate these assumptions in a variety of ways, some choosing alternate methods and self-presentation strategies that challenge particular ways of configuration by SN applications. The way that applications frame identity greatly influences the ways that people present their identities online, as do the other users. It is to be hoped that future versions of SNS will allow the user greater flexibility and agency. However, it is important to remember that people may not always present their identities, even given huge leeway, in ways that are inherently progressive or liberatory. However, the awkwardness and difficulty inherent in representing oneself in fixed, codified ways will need to be addressed if social networking applications are to move beyond the current fad to become truly useful services for internet users. There is a tendency among pundits and scholars to see technology as a neutral, or utopic good; excitement about novelty does not eliminate the need for scrutiny. 24 Marwick 2005 Power in Social Networking Services Bibliography Amin, A and Thrift, N. Cities: Reimagining the Urban. Cambridge, Massachusetts: Polity, 2002. Baym, N. Tune In, Log On: Soaps, Fandom, and Online Community. Thousand Oaks, CA: Sage Publications, 2000. boyd, d. “Friendster and Publicly Articulated Social Networks.” Conference on Human Factors and Computing Systems (CHI 2004). Vienna: ACM, April 24-29, 2004. boyd, d. "Revenge of the User: Lessons from Creator/User Battles." Talk at Emerging Tech 2004. San Diego, California: February 11, 2004. Davis, J. Myths Of Embodiment And Gender In Electronic Culture. 2004, <http://www.julietdavis.com/VC/paper.html> (1 June 2004). Emirbayer, M., and Goodwin, J. “Network Analysis, Culture, and the Problem of Agency.” American Journal of Sociology 99 (1994): 1411-1454. Friendster, Inc. Friendster. 2005, <http://www.friendster.com/> (25 May, 2005). Garton, L., Haythornthwaite, C., and Wellman, B. “Studying online social networks.” Journal of Computer-Mediated Communication 3 no.1 (1997), <http://www.ascusc.org/jcmc/vol3/issue1/garton.html> (18 February 2003). Goffman, E. The Presentation of Self in Everyday Life. (New York: Anchor Books, 1959). Golder, S. A Typology of Social Roles in Usenet. Unpublished senior honors thesis, Harvard University, Massachusetts, 2003. Granovetter, M. “The Strength of Weak Ties.” The American Journal of Sociology 78 no. 6 (1973): 1360-1380. 25 Marwick 2005 Power in Social Networking Services Grazian, D. Blue Chicago. Chicago: University of Chicago Press, 2003. Kolko, B. E. “Erasing @race: Going White in the (Inter)Face.” In Race in Cyberspace, ed. Kolko, B. E., Nakamura, L., and Rodman, G. B, 213-232. New York: Routledge, 2000. Krebs, V. “How to do social network analysis.” Orgnet.com. 2004, < http://www.orgnet.com/sna.html> (18 February 2004). Meskill, J. “Home of the Social Networking Services Meta List.” The Social Software Weblog. 14 February 2005, <http://socialsoftware.weblogsinc.com/entry/9817137581524458/> (30 June 2005). Mieszkowski, K. “Faking Out Friendster”. Salon.com. 14 August 2003, <http://www.salon.com/tech/feature/2003/08/14/fakesters/print.html> (7 May 2005). Morningstar, C., and Farmer, F. The Lessons of Lucasfilm’s Habitat. In Cyberspace: First Steps, M. L. Benedikt ed. (Cambridge: MIT Press, 1991) 273-302. “MySpace Ranks Seventh in List of Domains with Highest Number of Pageviews.” WWWCoder. 14 March, 2005. <http://www.wwwcoder.com/main/parentid/472/site/4544/266/default.aspx> (3 May 2005). 26