Faceted Id/entity:
Managing representation in a digital world

danah boyd
MIT Media Lab
Master's Thesis

Thesis document [pdf]
- Abstract
- Introduction
- Negotiating Identity in Social Interactions
- Reconsidering Social Interaction for the Digital Realm
- Self-Awareness in Social Interactions
- Digital Identity Management
- Example Applications
- Social Network Fragments: A Self-Awareness Application
- SecureId: An Identity Management Application
- Conclusion
- Bibliography

Related Projects
- Social Network Fragments
- SecureId

About the author


Chapter 1: Introduction

From its earliest days as a science fiction dream to its current commodified incarnation, the Internet has produced innumerable fantasies about a life free of physical and social constraints. Online society was to be utopian, prompting researchers and cyberanarchists alike to work towards this ideal. Unfortunately, as with all good dreams, we are reaching the moment of waking and becoming aware of the constraints of reality. Cyberspace is not our utopian fantasy; many of the social constraints that frame physical reality are quickly seeping into the digital realm.

Social interaction is a negotiation of identities between people in a given environment. One's identity is comprised of both a personal internal identity and a public social identity. As people engage socially, they project aspects of their internal identity into a social identity for others to perceive. Based on the situation, people only present a particular facet of their internal identity for consideration. Depending on their own need to self-monitor, an individual manages what is to be seen dependent on the environment, thereby creating a social performance where they offer different faces to convey different facets of their identity. The goal of such monitoring is to manage the impressions that others might perceive, to convey the appropriate information at the appropriate time.

In order to assess what is appropriate, people draw from situational and interpersonal contextual cues. By understanding the social implication of context cues and perceiving the reactions presented by others, an individual is given social feedback to adjust their behavior to fit the situation in the hopes of being perceived in the desired light. As people engage socially, they are continually drawing from their own experiences to perceive others and the environment and presenting aspects of their identity that they deem appropriate to the situation. Yet, this negotiation occurs with little conscious effort.

Digital social interaction is not as simple. The underlying architecture of the digital environment does not provide the forms of feedback and context to which people have become accustomed. The lack of embodiment makes it difficult to present oneself and to perceive the presentation of others. As people operate through digital agents, they are forced to articulate their performance in new ways. Additionally, the contextual information that they draw from does not have the same implications online. Situational context can be collapsed with ease, thereby exposing an individual in an out-of-context manner. Unlike physical architecture, the digital equivalent is composed of bits, which have fundamentally different properties than atoms. The interface to the digital world is explicitly constructed and designed around a user's desires. As with any fundamental differences in architecture, there are resultant differences in paradigms of use, interpersonal expectations, and social norms. Performing online requires that people be aware of and adjust to these differences so as to achieve the same level of social proficiency that they have mastered offline.

In this thesis, i begin by expanding on these ideas - drawing on previous work to unpack the ways in which people negotiate social interaction, analyzing the underlying differences between the digital and physical architectures as they relate to sociability, and discussing what adjustments must be made to properly negotiate social interactions in a digital world. In this discussion, i bridge different theories of behavior and communication to offer a new approach for conceptualizing context and context management online.

Using these theoretical ideas as a foundation, i articulate what adjustments i feel are needed in order to provide users with a more sociable environment. In particular, i emphasize the need for self-awareness and identity management capabilities. By being aware of their behavior, individuals are able to monitor their own presentation. Likewise, by having the tools to control what aspects of their identity are presented, people can more appropriately organize their presentation. Awareness and control can provide some of the missing feedback that inhibits certain types of social interaction. My goal in this thesis is to reflect on the existing forms of social feedback and mechanisms by which people engage socially, so as to offer suggestions for designers of digital system to more properly engage in human-centered development.

Assessing my perspective

I recognize that my biases frame the work in this thesis, particularly the perspective that i use to consider social interaction and regulation. As an academic, my notions of social interaction are entirely grounded in a Western, and primarily American perspective. As a researcher and system designer, i draw from a diverse set of disciplines, including sociology, psychology, cultural studies, queer theory, and computer graphics. While i am partially versed in all of these fields, i am by no means an expert in any of them. Yet, i come to this research as a technologist who is delving into the social sciences and as a long-time user of many of the sociable applications that are being discussed. In doing this work, my goal is to bridge the various disciplinary approaches as they relate to digital technology. As an activist in an American context, i value and seek to empower the individual, particularly those who are marginalized.

Issues of privacy and surveillance are embedded in my research. In handling these issues, i value the individual over corporations and governments and seek a privacy approach that makes data transparent to and controlled by their subject. In other words, i believe that an individual has complete rights to their own data and their presentation.

Although i take a performative approach to identity presentation, i do not believe that the individual is inherently fragmented. Instead, i see the modern individual as aware of and reacting to the diverse social climate that we are embedded within. In such an environment, i see the individual as managing multiple facets of their identity. I see social regulation as an operational force in social behavior, where the individual chooses how they react to such reactions. Although i do not believe that the digital world shall be the utopian space in which people can rid themselves of their prejudices, i do believe that it provides a novel social environment that allows people to interact in new ways. At the same time, i see the digital as limiting because of its architectural constraints. Thus, i am inclined to suggest structural and design adjustments to more adequately provide people with the level of expectations that they have developed in physical encounters.

I believe that the role of sociable designers should be to engage and empower users by developing human-centered applications. I believe in working with the needs and expectations of all users, particularly those who are marginalized in physical interactions. Rather than relying on market or legal forces for regulation, i believe in constructing an architectural environment that provides users with the necessary information to regulate primarily through social norms. I believe that users should have the ability to manage and present themselves as they deem appropriate while simultaneously maintaining control over all of their digital expression and content.

Designing social applications requires a fundamental understanding of both social interaction as well as the underlying architecture of digital environments. To understand social interaction requires a deeper understanding of how people perceive themselves and others and what motivates them to interact in particular ways. The digital world is explicitly structured and constructed to meet the needs of its inhabitants. The architecture provides different resources for users, not all of which resemble physical possibilities. Users recognize the digital as a place for social interaction and thus seek to engage socially, often bringing their own assumptions about what the underlying structure provides. While some may argue that the digital architecture should not focus on engaging users socially, i believe that this is a very desirable and valuable application, as it provides a new form of sociability across time and space. It is with this desire in mind that i believe that sociable designers should not only understand what architectural possibilities exist, but have an understanding of how they impact social behavior. With that understanding, they can design a space explicitly intended for social interaction.

It is my belief that sociable designers should focus on developing this understanding in order to empower individuals by designing appropriate interfaces. I am aware that my approach in doing so comes from a Western academic approach and an American understanding of digital life. Thus, my design approach focuses on the needs and interests of the Western world, not because i believe that these issues are not applicable elsewhere, but because i cannot dutifully address them.

Goal and purpose

This thesis focuses on a particular aspect of underlying sociability issues, addressing the role of context, self-awareness, and identity management in social interaction. In doing so, i highlight how the architectural differences of the physical and digital realms affect perception and social behavior. With my biases known, my goal is to assess how sociable designers can present awareness to create a digital environment that more adequately gives users control over their social interactions.

This thesis offers three novel contributions to this area of research. After grounding the discussion in different notions of social interaction, i articulate a new theory of how context operates in the digital realm, focusing on how the underlying architectural differences require a new set of considerations. In particular, i tackle the problems that occur when situational contextual information is collapsed and how users reclaim this. As both contextual feedback and self-awareness are necessary for those seeking appropriate social presentation, i discuss the importance of self-awareness within the digital realm. In doing so, i discuss current approaches to self-awareness and offer a sample design approach for providing self-awareness. Contextual understanding and personal self-awareness are the building blocks that people use to properly control their identity and presentation during social interactions. Next, i focus on the relevance of identity management in giving people control over their social interactions. In this section of the thesis, i discuss current mechanisms for management and control suggest a theoretical framework for conceptualizing these issues. Finally, i introduce and critique two sample applications intended to test my theories. First, i discuss Social Network Fragments, a tool designed to reveal the social network structure that emerges in one's email interactions. Following this, i analyze SecureId as a prototype tool for identity management.

The purpose of this thesis is to delineate important issues that sociable designers should consider when they develop structures intended to encourage social interaction. In doing so, i address both theoretical and computational contributions to this area of research. While i sketch a conceptual model for addressing these issues, the prototypes that are discussed reveal the challenges that we, as researchers and designers, must face. Rather than providing solutions, they expose the weaknesses in this area of research and suggest paths for future research. Explicitly structuring a system for social interaction requires overcoming many obstacles, as each new interface presents new confounding social effects. Attempts to mimic the physical world are flawed because the underlying structure is so different. Yet, to determine how to break from those assumptions and provide users with the necessary information requires far more than an understanding of social behavior. Thus, this thesis only provides the first level of information that is necessary to enrich the social atmosphere of the digital realm.

Motivation for aiding social interaction

From its inception, a primary use of the Internet has been to engage people in social interaction. From email to Usenet to instant messenger, some of the most popular applications have focused on building community and aiding in communication. Yet, while these systems are quite popular, the architecture also restricts the types of social interactions possible online. Online mediums are quite valuable for quickly sharing data, but they prove limited in providing the support necessary for building community. Community requires trust, yet building trust online requires understanding how trust is built and designing systems accordingly (Bos, et. al 2002; Rocco, et. al 2000). At the same time, notions of trust and privacy are not universal, which is problematic both for designers as well as participants. As people do not maintain the same notions of trust, the system must provide for negotiated ideas about trust and privacy.

Trust is complicated by the lack of consistent communication techniques and expectations. As Saville-Troike (1982) noted in reference to physical interactions, differing notions of communicative competence create misunderstandings; this also applies online, although the likelihood of differing communicative expectations is greater. Messages are often misinterpreted, resulting in flame wars or otherwise unnecessary arguments. Yet, assessing a situation for communicative norms requires the ability to determine interpersonal context; online, other people are difficult to see. Just as presence is difficult to ascertain, so are the cues that people tend to embody. Context takes on a different role in the digital realm, as does presentation of one's identity. Lack of embodiment and feedback about the situation and people make the digital world fundamentally different.

These differences should be embraced and appreciated. Although access is not universal, the digital era allows certain groups of people to connect across vast distances in unprecedented ways. Information can be accessed with ease and spread rapidly around the globe. Anyone online can publish their thoughts in a public space and connect with people who have similar thoughts. While the digital opportunities are invaluable, understanding and working with the differences to provide users with a more sociable space can only enhance the possibilities of the digital realm.

My goal is not to replicate physical social interactions, but to learn from them to ascertain what people need and want in social environments. Through experience with the physical world, people have come to understand how they can operate their bodies to convey thoughts. The ease with which people present themselves comes through regular interactions. By understanding the fundamental structures that people use to engage naturally, sociable designers can build systems that provide different, but equally comfortable environments for social interaction. At the same time, while marginalized individuals are limited in what they can convey and how they may convey it offline, they have a new level of freedom online to present themselves without the implications of their bodies automatically associated with their presentation.

I believe that many people are hesitant to join digital communities because of problematic social norms and other fears of privacy. As social norms do not operate as a regulatory force, there are many incidents of people abusing the freedoms that the digital world provides. For example, as discussed in "A Rape In Cyberspace," one individual chose to use his account to harass others, resulting in collective aggravation without a real mechanism for stopping the behavior (Dibbell 1993). Without socially normative regulation or effective feedback channels, the digital environment makes people feel unsafe and powerless.

Rather than requiring that users accommodate for the current interface designs, i believe that designers should assess what people want in their social environments. Current interfaces only address a limited segment of the population and those designs make it difficult for people to maintain properly segmented lives online. In this thesis, i discuss what people are seeking when they go online and propose suggestions for designing such systems, focusing on empowering users through design.

Thesis structure

In order to develop the framework from which sociable designers, including myself, can operate, i begin this thesis by analyzing social interaction. Drawing from various social science approaches, Chapter 2 discusses topics such as the role of one's personality in social interactions. In particular, i take a Goffman-esque approach to discuss the mechanisms by which people perform and negotiate identity, relating this to the relevance of context in determining appropriate forms of interaction, and the ways in which we construct ourselves and others based on that information, our roles, and the facets of our identity. I address theories of self-monitoring, using this to discuss how people differ in the ways in which they negotiate social environments. In this chapter, i focus on physical interactions, using theories that speak almost exclusively to face-to-face interaction. This chapter establishes the framework for this thesis, grounding social behavior in a theoretical discussion. By integrating a diverse set of concepts, the discussion gives the reader the basic background for considering the issues surrounding social interaction in the digital world.

While Chapter 2 is focused on social behavior in the physical world, Chapter 3 explores some of the ways social behavior is altered in the transition to digital interactions. After analyzing some underlying structural differences, this chapter examines how changes in the notion of context affect different aspects of social behavior, from performance to regulation. While situational and interpersonal contextual information is usually available during physical interactions, online it is often missing, misleading or collapsed. Because of its impact on the many arenas of social interaction, contextual differences are crucial for understanding social shortcomings of the digital environment. Thus, i provide an extended example of collapsed contexts and discuss how people attempt to manage contextual information locally. Chapter 3 also addresses the problems that emerge when bodies are no longer the agent through which people negotiate their interactions. Ultimately, Chapter 3 deconstructs the architectural differences in order to address what is missing when context and embodiment are altered.

In order to provide users with appropriate mechanisms for presenting themselves, i propose a two-tiered approach. First, users should have tools to be aware of themselves and others. Second, they should have tools to manage their identity and presentation.

Given this perspective, Chapter 4 introduces digital self-awareness tools. By presenting a selection of current tools, i discuss the motivations of designers in providing users with necessary feedback mechanisms, including visualization tools and data collection systems. From the perspective of empowerment, i critique this area of work and suggest desired improvements. After discussing what types of awareness people desire, i offer an example scenario and tool called Digital Mirror that is intended to provide interactive digital reflection.

Chapter 5 tackles the issues surrounding identity management, again presenting currently existing tools as well as discussing apparent needs. Here, i critique current management systems, most notably Microsoft's Passport. I also discuss why management is necessary for social regulation and articulate what is necessary for users to acquire control over their identity presentation. I suggest a set of design standards that sociable designers should consider in order to more adequately meet the needs of multi-faceted individuals.

Following this, i switch modes to analyze the sample prototypes that i helped design to test these theoretical concepts. Chapter 6 introduces the applications section of this thesis, while Chapter 7 and Chapter 8 detail Social Network Fragments and SecureId, including the conceptual theory, the algorithms and the design approach.

Chapter 7 reflects on the design and concepts behind Social Network Fragments, a visualization tool that i built in collaboration with Jeff Potter. Beginning with an introductory background to social networks, i introduce the motivation behind this awareness tool, reflecting on the importance of social networks in understanding oneself as a multi-faceted individual. From here, i introduce the relevant algorithms and the design of the system so that the curious reader may understand the application. By analyzing the images produced from a sample dataset, i critique the application as a tool for awareness and discuss the issues unveiled in the process of developing the system. Specifically, Chapter 7 critiques the tool from a design perspective, analyzing the problems that arise when conveying highly dimensional data on a visual plane.

By focusing on the design issues that arose in developing SecureId, Chapter 8 analyzes the issues in developing a tool for identity management. First, i discuss conceptual aspects of the system, such as knowledge-based security of identity facets. Following this, i analyze SecureId through a series of images drawn from the prototype. In developing this prototype, i was able to reflect on the amount of work necessary to make the theoretical ideals of Chapter 5 a reality. Thus, the majority of this chapter exposes the problems that i encountered as i set out to design a tool for identity management.

Finally, in Chapter 9, i integrate these ideas, discussing the users' need to have appropriate cues for social interaction, the impact of the digital architecture, and offering an approach for designers that includes giving users self-awareness and management tools. My goal is to motivate designers to focus on designing systems that empower users, as this would only create more desirable sociable environments. While this document articulates much of the conceptual work that must be considered, the applications are embryonic. Thus, throughout this thesis, and most notably in Chapter 9, i argue for further research.