http://nms.sagepub.com/
New Media & Society
http://nms.sagepub.com/content/early/2010/06/22/1461444810365313
The online version of this article can be found at:
DOI: 10.1177/1461444810365313
published online 7 July 2010New Media Society
Alice E. Marwick and danah boyd
the Imagined Audience
I Tweet Honestly, I Tweet Passionately: Twitter Users, Context Collapse, and
Published by:
http://www.sagepublications.com
can be found at:New Media & SocietyAdditional services and information for
http://nms.sagepub.com/cgi/alertsEmail Alerts:
http://nms.sagepub.com/subscriptionsSubscriptions:
http://www.sagepub.com/journalsReprints.navReprints:
http://www.sagepub.com/journalsPermissions.navPermissions:
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Article
Corresponding author:
Alice E. Marwick, 239 Greene Street 7th Floor, NewYork, NY 10003, USA.
Email: alice.marwick@nyu.edu
I tweet honestly, I tweet
passionately:  Twitter users,
context collapse, and the
imagined audience
Alice E. Marwick
NewYork University, USA
danah boyd
Microsoft Research, USA
Abstract
Social media technologies collapse multiple audiences into single contexts, making it
difficult for people to use the same techniques online that they do to handle multiplicity
in face-to-face conversation. This article investigates how content producers navigate
‘imagined audiences’ on Twitter.We talked with participants who have different types
of followings to understand their techniques, including targeting different audiences,
concealing subjects, and maintaining authenticity. Some techniques of audience
management resemble the practices of ‘micro-celebrity’ and personal branding, both
strategic self-commodification. Our model of the networked audience assumes a manyto-many
communication through which individuals conceptualize an imagined audience
evoked through their tweets.
Key words
audience, context, identity, micro-celebrity, self-presentation, social media,Twitter
Introduction
We present ourselves differently based on who we are talking to and where the conversation
takes place – social contexts like a job interview, trivia night at a bar, or dinner with
a partner differ in their norms and expectations. The same goes for socializing online.
new media & society
XX(X) 1–20
© The Author(s) 2010
Reprints and permission: sagepub.
co.uk/journalsPermissions.nav
DOI: 10.1177/1461444810365313
http://nms.sagepub.com
New Media Society OnlineFirst, published on July 7, 2010 as doi:10.1177/1461444810365313
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
2		 new media & society XX(X)
Participants have a sense of audience in every mediated conversation, whether on instant
messenger or through blog comments. This audience is often imagined and constructed
by an individual in order to present themselves appropriately, based on technological
affordances and immediate social context. Studies of identity presentation on profilebased
sites, such as social network sites and personal ads, have demonstrated that profile
owners are attentive to audience (e.g. boyd, 2006b; Ellison et al., 2006). The need for
variable self-presentation is complicated by increasingly mainstream social media technologies
that collapse multiple contexts and bring together commonly distinct audiences.
This paper examines how people using the microblogging site Twitter imagine their
audiences and what strategies they use to navigate networked audiences. Our findings
shed light on how audience changes in networked environments.
Imagining the audience online
Every participant in a communicative act has an imagined audience. Audiences are not
discrete; when we talk, we think we are speaking only to the people in front of us or on the
other end of the telephone, but this is in many ways a fantasy. (Social norms against eavesdropping
show how ‘privacy’ requires the participation of bystanders.) Technology complicates
our metaphors of space and place, including the belief that audiences are separate
from each other. We may understand that the Twitter or Facebook audience is potentially
limitless, but we often act as if it were bounded. Our understanding of the social media
audience is limited. While anyone can potentially read or view a digital artifact, we need
a more specific conception of audience than ‘anyone’ to choose the language, cultural
referents, style, and so on that comprise online identity presentation. In the absence of
certain knowledge about audience, participants take cues from the social media environment
to imagine the community (boyd, 2007: 131). This, the imagined audience, might be
entirely different from the actual readers of a profile, blog post, or tweet.
Joshua Meyrowitz’s germinal work No Sense of Place (1985) applied situationist
theory to the changes brought about by electronic media like television and radio.
Situationism maintains that people react to situations based on context rather than fixed
psychological traits. Situations, and thus social order, are collectively produced by participants
(Garfinkel, 1967). Meyrowitz theorized that electronic media eliminated walls
between separate social situations, contributing to the rapid social change that characterized
the United States in the 1960s. Similarly, self-presentation theory has been used to
understand the further combination of audience by digital media. Self-conscious identity
performances have been analyzed in internet spaces like social network sites (boyd, 2007;
Livingstone, 2005), blogs (Hodkinson and Lincoln, 2008; Reed, 2005), dating sites
(Ellison et al., 2006) and personal homepages (Papacharissi, 2002; Schau and Gilly, 2003).
Personal homepages, arguably the first multi-media online identity presentations, are
highly managed and limited in collaborative scope; people tend to present themselves in
fixed, singular, and self-conscious ways. Papacharissi describes the personal homepage as
‘a carefully controlled performance through which self presentation is achieved under
optimal conditions’ (2002: 644). Schau and Gilly found that personal homepage creators
thought of their work as constructed for the public; even if they focused on friends or family,
creators ‘acknowledge(d) the potential for the audience to be unlimited and undefined’
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 3
(2003: 391). Similarly, Robinson argues ‘the very construction of the homepage presumes
the expectation of the virtual “generalized other”’ (2007: 104). She notes that personal
homepages are more atomized and isolated than blogs and message boards, which
presume ongoing communication with the audience. She writes, ‘the “I” is constantly
redefined as the “me” in response to this interactional commentary’ (2007: 104).
The specifics of the imagined audience are more important in social media that
involve greater interaction with readers than personal homepages. Professional writers’
sense of ‘audience awareness’ factors greatly into their writing, in terms of goals, vocabulary,
technique, and subject matter (Berkenkotter, 1981). Like many writers, bloggers
write for a ‘cognitively constructed’ audience, an imagined group of readers who may
not actually read the blog (boyd, 2006a). Actual readers are present only in digital
ephemera like server logs or comments. The imagined audience exists only as it is written
into the text, through stylistic and linguistic choices (Scheidt, 2006). Through the
process of labeling connections as ‘Friends’, social network sites require participants to
publicly articulate connections, thereby enabling them to write their audience into being
(boyd, 2006b).
In goal-oriented spaces like dating sites, people are highly conscious of audience.
Ellison et al. (2006) found that personal ads were constructed with a hyper-aware selfconsciousness,
as users knew that misspellings, cultural references, and even time stamps
were likely to be scrutinized by potential suitors. Similarly, social network site users
select ‘markers of cool’based on an imagined audience of friends and peers. Liu’s (2007)
study of ‘taste cultures’ on social network site profiles found that participants listed
favorite books, music, movies, and TV shows to construct elaborate taste performances,
primarily to convey prestige, uniqueness, or aesthetic preference.
The microblogging site Twitter affords dynamic, interactive identity presentation to
unknown audiences. Self-presentation on Twitter takes place through ongoing ‘tweets’
and conversations with others, rather than static profiles. It is primarily textual, not
visual. The potential diversity of readership on Twitter ruptures the ability to vary selfpresentation
based on audience, and thus manage discrete impressions.
Twitter
Twitter is a microblogging site, originally developed for mobile phones, designed to let
people post short, 140-character text updates or ‘tweets’ to a network of others. Twitter
prompts users to answer the question ‘What are you doing?’, creating a constantlyupdated
timeline, or stream, of short messages that range from humor and musings on
life to links and breaking news. Twitter has a directed friendship model: participants
choose Twitter accounts to ‘follow’ in their stream, and they each have their own group
of ‘followers’. There is no technical requirement of reciprocity, and often, no social
expectation of such. Tweets can be posted and read on the web, through SMS, or via
third-party clients written for desktop computers, smartphones, and other devices. These
different access methods allow for instant postings of photos, on-the-ground reports, and
quick replies to other users. The site was launched in 2006, and broke into the mainstream
in 2008–09, when accounts and media attention grew exponentially. In May
2009, The Nielsen Company reported that Twitter had approximately 18.2 million users,
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
4		 new media & society XX(X)
a growth rate of 1448 percent from May 2008. Today, the most followed Twitter accounts
represent public figures and celebrities, from US President Barack Obama to actor
Ashton Kutcher and pop star Britney Spears.
Twitter and audience
As in much computer-mediated communication, a tweet’s actual readers differ from its
producer’s imagined audience. Twitter allows individuals to send private messages to
people they follow through direct messages (DMs), but the dominant communication
practices are public. A convention known as the ‘@reply’ (consisting of the @ sign and
username) lets users target a conversation to or reference a particular user, but these
tweets can be viewed by anyone through search.twitter.com, the public timeline, or the
sender’s Twitter page (Honeycutt and Herring, 2009 provide a detailed discussion of @
replies).
On Twitter, there is a disconnect between followers and followed. For instance,
musician John Mayer (johncmayer) is followed by 1,226,844 users, but follows only
47. While followers provide an indication of audience, this is imprecise. When an individual’s
account is public, anyone – with or without a Twitter account – can read their
tweets through the site, RSS, or third-party software. The vast majority of Twitter accounts
are public. Those who choose to protect their accounts can restrict their audience, but the
lists of followers on both public and protected accounts indicate only a potential audience,
since not everyone who follows a user reads all their tweets.
Tweets are also spread further when participants repost tweets through their accounts.
This practice, commonly referred to as ‘retweeting’, can introduce content to new audiences
(boyd et al., 2010). While the dominant norm is to use @username to cite the original
author or attribute the person who spread the message, retweeted messages are often
altered and may lose any reference to the original. Additionally, it is not uncommon for
people to forward tweets via email or by copying and pasting them into new communication
channels. Furthermore, various tools allow users to repost tweets to Facebook,
MySpace, and blogs.
Given the various ways people can consume and spread tweets, it is virtually impossible
for Twitter users to account for their potential audience, let alone actual readers. Yet,
this inability to know the exact audience does not mean that tweets are seen by infinite
numbers of people. As with blogs (Shirky, 2005), nearly all tweets are read by relatively
few people – but most Twitterers don’t know which few people. Without knowing the
audience, participants imagine it.
HowTwitter users imagine audiences
Methodology
To find out how Twitter users imagine their audience, we asked them directly. We posted
questions to our own followers (many of whom retweeted our question to their followers)
and sent @reply questions to a sample of users whose tweets appear in the public
timeline, every person in the 300 most-followed accounts (249 total), and a subset of
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 5
users with 1000–15,000 followers. While all the Twitter accounts we reference are public,
we anonymized all of our informants except the highly-followed users. Our questions
included: ‘Who do you imagine reading your tweets?’ and ‘Who do you tweet to?’ Later,
we asked: ‘What makes an individual seem “authentic” on Twitter? (Or what does
it mean to be authentic?)’ and ‘What won’t you tweet about? What subjects are inappropriate
for Twitter?’ Our goal in approaching different types of users was not to get a
representative sample of Twitter users but to elicit potentially diverse perspectives.
Given the issues mentioned above, we are unable to assess how many people saw our
tweets. We received 226 responses from 181 Twitter users through direct messages or @
replies to our queries. The responses we received revealed many different perspectives
on audience. While we can neither quantify audience management techniques nor
account for all potential perspectives, the responses we received provide valuable insight
into some core differences in conceptualizing audience. Future empirical work might
examine the prevalence of these strategies and their relationship to follower counts,
demographics, or genres of content.
To whom am I speaking?
Our informants conceptualized their audience on Twitter in diverse and varied ways. Most
responses we received focused on abstract categories of people (e.g. ‘friends’), but a few
indicated that their audience was articulated through the service itself. For example, an
informant defined his audience as ‘the overlap between my followers and my following’.
Respondents with relatively few followers typically spoke about friends, but some
focused on themselves. Respondents with large followings commonly described their
audience as ‘fans’. Of course, some have multiple audiences in mind:
I think I write to the people I follow and have twittered something recently. And I also tweet to
myself. Is that wrong?
I guess I’m tweeting to my friends, fans... and talking to myself.
Although some respondents emphasized that they speak to friends through Twitter, what
they mean varies. Part of the difficulty is that ‘friends’ is an overloaded term in social
media (boyd, 2008). One user described her friends as people she followed, while another
talked about writing to her ‘IRL friends’ to signal people she knew outside of Twitter.
Such users imagined their audience as people they already knew, conceptualizing Twitter
as a social space where they could communicate with pre-existing friends. This follows
the argument that Twitter’s strength is in its encouragement of ‘digital intimacy’
(Thompson, 2008). Many tweets are phatic in nature (Miller, 2008) and serve a social
function, reinforcing connections and maintaining social bonds (Crawford, 2009). One
respondent wrote, ‘I guess it’s like a live diary to all my friends. I post what they might
find interesting or know they will have an opinion on.’
When respondents referred to their audience as ‘me’, they also meant different things.
Some thought of Twitter as a diary or record of their lives. Others saw the service as a
space where they expressed opinions for themselves rather than others. Emphasizing
‘me’ may also be a self-conscious, public rejection of audience:
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
6		 new media & society XX(X)
Myself. It is MY Twitter account so, it’s mostly about me.
< Who do you tweet *to*?> No one & I love that. Or maybe myself five min. ago: I write the
tweets I want to read.
I don’t tweet to anybody; I just do it to do it.
Although these individuals may not direct tweets to others, they are not tweeting into a
void; they all have followers and follow others. Their emphasis on ‘me’ implies that for
them, Twitter is personal space where other people’s reactions do not matter. Similarly, a
few people saw crafting tweets for a particular audience as problematic:
As an individual (not org or corp) it’s worth it 2 me 2 lose followers 2 maintain the wholeness/
integrity of who/ what/how I tweet.
when I tweet, I tweet honestly, I tweet passionately. Pure expression of my heart.
What emerges here is not that these individuals lack an audience, but that they are
uncomfortable labeling interlocutors and witnesses as an ‘audience’. In bristling over the
notion of audience, they are likely rejecting a popularly discussed act of ‘personal branding’
as running counter to what they value: authenticity. In other words, consciously
speaking to an audience is perceived as inauthentic.
The strategic use of Twitter to maintain followers, or to create and market a ‘personal
brand’, is part of a larger social phenomenon of using social media instrumentally
for self-conscious commodification. In this process, strategically appealing to
followers becomes a carefully calculated way to market oneself as a commodity in
response to employment uncertainty (Hearn, 2008; Lair et al., 2005). As Dan Schawbel,
author of Me 2.0: Build a Powerful Brand to Achieve Career Success, writes on the
blog Mashable:
Today, Twitter has roughly 6 million users and is projected to grow to 18.1 million users by
2010. With all those people, the chances for networking are endless and connecting with new
people can lead to career opportunities, so it is essential that your personal brand exists on the
service … By leveraging the Twitter platform to build your brand you can showcase yourself to
a huge and growing audience. (2009)
Using Twitter to carefully construct a ‘meta-narrative and meta-image of self’ (Hearn,
2008) is part of what Jodi Dean (2002) calls the ‘ideology of publicity’, in which we
value whatever grabs the public’s attention. Publicity culture prizes social skills that
encourage performance (Sternberg, 1998); people are rewarded with jobs, dates, and
attention for displaying themselves in an easily-consumed public way using tropes of
consumer culture. In contrast, tweeting for oneself suggests a true-to-self authenticity,
untainted by expectations. Of course, authenticity is a social construct (Grazian, 2003)
and it is unlikely that anyone could tweet context-independently with no concern with
audience, given our understanding of audience influence on self-presentation (Blumer,
1962; Goffman, 1959). We are interested not in an absolute sense of authenticity, but in
what Twitter users consider ‘authentic’.
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 7
Other respondents suggested that audience conceptions were tweet-dependent. From
this perspective, Twitter is a medium, like telephony or email, that can be used for many
different purposes:
isn’t tweeting (like all things) situational? Try replacing the word ‘tweet’ in that Q [with]
‘email’.
Q->A: depends on the topic& intensity of connection: a)RT for public b)@/DM for followers/
friends, c) “thinking aloud for myself”;-)
This implies that users write different tweets to target different people (e.g. audiences).
This approach acknowledges multiplicity, but rather than creating entirely separate, discrete
audiences through the use of multiple identities or accounts, users address multiple
audiences through a single account, conscious of potential overlap among their audiences.
However, the difference between Twitter and email is that the latter is primarily a
directed technology with people pushing content to persons listed in the ‘To:’field, while
tweets are made available for interested individuals to pull on demand. The typical email
has an articulated audience, while the typical tweet does not. Email is also usually private,
while Twitter is primarily public. Notably, people avoid broaching many topics on
Twitter precisely for this reason.
Another approach some respondents took was to conceptualize their audience as an
‘ideal’ person:
I imagine my audience as a fellow nerd, who gets a say in my amusement, confusion, or
disappointment at whatever just happened.
In super-deep novel-writing mode tho I doubt it looks like it, so can’t talk, but basically I
imagine a few “ideal readers”.
I think of a room filled with friends when I tweet. I assume people like me that are reading my
tweets.
The ‘ideal’ reader is a well-known concept for writers, who often write ‘to’ an imaginary
interested party. In the tweets above, the ideal reader is conceptualized as someone similar
to the writer, who will presumably share their perspective and appreciate their work.
Ethnographies of television production have similarly found that producers ‘imagine
others like and unlike themselves, (re)constructing their own identities in the process of
constructing the imagined audience’ (Peterson, 2003: 161). The ideal audience is often
the mirror-image of the user.
Strategic audiences
In contrast to general notions of imagined audiences, Twitter users with numerous followers
expressed specific, pragmatic understandings of audience. A few mentioned reallife
friends, family and co-workers, but others with 100,000+ followers suggested that
they imagined their audience as a fan base or community with whom they could connect
or manage:
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
8		 new media & society XX(X)
Adventuregirl:	when I tweet I thnk of all my Sweet Tweets and sharing my life’s advenutes
w/them- and luv hearing there’s 2!
Padmasree:	 I mostly think of the community (more than friends)
This approach can be understood through the lens of ‘micro-celebrity’. Senft describes
micro-celebrity as a communicative technique that ‘involves people “amping up” their
popularity over the Web using techniques like video, blogs, and social networking sites’
(2008: 25). Micro-celebrity implies that all individuals have an audience that they can
strategically maintain through ongoing communication and interaction. Twitter is used
this way by many people – including marketers, technologists, and individuals seeking
wide attention – to establish a presence online. Likewise, by embracing social media to
engage directly with their audience, many traditional celebrities and public officials
embrace the techniques of micro-celebrity.
Users with large numbers of followers reasonably conceptualized and navigated their
audiences tactically. For example, Casey Wright (100,000 followers) answered that he
assumed ‘a broad audience with disparate tastes.’ When asked if he tailored tweets to
different parts of the audience, he answered, ‘Not really. I don’t think any tweet reaches
everyone but they all appeal to someone. I try to mix it up.’ The specificity of audience
understanding was striking among some with large followings:
Nansen:	I think of 1. political messaging 2. new friends 3. information 4. news
GuysReplies:	My tweets are news broadcasts ala NYTimes or StumbleUpon with Alltop
plugs.
Brooksbayne:	all of the above, but i have different “silos” for convos here. hashtags help with
all that.
[Authors]:	 What are your different silos?
Brooksbayne:	 politics, foodie, tech, social media, music biz, brands, and bacon!
Nansen, a conservative activist with 110,000 followers, has definite goals for Twitter:
maintain consistent political messaging, create relationships with new friends, provide
information, and spread news. Guy Kawasaki (154,000 followers) views his feed primarily
as a news broadcast. Brooks Bayne (95,000 followers) recognized the diversity of his
audience and used hashtags (keywords preceded by the # sign) to direct tweets to interested
followers. These users were personally invested in maintaining high follower numbers
and used several techniques to attract attention.
That is not to suggest there is an absolute divide in practice between the heavily followed
and those who are not. Instead, knowledge of the audience functions more as a
continuum. Several highly followed users did not mention trying to build and maintain
audience or feigned unawareness:
TychoBrahe:	Honestly, I have no idea who reads them. Hopefully a very small group of very
forgiving people!
Others acknowledged their visibility but didn’t see their actual audience as their intended
audience. Jason Goldman, a Twitter employee, said ‘I sometimes think about what my
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 9
girlfriend or coworkers or mom would think. I don’t think about “audience” really … If
I think about audience before tweeting (mostly not true) I think “would my friends dig
this.”’Movie blogger Harry Knowles said, ‘I imagine that my friends are reading mostly,
but with the knowledge there’s a greater voyeuristic society tuned in.’ Of course, just
because highly followed users claim that they’re not focused on audience does not mean
they are not. They may also be aware of the value of being perceived as authentic.
We also talked to users with fewer followers who had strategic plans for their audiences.
In choosing what to put forward, they often learn to present what is well received.
like my stream 1/3 humors, 1/3 informative, 1/3 genial and unfiltered, transparency is so chic.
try to tweet the same way.
Who do I imagine when I tweet? I think, I imagine myself getting my opinion out to hundreds
of ppl who might care:-)
U know I don’t know who reads them, but when I tweet sumthin contrvsial or interesting I find
a get a cple more followers.
The strategies of micro-celebrity are not only used by people with large numbers of followers.
Many users consciously use Twitter as a platform to obtain and maintain
attention, by targeting tweets towards their perceived audience’s interest and balancing
different topic areas.
A variety of imagined audiences stems from the diverse ways Twitter is used: as a
broadcast medium, marketing channel, diary, social platform, and news source. It is a
heavily-appropriated technology, which participants contextualize differently and use
with diverse networks. The networked audience is an abstract concept and varies among
Twitter users, in part because it is so difficult to ascertain who is actually there.
Navigating multiple audiences
The need to navigate
Like many social network sites, Twitter flattens multiple audiences into one – a phenomenon
known as ‘context collapse’. The requirement to present a verifiable, singular
identity makes it impossible to differ self-presentation strategies, creating tension as
diverse groups of people flock to social network sites (boyd, 2008). Privacy settings
alone do not address this; even with private accounts that only certain people can read,
participants must contend with groups of people they do not normally bring together,
such as acquaintances, friends, co-workers, and family. To navigate these tensions,
social network site users adopt a variety of tactics, such as using multiple accounts,
pseudonyms, and nicknames, and creating ‘fakesters’ to obscure their real identities
(Marwick, 2005). The large audiences for sites like Facebook or MySpace may create a
lowest-common denominator effect, as individuals only post things they believe their
broadest group of acquaintances will find non-offensive. Similarly, Twitter users negotiate
multiple, overlapping audiences by strategically concealing information, targeting
tweets to different audiences and attempting to portray both an authentic self and an
interesting personality.
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
10		 new media & society XX(X)
But why do users need to navigate multiplicity? In his seminal text The Presentation
of Self in Everyday Life (1959), Erving Goffman conceptualized identity as a continual
performance. Goffman analyzed people’s practices using a dramaturgical metaphor, suggesting
that we can understand individuals as actors who tailor self-presentation based
on context and audience. He proposed that in any given situation, people, like actors,
navigate ‘frontstage’ and ‘backstage’ areas; a workplace’s office space and meeting
rooms might be frontstage, while more candid talk takes place backstage at after-work
happy hour. Goffman’s work is often grouped with symbolic interactionism, a sociological
perspective which holds that meaning is constructed through language, interaction,
and interpretation (Blumer, 1962; Strauss, 1993). Symbolic interactionism claims that
identity and self are constituted through constant interactions with others – primarily,
talk. In other words, self-presentation is collaborative. Individuals work together to
uphold preferred self-images of themselves and their conversation partners, through
strategies like maintaining (or ‘saving’) face, collectively encouraging social norms, or
negotiating power differentials and disagreements.
Goffman maintained that this becomes a process of ‘impression management’, where
individuals habitually monitor how people respond to them when presenting themselves.
This process is self-conscious in situations of intense scrutiny, like first dates and job
interviews, but is habitual even in relaxed social situations. Self-monitoring leads people
to emphasize or de-emphasize certain things, responding to further feedback in a
dynamic, recursive process (Leary and Kowalski, 1990: 43). Thus, self-presentation
changes based on audience factors, such as friendship ties (Tice et al., 1995), status differentials
(Leary and Kowalski, 1990: 38), and racial differences (Fleming and Rudman,
1993). Even in difficult circumstances, people are skilled at using gesture, language, and
tone to manage impressions face-to-face (Banaji and Prentice, 1994).
Most of these studies draw from data and observations that involve people interacting
face-to-face, where it is fairly easy to gauge the gender, race, status, etc. of the
audience. Removing this ability creates tensions. Meyrowitz (1985) gives the example
of Black Power advocate Stokely Carmichael, who typically used different styles when
presenting to black and white audiences. Speaking on broadcast television, Carmichael
could not appear ‘authentic’ to both audiences and had to choose between a black or
white rhetorical style. He chose the former, engaging his black audience but alienating
white viewers. In today’s media-saturated landscape, politicians and celebrities use
‘polysemy’ or coded communication to simultaneously appeal to different, even oppositional
audiences (Albertson, 2006; Fiske, 1989). Madonna’s early image exemplifies
polysemy. She was interpreted differently by young women, who responded to her feminist
message, and young men, who responded to her sexy persona (Fiske, 1989).
Similarly, George W. Bush sprinkled coded references to hymns, Bible verses, and
Evangelical culture throughout his speeches to appeal to his base without alienating
others (Albertson, 2006).
Social media thus combines elements of broadcast media and face-to-face communication.
Like broadcast television, social media collapse diverse social contexts into one,
making it difficult for people to engage in the complex negotiations needed to vary identity
presentation, manage impressions, and save face. But unlike broadcast television,
social media users are not professional image-makers, and rather than giving a speech on
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 11
television, they are often corresponding with friends and family. By necessity, Twitter
users maintain impressions by balancing personal/public information, avoiding certain
topics, and maintaining authenticity.
Balancing expectations of authenticity
The imagined audience affects how people tweet. People with few followers, who use
the site for reasons other than self-promotion, generally see Twitter as a personal space
where spam, advertising, and marketing are unwelcome. Following the paradigm of
symbolic interactionism, identity on Twitter is constructed through conversations with
others. Tweets are formulated based partially on a social context constructed from the
tweets of people one follows. Participants must maintain equilibrium between a contextual
social norm of personal authenticity that encourages information-sharing and phatic
communication (the oft-cited ‘what I had for breakfast’ example) with the need to keep
information private, or at least concealed from certain audiences. The tension between
revealing and concealing usually errs on the side of concealing on Twitter, but even users
who do not post anything scandalous must formulate tweets and choose discussion topics
based on imagined audience judgment. This consciousness implies an ongoing frontstage
identity performance that balances the desire to maintain positive impressions with
the need to seem true or authentic to others.
This concept of ‘authenticity’ is a popular one. We refer to the ‘real me’ and authentic
experiences, artifacts, and people. However, there is no such thing as universal authenticity;
rather, the authentic is a localized, temporally situated social construct that varies
widely based on community. Grazian’s study of blues bars in Chicago defines authenticity
as conforming to an idealized representation of reality. The authentic is always manufactured,
and always constructed in ‘contradistinction to something else’ (2003: 13). In
other words, for something to be deemed authentic, something else must be inauthentic.
However, this dichotomy is false when we note that both the performance of authenticity
and inauthenticity are equally constructed by discourse and context (Cheng, 2004). What
we consider authentic constantly changes, and what symbols or signifiers mark a thing
as authentic or inauthentic differ contextually.
The fact that we constantly vary self-presentation based on audience reveals authenticity
as a construct: are we more or less authentic with our book club or gym partner?
Whether we are viewed as authentic depends on the definition imposed by the person
doing the judging. As in much social media, participants’ understanding of authenticity
on Twitter varies. For fashion bloggers, the ability to assemble an outfit that reflects a
personal aesthetic and knowledge of larger trends marks one as authentically stylish
and fashionable. For open-source geeks, on the other hand, ignorance of current trends
marks authenticity, emphasizing instead mental acuity and knowledge of software and
information law. Since authenticity is constituted by the audience, context collapse
problematizes the individual’s ability to shift between these selves and come off as
authentic or fake. We observed Twitter users using two techniques to navigate these
tensions: self-censorship and balance. People refrain altogether from discussing certain
topics on Twitter, while others balance strategically targeted tweets with personal
information.
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
12		 new media & society XX(X)
Self-censorship
Some people we spoke with suggested that they simply would not broach certain topics
on Twitter. Self-censorship can be a useful technique in the face of an imagined audience
that includes parents, employers, and significant others. Some respondents assumed anyone
could potentially read their tweets, making it impossible to discuss controversial or
personal topics:
anything i’d consider TMI (to spare my followers): family problems, relationship rants, etc.
This ain’t FB.
bathroom activity, romantic relationships, complaining about an employer
Won’t Tweet anything too personal, TMI about self/others, dead horse areas from subjects like
religion/politics/sports
Subjects mentioned included dating, sex, relationships, marital problems; Too Much
Information (TMI) about bodily functions and the like; criticism of one’s job; and controversial
or negative topics that might alienate followers. Without the ability to vary
information flow based on audience, participants could not risk a sensitive topic being
viewed by the wrong person.
Twitter can be viewed as a public space that should be carefully policed:
i’m very conscious that twitter is public. i wouldn’t tweet anything i didn’t want my mother/
employer/professor to see
I tweet about anything I would say in a lobby. Beyond that, each tweet is influenced by the
tweets around that unique moment.
Interestingly, one user views Facebook as an appropriate place for ‘family problems or
relationship rants’. While many respondents self-censored sex and dating information,
these topics are often discussed on blogs, Facebook, and LiveJournal. The information
was not too sensitive to ever be revealed, but it could not be revealed on Twitter, which
was seen by some as a ‘professional’ environment with potential professional costs:
I think it all depends on what the intended purpose for your twitter account. Professionals
should beware how they rep their cos i got threatened w/ lawsuit and loss of work bc of one of
my tweets. quite careful now in what i tweet. or try to be!
Work concerns influenced what people tweeted about as well as what they self-censored.
For instance, a freelancer said ‘I always keep my clients in mind. I want to convey intelligence
and professionalism, and diversity – I want to be seen as interested in a lot of
things.’ She could present herself appropriately against the social context of an imagined
audience of other professionals.
Participants maintain a public-facing persona to manage impressions with potential
readers. Context collapse creates an audience that is often imagined as its most sensitive
members: parents, partners, and bosses. This ‘nightmare reader’is the opposite of the ideal
reader, and may limit personal discourse on Twitter, since the lowest-common-denominator
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 13
philosophy of sharing limits users to topics that are safe for all possible readers. While
people do talk about controversial subjects on Twitter, our respondents show that some are
more likely to avoid personal topics that imply true intimacy and connection between followers.
Instead, they may frame Twitter as a place where the strictest standards apply.
Balance
For Twitter users trying to build audience, personal authenticity and audience expectations
must be balanced. To appeal to broad audiences, some popular Twitter users maintained
that they had to continually monitor and meet the expectations of their followers.
However, given context collapse, their followers had different preferences for revealing
personal information versus focusing on informative topics. Our respondents described
an ongoing loop of impression management as they altered this mix based on audience
feedback.
Soraya Darabi, the social media strategist for the New York Times, said, ‘I’m constantly
aware of my followers.’ She uses tools like Twittersheep (http://www.twittersheep.com),
developed by her company’s research and development staff, to track what
her 472,000 followers are interested in. Soraya knows her audience is interested in
‘media and marketing’, so she focuses on those topics. At the same time, she tries to
interject her own personality and passions – like music – to retain an authentic voice.
Soraya said:
Say you’re an author, a book aficionado. Most [of your followers] have tagged music as a
passion. You might want to throw them a bone about your favorite song. There are a lot of Venn
diagram overlaps in this community. It’s to your advantage to be as much as part of a community
as possible which means engaging with people’s interests.
Brandon Mendelson, an activist with more than 700,000 users, defined his audience as
‘my “tribe,” people who are interested in leading change in their organizations or day-today
life by using emerging technology and people interested in helping others through
social networks’. Brandon primarily focuses on subjects that appeal to this tribe, but he
agreed with Soraya that a mix of personal and professional is necessary for active engagement
on Twitter:
Occasionally I’ll get a person not happy about how often I tweet, which is quite a bit, and if I
tweet about something personal about my college plans. I always tell them I tweet about what
I want to tweet and that social networks are a personal platform. By not sharing personal
information I’m not building a strong relationship with my audience.
Other respondents’ comments echoed this view:
Authenticity Rule 1: Include personal w/ promotional. “Bags under my eyes from from staying
up ‘til 4 accepting friend requests. ”
to me, authenticity means being human; tweets include mix of ups, down, personal, professional.
v.little robot or corporate speak
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
14		 new media & society XX(X)
mix of work and social is interesting; agree it creates authenticity, but some find it annoying/
distracting
‘Corporate-speak’ or ‘work’ topics were seen as less authentic than personal, ‘human’
revelations. However, the intimacy of these revelations is limited. Note that both Soraya
and Brandon’s examples of personal topics are relatively innocuous. Their decisions to
reveal personal information are strategic, and often framed as a way to reinforce relationships
with followers. Soraya noted:
I don’t put romantic or deeply personal information on Twitter. I do say when I’m spending
time with X who they are, but typically that person is in new media and it may look good for
professional purposes to say I’m having lunch or dinner with X. It also serves as a call to action
to the newsroom that person X is in town.
These exemplify highly self-conscious identity presentations that assume a primarily
professional context. Revealing personal information is seen as a marker of authenticity,
but is strategically managed and limited. Similarly, several respondents mentioned that
concealing personal information was a way to avoid alienating followers, deliberately
avoiding topics that their followers might not agree with:
I try very hard not to Tweet hate speech, anything divisive, try to send messages that will bring
people together
politics and religion can be a little dangerous, ‘cus you never know which of your followers you
might offend.
Keeping balance is tricky; both Soraya and Brandon mentioned criticism from real-life
friends. Soraya has a few close friends whom she asks to critique her Twitter stream to
ensure she is striking a good balance: ‘I run blind checks – do I look too much like a
marketer? Am I tweeting too personally?’ The mix of personal and informative tweets
from users like Soraya and Brandon allowed them to maintain multiple audiences that
included both personal friends and professional contacts. Rather than appealing simultaneously
to multiple audiences, each mixed tweets with different target audiences to
maintain their broad appeal. This technique resembles the polysemic and coded communicative
strategies of image-management experts.
Micro-celebrity, conceptualized as a learned practice supported by the infrastructure of
social media, can create tension. Twitter’s directed friendship model replaces ‘friends’with
‘followers’ and prominently displays the number of followers on each person’s Twitter
page, creating a quantifiable metric for social status. The ability to strategically appeal to
broad audiences and retain the attention of others is publicly valued through third-party
services that rank people according to their number of followers. Micro-celebrity practices
like interacting directly with followers, appealing to multiple audiences, creating an affable
brand and sharing personal information are rewarded, and consequently encouraged, in
Twitter culture. The ability to attract and command attention becomes a status symbol.
At the same time, micro-celebrity practice can be seen as inauthentic. When asked to
describe ‘authenticity’ on Twitter, respondents placed it in direct opposition to strategic
self-promotion:
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 15
When I present the concept of authenticity I usually mean no marketing speak, don’t pretend
you know everything. Be yourself.
High honesty about what you’re here for. Don’t pretend to be my friend if you’re here for
promotion. (Promo is fine. Lying isn’t.)
Once I feel they’ve crossed the threshold of caring more about status/followcount or trendy
topics than their followers.
This view of micro-celebrity practice assumes an intrinsic conflict between self-promotion
and the ability to connect with others on a deeply personal or intimate level. Some view
strategic audience management as dishonest ‘corporate-speak’ or even ‘phony, shameless
promotion’. The encroachment of presumably profit- or status-driven ‘public’ techniques
into ‘private’ social spaces is met with stiff resistance from people used to
interactional norms that do not involve the commodification of social ties. We might ask
if ‘public’ space is becoming synonymous with ‘commercial’, and if alternative models
of publicity and attention can thrive within the networked audience environment.
From the broadcast to the network
Twitter is an example of a technology with a networked audience. Media audiences are
always imaginary, whether they exist in the writer’s mind or as the target demographics
for a sitcom. But while Fiske (1989) argues that the broadcast audience is a fiction meant
to serve the needs of media institutions, the writer’s audience services the writer. These
two models of audience help contextualize the networked audience and its impact on
online social behavior.
The writer’s audience
Writers have long grappled with conceptions of audience because writers, unlike speakers,
are separated from their audiences. In his essay ‘The writer’s audience is always a
fiction’, Walter Ong (1975) argued that writers imagine an audience appropriate to their
topic and form, and use textual cues to write that audience into being. Writers write for
and to this fictionalized audience, adapting to their imagined expectations. Lisa Ede and
Andrea Lunsford (1984) distinguished further between the audience addressed – the
actual readers of a piece of writing – and audience invoked, the audience constructed by
the writer. Published writers are often told to tailor books to particular demographics;
these ‘future readers’ are a fiction about the audience addressed.
The imagined audience of social media strongly resembles Ong’s fictionalized audience.
While Facebook or Twitter users don’t know exactly who comprises their audience
addressed, they have a mental picture of who they’re writing or speaking to – the
audience invoked. Much like writers, social media participants imagine an audience and
tailor their online writing to match.
The broadcast audience
Today, when we think about ‘audience’, we imagine people watching movies in a theater,
or at home watching television. This model of the audience has historically been viewed
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
16		 new media & society XX(X)
pejoratively in media studies – as an unidentifiable mass who passively consume
(Livingstone, 2005: 24). It is also intrinsically institutional; broadcasters assume an audience
that is anonymous, static, and geographically bounded (Drotner, 2005). One-to-many
communication implies a single broadcaster distributing content through a complex
entertainment structure that reaches audiences who cannot respond back.
The broadcast model has been complicated through studies of active audiences and
through the fragmentation and dispersal of mass audience. Active audience theory maintains
that the meaning of a media text is negotiated by the audience; rather than consuming
blindly, audiences use interpretive lenses and bring individual experiences to bear
when making meaning from media (e.g. Fiske, 1989; Radway, 1984). With the advent of
cable television in the 1970s, the VCR in the 1980s, the DVD player in the 1990s and the
ubiquity of home broadband and video games, audiences have splintered. The top-rated
shows on television are viewed by a fraction of the audience that watched 1960s network
television; niche networks and targeted media have proliferated (Turner, 2009).
As a result, the idea of the ‘audience’as a stable entity that congregates around a media
object has been displaced with the ‘interpretive community’, ‘fandom’, and ‘participatory
culture’, concepts that assume small, active, and highly engaged groups of people who
don’t just consume content but produce their own as well (Baym, 1999; Jenkins, 2005).
In decentralized communication networks like mobile phones and email, ‘audience’
describes how a communicative medium mediates a relationship between content producers
and receivers (Drotner, 2005: 196), requiring a more interpersonal and flexible
model. New media has changed the broadcast model of the audience, decentralizing
media production and distribution (Benkler, 2006). The network changes it further.
The networked audience
The networked audience combines elements of the writer’s audience and the broadcast
audience. It consists of real and potential viewers for digital content that exist within a
larger social graph. These viewers are connected not only to the user, but to each other,
creating an active, communicative network; connections between individuals differ in
strength and meaning (Haythornthwaite, 2002; Boase et al., 2006). Just as the broadcast
audience flattened separated demographic groups into a mass audience, the networked
audience combines a person’s social connections, revealing the fiction of discrete faceto-face
audiences. While the broadcast audience is a faceless mass, the networked audience
is unidentified but contains familiar faces; it is both potentially public and personal.
Like the broadcast audience, the networked audience includes random, unknown individuals,
but, unlike the broadcast audience, it has a presumption of personal authenticity
and connection. Social media participants are far more concerned with parents or
employers viewing their Twitter stream than a complete stranger.
In contrast to the imagined broadcast audience, which consumes institutionally-created
content with limited possibilities for feedback, the networked audience has a clear
way to communicate with the speaker through the network. This opportunity for communication
influences how speakers respond and what content they create in the future.
Audience members take turns creating and producing content, and in this ‘many-to-many’
model the network constantly centers on who is talking, responding, or replying. Social
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 17
media environments become a place where person-to-person conversations take place
around user-generated content amidst potentially large audiences.
The networked audience contains many different social relationships to be navigated,
so users acknowledge concurrent multiple audiences. Just as writers fictionalize
the audience within the text in their audience addressed, Twitter users speak directly to
their imagined audience. For instance, some ask their followers questions that assume
a particular collective knowledge. They target tweets to specific audience members,
and conceal or reveal information based on who they imagine to be listening. Some
construct a sophisticated model of who may be reading their tweets based on linguistic,
cultural, and identity markers in their Twitter stream. Managing the networked audience
requires monitoring and responding to feedback, watching what others are doing
on the network, and interpreting followers’ interests. The network is therefore a collaborator
in the identity and content presented by the speaker, and the imagined audience
becomes visible when it influences the information Twitter users choose to
broadcast.
Networked media brings the changes Meyrowitz described to interpersonal interactions.
In sites like Twitter and Facebook, social contexts we used to imagine as separate
co-exist as parts of the network. Individuals learn how to manage tensions between public
and private, insider and outsider, and frontstage and backstage performances. They
learn how practices of micro-celebrity can be used to maintain audience interest. But
Twitter makes some intrinsic conflicts visible. On the one hand, Twitter is seen as an
authentic space for personal interaction. On the other, social norms against ‘oversharing’
and privacy concerns mean that information deemed too personal may be removed from
potential interactions. Similarly, the desire to have ‘fans’ or a ‘personal brand’ conflicts
with the desire for pure self-expression and intimate connections with others. In combining
public-facing and interpersonal interaction, the networked audience creates new
opportunities for connection, as well as new tensions and conflicts.
References
Albertson B (2006) Dog whistle politics, coded communication, and religious appeals. American
Political Science Association and International Society of Political Psychology. Princeton, NJ:
Princeton University.
Banaji MR, Prentice DA (1994) The self in social contexts. Annual Review of Psychology 45:
297–332.
Baym N (1999) Tune In, Log On: Soaps, Fandom, and Online Community. New York: Sage.
Benkler Y (2006) The Wealth of Networks. New Haven, CT: Yale University Press.
Berkenkotter C (1981) Understanding a writer’s awareness of audience. College Composition and
Communication 32(4): 388–99.
Blumer H (1962) Society as symbolic interaction. In: Rose A (ed.) Human Behavior and Social
Process: An Interactionist Approach. New York: Houghton-Mifflin.
Boase J, Horrigan JB, Wellman B, and Rainie L (2006) The Strength of Internet Ties. Pew Internet
and American Life Project Report, January 25.
boyd d (2006a) A blogger’s blog: Exploring the definition of a medium. Reconstruction 6(4). URL
(consulted July 2009). Available at: http://reconstruction.eserver.org/064/boyd.shtml
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
18		 new media & society XX(X)
boyd d (2006b) Friends, friendsters, and Myspace top 8: Writing community into being on social
network sites. First Monday 11(2).
boyd d (2007) Why youth <3 social network sites: The role of networked publics in teenage social life.
In: Buckingham D (ed.) Youth Identity and Digital Media. Cambridge MA: MIT Press, 119–42.
boyd d (2008)Taken out of context:American teen sociality in networked publics. PhD Dissertation,
University of California Berkeley.
boyd d, Golder S, and Lotan G (2010) Tweet, tweet, retweet: Conversational aspects of retweeting
on Twitter. HICSS-43. IEEE: Kauai HI, January 6.
Cheng VJ (2004) Inauthentic: The Anxiety Over Culture and Identity. New Jersey: Rutgers
University Press.
Crawford K (2009) These foolish things: On intimacy and insignificance in mobile media. In: Goggin
G, Hjorth L Mobile Technologies: From Telecommunications to Media. New York: Routledge.
Dean J (2002) Publicity’s Secret: How Technoculture Capitalizes on Democracy. Ithaca, NY:
Cornell University Press.
Drotner K (2005) Media on the move: Personalized media and the transformation of publicness.
In: Livingstone S (ed.) Audiences and Publics: When Cultural Engagement Matters for the
Public Sphere. Portland, OR: Intellect, 187–212.
Ede L, LunsfordA(1984)Audience addressed/audience invoked: The role of audience in composition
theory and pedagogy. College Composition and Communication 35(2): 155–71.
Ellison N, Heino R, and Gibbs J (2006) Managing impressions online: Self-presentation processes
in the online dating environment. Journal of Computer-mediated Communication 11(2). URL
(consulted June 2009). Available at: http://jcmc.indiana.edu/vol11/issue2/ellison.html
Fiske (1989) Moments of television: Neither the text nor the audience. In: Seiter E et al. (eds)
Remote Control: Television Audiences and Cultural Power. London: Routledge
Garfinkel H (1967) Studies in Ethnomethodology. Malden, MA: Polity Press/Blackwell Publishing.
Goffman E (1959) The Presentation of Self in Everyday Life. Garden City, NY: Doubleday.
Grazian D (2003) Blue Chicago. Chicago, IL: University of Chicago Press.
Haythornthwaite C (2002) Strong, weak and latent ties and the impact of new media. The
Information Society 18(5): 385–401.
Hearn A (2008) ‘Meat, mask, burden’: Probing the contours of the branded ‘self.’ Journal of
Consumer Culture 8(2): 197–217.
Hodkinson P, Lincoln S (2008) Online journals as virtual bedrooms? Young people, identity and
personal space. Young 16(1): 27–46.
Honeycutt C, Herring SC (2009) Beyond microblogging: Conversation and collaboration via
Twitter. Proceedings of the Forty-second Hawaii International Conference on System Sciences
(HICSS-42). Los Alamitos, CA: IEEE Press.
Jenkins H (2005) Convergence Culture. New York: NYU Press.
Lair D, Sullivan K, and Cheney G (2005) Marketization and the recasting of the professional self.
Management Communication Quarterly 18(3): 307–43.
Leary MR, Kowalski, M (1990) Impression management. A literature review and two-component
model. Psychological Bulletin 107(1): 34–47.
Liu H (2007) Social network profiles as taste performances. Journal of Computer-mediated
Communication 13(1) article 13. URL (consulted June 2009). Available at : http://jcmc.indi-
ana.edu/vol13/issue1/liu.html
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
Marwick and boyd	 19
Livingstone S (2005) On the relation between audiences and publics. In: Livingstone S (ed.)
Audiences and Publics: When Cultural Engagement Matters for the Public Sphere. Portland,
OR: Intellect, 17–42.
Marwick A (2005) I’m More Than Just a Friendster Profile: Identity, Authenticity and Power in
Social Networking Services. Chicago, IL: Association for Internet Researchers.
Meyrowitz J (1985) No Sense of Place: The Impact of Electronic Media on Social Behavior. New
York: Oxford University Press.
Miller V (2008) New media networking and phatic culture. Convergence 14(4): 387–400.
The Nielsen Company (2009) Nielsen’s Social Media QuickTake: May 2009. Nielsen.com.
URL (consulted June 2009). Available at: http://blog.nielsen.com/nielsenwire/wp-content/
uploads/2009/06/nielsen_pr_090619.pdf
Ong W (1975) The writer’s audience is always a fiction. PMLA 90(1): 9–21.
Papacharissi Z (2002) The presentation of self in virtual life: Characteristics of personal home
pages. Journalism and Mass Communication Quarterly 79(3): 643–60.
Peterson MA (2003) Anthropology and Mass Communication: Media and Myth in the New
Millennium. Canada: Berghahn Books.
Radway J (1984) Reading the Romance. North Carolina: University of North Carolina Press.
Reed A (2005) ‘My blog is me’: Texts and persons in UK online journal culture (an anthropology).
Ethnos 70(2): June.
Robinson L (2007) The cyberself: The self-ing project goes online, symbolic interaction in the
digital age. New Media & Society 9(1): 93–110.
Schau HJ, Gilly MC (2003) We are what we post? Self-presentation in personal web space. Journal
of Consumer Research 30(3): 385–404.
Schawbel D (2009) HOW TO: Build Your Personal Brand on Twitter. Mashable.com, 20 May.
URL (consulted July 2009). Available at: http://mashable.com/2009/05/20/twitter-personal-
brand/
Scheidt LA (2006) Adolescent diary weblogs and the unseen audience. In: Buckingham D, Willett R
Digital Generations: Children, Young People and New Media. New Jersey: Lawrence Erlbaum
Associates.
Senft T (2008) Camgirls: Celebrity and Community in the Age of Social Networks. New York:
Peter Lang.
Shirky C (2005) Power-laws, weblog and inequality. In: Lebkowski J, Ratcliffe M (eds) Extreme
Democracy. URL (consulted July 2009). Available at: http://extremedemocracy.com/chapters/
Chapter%20Three-Shirky.pdf
Sternberg E (1998) Phantasmagoric labor: The new economies of self-presentation. Futures 30(1):
3–28.
Strauss A (1993) Continual Permutations of Action. New York: Aldine de Gruyter.
Thompson C (2008) Brave New World of Digital Intimacy. The New York Times, 7 September.
URL (consulted February 2009). Available at: http://www.nytimes.com/2008/09/07/
magazine/07awareness-t.html
Tice DM, Butler JL, Muraven MB, and Stillwell AM (1995) When modesty prevails: Differential
favorability of self-presentation to friends and strangers. Journal of Personality and Social
Psychology 69:1120–38.
Turner G (2009) Television Studies After TV. New York: Routledge.
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from
20		 new media & society XX(X)
Alice E. Marwick is a PhD Candidate at New York University’s Department of Media,
Culture, and Communication and a researcher at the Berkman Center for Internet and
Society on the Youth and Media Policy Project. She has a BA from Wellesley College
and an MA from the University of Washington.
danah boyd is a researcher at Microsoft Research and a Fellow at Harvard University’s
Berkman Center for Internet and Society. Her research focuses on youth engagement with
social media. She is a co-author of Hanging Out, Messing Around, and Geeking Out: Kids
Living and Learning with New Media (MIT Press, 2009). [email: dmb@microsoft.com]
at Bobst Library, New York University on July 8, 2010nms.sagepub.comDownloaded from