Sam Hinton and Larissa Hjorth - Understanding Social Media-SAGE Publications (2013)
Sam Hinton and Larissa Hjorth - Understanding Social Media-SAGE Publications (2013)
Understanding Social Media provides a critical and timely conceptual toolbox for
navigating the evolution and practices of social media. Taking an interdisciplinary
and intercultural approach, this book provides a clear and concise explanation of
the key concepts but also goes beyond specific brands, sites and practices to show
social media
readers how to place social media more critically within the changing media and
cultural landscape.
This book is essential reading for students of media studies and cultural studies.
Sam Hinton is a senior lecturer in media and the head of the discipline of Media
Arts and Graphic Design at the University of Canberra.
ISBN 978-1-4462-0120-6
ISBN 978-1-4462-0121-3
Acknowledgements vi
8 Conclusion 136
Glossary 140
References 145
Index 157
F irstly, the authors would like to thank series editors Jen Webb and
Tony Shirato and Sage’s Mila Steele for their help in developing this
publication.
Secondly, we would like to thank the wonderful community of researchers
who are exploring social media for all your insights, inspiration and thoughtful
provocations. We would also like to acknowledge the support of the University
of Canberra and RMIT University, Melbourne.
Sam would like to thank his family, friends and colleagues in media, arts and
production, who gave him both the time and space to bring this book together,
and to his co-author, Larissa Hjorth, whose energy and intelligence pushed this
book through from draft to conclusion. Sam would like to dedicate this book
to his family: Nicole, Catherine and Sarah – you three continue to humble me.
Larissa would like to thank the Australia Research Council for a discovery
grant (DP0986998) that allowed her the time to conduct empirical research
for this book. She would also like to thank family and friends and dedicates
this book to her son, Jesper, and brother, Greg.
I t’s 9 p.m. and in the busy streets of Shanghai, a Chinese teenager takes a
picture via her iPhone and loads it up to her social network site (SNS),
Renren. Meanwhile, a high school boy in Manila logs on to his Facebook
page via his personal PC to say hello to his aunty who has just woken up in
LA. Elsewhere, two university friends in Seoul stay up late to play a social
media game in their local PC bang (PC internet room), while another univer-
sity student in Beijing logs on to FarmVille to play with their parents in a
far away village. In London, two high school students sit together editing
and commenting on their Facebook pictures in a local café. At the same
time, two old friends meet up face-to-face in New York for breakfast,
thanks to Facebook Places.
These fleeting moments of connection take place through flashes of light,
across deep-sea cables and microwave pulses that bounce invisibly between
orbiting satellites. If we could see these connections plotted around the globe,
the world would be illuminated like an exquisite decoration, shimmering
with the mediated social interactions of many of its nearly seven billion
inhabitants. Today, social media in its many forms accounts for a great deal
of this mediated activity. Social media bleeds across platforms (desktop com-
puters, mobile phones, tablets and on modern network-capable televisions),
across social and media contexts, and creates various forms of presence.
As smartphones continue to move into mainstream everyday life in many
urban settings globally, the demographics of social media are also changing.
Once upon a time, SNSs were just for the young (boyd and Ellison 2007), but
today young and old can be seen using SNSs in everyday life.
As SNSs evolve, the term ‘social media’ is also developing to encompass the
growing and often unwieldy sphere of contemporary online media practice.
For Melissa Gregg, social media accompanies a movement towards ‘presence
bleed’, ‘where boundaries between personal and professional identities no
longer apply’ (2011: 2). Underlying this concept is an array of questions about
the changing nature of what is public and what is private, and where work ends
and life begins, as social media infiltrates every facet of everyday life.
For danah boyd, SNSs are a genre of what she calls ‘networked publics’
(2011), that is, public groupings that are structured by the logic and reality
of computer networks. For Mark Andrejevic, the ‘networked sociabilities’ of
SNSs are ordered by a ‘separation of the user from the means of socialising,
thus permitting “storable and sortable” collections of social data’ (Andrejevic
2011: 311). In other words, the social activities of the user can be easily
adapted by social media companies into ‘user profiles’ that are then sold to
advertisers. This phenomenon has led media activist and theorist Geert
Lovink to argue for alternative models of social media beyond the stronghold
of mainstream companies like Facebook and Google (2012).
Understanding Social Media attempts to engage with some of these
complex debates about the definitions of social media. We reflect upon the
differences between SNSs and social media and how the rise in devices
such as smartphones and locative media services such as Facebook Places,
Google Maps, and Foursquare are changing the fabric of social media. We
acknowledge that social media is currently transforming definitions of both
‘social’ and ‘media’.
Social media impacts on the way in which we think, experience and
practise ‘online media’. It is no longer merely a form of teen socialising – it
has become an integral part of everyday life. In turn, this influences how
we reflect and engage with friends, family, colleagues and politics. Social
media further amplifies the changes in the media landscape and as it does,
it provides new avenues for dissemination and engagement. For some crit-
ics, social media is part of the rise of participatory culture which empowers
users (Jenkins 2006) to produce their own content, to become ‘produsers’
(Bruns 2005). For others, social media is part of broader ‘structural
affordances of a capitalist economy’ (Andrejevic 2011: 312) in which users’
free labour is exploited for the benefit of corporations (Kücklich 2005;
Andrejevic 2011; Lovink 2012). For still others, the relationship between
production and consumption has now altered and should not be understood
in the same way that sociologists understood production in an industrial
context (Banks and Humphreys 2008). The widely varying ideas, criticisms
and exhortations about social media reflect the complex social processes
that it engages with.
In order to address these issues, and provide a framework for understanding
the many different concepts and theories that inform the debates surrounding
social media, we deal with four recurrent themes throughout this book. These
themes are: empowerment/control, online/offline, the role of the local/
cultural (especially in non-Anglophonic contexts) and the ‘intimacy turn’.
The first two themes present as dichotomies. However, rather than seeing
them as an ‘either/or’ relationship, we suggest that the reality is somewhere
in the middle, and their apparent oppositions provide us with two ways in
which we can critically examine social media. So, social media is neither
entirely empowering nor entirely controlling. In fact, it is often both.
Likewise, it is important to acknowledge that social media contains offline
modes of engagement: it is never entirely just an online phenomenon. Given
that social intimacy has always been mediated (Hjorth 2005), the online/
offline tension suggests new entanglements for social interaction, notions of
presence and its impact upon public and private spaces. The relationships
that people have online are always shaping, and shaped by, the offline. This
may be as obvious as the recognition that our online friends are frequently
people we know from home or work. Or it may be as subtle as understand-
ing that how we behave online, who we choose to make friends with online,
and how we use our online time are all influenced by the reality of our
offline lives.
The final two themes – non-Anglophonic contexts and the intimacy
turn – respectively focus our attention upon the global and local nature of
social media. In this book, we want to emphasise the truly global nature
of social media by mixing Anglophonic and non-Anglophonic perspectives.
We seek to bring balance and emphasise the socio-cultural nature of social
media in light of the dominance of Anglophonic approaches (Goggin and
McLelland 2009). This not only serves to highlight how different cultural
perspectives change the meaning of certain concepts (like privacy, for example),
but also serves as a foil to help us better understand how our own cultural
practices involve assumptions and tacit knowledge. The local continues to
play a key role in the uneven global evolution of social media across a variety
of platforms, modes of presence, contexts and media.
Even intimacy operates upon micro (individual), meso (social) and macro
(cultural) levels as a glue for social relationships. When we refer to the ‘inti-
macy turn’, we are drawing attention to the way in which the concept of
intimacy can be used to understand some of the erosions between public and
private spaces, and between work and leisure in contemporary societies
(Berlant 1998). To put it another way, social media affords certain kinds of
social performance that involve making intimacy more public. For example,
when a Facebook user takes and uploads self-portraits, they may well only
intend these to be viewed by an audience of close friends even though they
have their privacy settings set to public. However, with Facebook owning the
copyright of the pictures, which are available to be seen by anyone with a
Facebook account, how the pictures would be recontextualised and further
consumed is complex and little understood. Alternatively, a user might
have their personal photo albums set to be viewable only by family and
friends, and so the viewing activity becomes a semi-public performance.
Once the internet changed the world; now the world is changing the internet.
Its mainstreaming is well and truly over, and the forgettable Web 2.0 saga has
run its course. Now that society has overruled their freewheeling ethic, the
notion of the internet as an exceptional, unregulated sphere evaporates. The
moment of decision bears upon us: which side are you on? (Lovink 2012: 1)
As Lovink’s quote above suggests, the internet is coming of age. With more
than a decade of use in many parts of the world, the internet is embedded
in the everyday. But along with its uneven development across the globe
come issues concerning power and locality. There are many internets across
the world, accessed and used in a variety of ways. In this chapter we engage
with one of the major themes that underlie the emergence of social media:
the tension between control and freedom and between exploitation and
empowerment. We look at how business interests have attempted to com-
mercialise the internet and how, over the course of a decade, they have
shifted their strategies in order to align with how people are actually using
the internet. This transition to user-focused business models is represented
in the term ‘Web 2.0’.
To begin this chapter, we indulge in a brief discussion of the web, high-
lighting the key technical features and its relationship with the development
of the internet. When we think of the internet we must acknowledge that it
encompasses multiple definitions and experiences. Rather than ‘one’ internet,
there are multiple, intersecting imaginings and understandings of the
internet that are informed by the user’s background and experiences.
The internet is not a parallel universe (as was suggested by early writings on
cyberspace) but rather has always been a part of everyday life. Today, with
the popularity of ‘always on’ mobile media allowing users to perpetually surf
across social and locative media apps, the internet has become an embedded
part of mundane social life.
For Lovink, the rise of Web 2.0 heralded a new definition of the term
‘social’ that no longer evokes the possibility for democratic empowerment
and change as it did in the nineteenth or twentieth centuries. Instead, the
‘social’ has been tamed (2012: 6). As Lovink notes, ‘[p]latforms come and go,
but the trend is clear: the networks without cause are time eaters, and we’re
only being sucked deeper into the social cave without knowing what to look
for’ (2012: 6). Representing internet culture as being ‘caught between self-
referentiality and institutional arrangements’ (2012: 2), Lovink observes that:
today. Prior to the advancement of the web, the internet was made up of a
series of computers, connected to each other through numerous diverse
methods but sharing a common basic data transfer protocol called TCP/IP.
Each of the computers connected to each other via the internet were able
to share data with each other, and as the internet developed, the ability to
locate (or ‘discover’) resources became an increasingly large problem. The
web provided an interface that allowed people to discover and access inter-
net resources quickly and easily.
The TCP/IP protocol describes how data on one computer can be trans-
ferred to another computer across a vast network made up of anything
from physical copper wires to wireless satellite connections. TCP/IP does
not decipher the data, it just deals with lumps of zeros and ones – called
packets – and makes sure each packet is delivered to the correct computer
in the right order without any loss of data. On top of this protocol sit ‘appli-
cation protocols’, which are concerned with making sense of data. Email is
one such application protocol. The email protocols describe how an email
can be turned into small chunks of information, sent over the TCP/IP pro-
tocol to another computer and then reassembled as an email at the other
end. The web constitutes just one of these application protocols – a protocol
called HTTP, or hypertext transfer protocol – but because the web is the
main way in which we interact with the internet on a daily basis, people
often conflate the terms ‘internet’ and ‘web’. In day-to-day circumstances
that is probably fine, but when we think about these things critically it is
important to be precise.
For most people, the primary experience with the internet is through the
web browser – of which there are many different brands such as Firefox,
Internet Explorer, Safari and Google Chrome. While each of these browsers
has slightly different features, they all use the same HTTP application pro-
tocol for sharing information across the internet, and they are all designed
to assemble text, images, video and interactive components together into
one coherent interface. It is this interface – which we call the web page –
that constitutes our experience of the web. Because the browser is so good
at assembling different kinds of media and presenting them in a single uni-
fied interface, the web browser becomes most people’s entire interface with
the internet.
The other important piece of software on the internet that makes the web
possible is the web server. The server is a computer program that is con-
stantly running on a computer that is always connected the internet. The
software waits until someone contacts it, and then responds to this contact
by sending data – mostly text and images, but often also things like video or
perhaps interactive content like games. When you type in a website address
the web can also allow computers on the internet to share images, audio,
video and other forms of media. When these are mixed together, the result
is a multimedia interface – a single view that can incorporate elements of
many different media. Almost every web page you access contains a variety
of media elements.
The web quickly became the quintessential ‘killer app’ – a phrase that
denotes a software application that is so successful that it sells the plat-
form that it runs on. Early killer apps sold computer platforms, and the
web, if seen in a similar light, has sold the internet, at least in a meta-
phorical sense. The growth in internet use is closely correlated to the dev
elopment of the web and, although we should always be careful not to
confuse correlation with causation, it seems quite clear here that the web’s
ability to bring together multiple digital media sources through a single
easy-to-use interface was a significant innovation in the development of
the internet.
WEB 1.0
Web 1.0 is a phantom term, constructed after the event. People did not speak
of Web 1.0 until after the concept of Web 2.0 had been defined. In this
respect Web 1.0 is an arbitrary historicisation, not much different from the
way historians used to break history into the time before the birth of Jesus
Christ (BC), and after (AD). Just as nobody walked around Europe 2,500
years ago talking about how nice it was to be living in 500BC, nobody talked
about Web 1.0 until the term Web 2.0 emerged. The tag ‘2.0’ evokes the idea
of software versioning and its associated marketing,1 and so suggests that
Web 1.0 was less evolved, less sophisticated and less refined.
In the introduction to this chapter we suggested that the web was a
significant factor in the rapid development of the internet. In a field that
is quite liberally scattered with hyperbole and invective, the one fact that is
probably fair to say is that the growth of the internet has been nothing short
of remarkable. With so many people going online so quickly, it was only a
matter of time before companies started to realise the potential of this
medium. Where there are people, there are markets, and the internet, which
is by design a two-way digital medium, had the potential to offer up some-
thing that mass media broadcast technologies could not: a source of highly
detailed information about audiences.
Inevitably, initial attempts by large companies to control this digital envi-
ronment were based upon their experience with traditional media. Television
used content to aggregate audiences with similar interests: a science-fiction
show attracted a certain demographic, soap operas another, and so on.
Consequently, an apparently reasonable strategy in the online space was to
accumulate attention, in a similar way, and then to sell products and services
to a captive audience. Attracting attention in the online environment proved to
be relatively easy, but turning that attention into money was problematic.
Commercialisation of the web has not been the straightforward process
that many early internet entrepreneurs felt it might be. Roger Clarke (1999)
has suggested that one of the key problems for those looking to make money
from the internet in the early days was a simple unwillingness of internet
users to pay for online services. Straightforward subscription models were
only marginally successful and many initial attempts to set up internet pay
sites were undermined by other sites that gave away information in a rush
to build large user-bases.
Web 1.0 emerged out of a desire to make money from internet users, or to
‘monetise’ them (to use the rather ugly word that is widespread in marketing
and business circles). It also built on pre-internet dreams which involved
computer services of some kind that would be delivered to the home: at this
time, the idea of the domestic computer service has been an ongoing theme
in the media, information technology and telecommunications sectors for
almost four decades (Haddon 1999). Attempts by businesses to establish
information networks in people’s homes before the internet emerged as a
viable domestic networking technology – such as Videotex – were largely
unsuccessful and generally very expensive mistakes for the companies that
backed them.
Nevertheless, undeterred by past experience, and spurred on by the phe-
nomenal growth rate of the internet subscription base, entrepreneurs courted
the internet as a commercial domestic network as soon as user numbers began
growing in the mid-1990s. Despite initial enthusiasm, the commercialisation
of internet users (as distinct from commercialisation of internet access, which
was very successful for major ISPs and telecommunications companies)
proved to be highly elusive.
Wired magazine claims that it was the first organisation to launch banner
ads on its website Hotwired in October 1994 (Clarke 1999). Other companies
followed this trend, and there soon began a rapid appearance of advertising
banners on search engines and other websites. This was a substantial
imposition on the slow, low-bandwidth connections of the time. Although
this kind of advertising has since proven one of the more effective ways for
websites to make money, these initial attempts were less than successful.
As Clarke pointed out:
This initial failure was blamed on technological and economic factors. The
argument was that the web was simply not (yet) capable of maintaining
internet commerce, the network was not technically sophisticated enough
and slowed things down too much, and as soon as sites and services became
popular they crumpled under the weight of user attention. There was a
certain amount of truth to this, although there was an underlying lack
of interest in actually attempting to understand how people were using
the internet, and how this affected business models that were still treating
internet users like TV audiences.
As an extension of success enjoyed by some of the non-internet online
services that had emerged in the 1980s (Compuserve and AOL, for exam-
ple), some internet entrepreneurs attempted to create internet services that
provided the same kind of isolated walled-off space, but on the internet
rather than on a proprietary online system. Microsoft even attempted to
establish the Microsoft Network (MSN) as its own proprietary network
quite apart from (and perhaps imagined to compete with) the internet. The
idea was to get users to sign up for an online service that integrated directly
into Microsoft’s desktop environment – Microsoft’s plan here was to use
their market dominance in operating systems to establish a new online
service. Generally speaking, such ‘gated’ areas of the internet also failed to
appeal to users, and were not successful in generating revenue. Why volun-
tarily stay behind a walled-off zone that you have to pay subscription fees
to live within when there’s a free garden of earthly delights just a modem’s
dial away?
The failure of MSN was a particularly significant experience for Microsoft
who, until this time, had gone from strength to strength with almost every
new product or idea achieving immense commercial success. Within months
of launching MSN, Microsoft relegated the service to a content-aggregation
node of the internet, making a hasty about-face. A previously unknown soft-
ware company – Netscape – became a multi-million dollar business virtually
overnight by giving its web browser software to users and selling web server
software to companies.2 One of the young engineers who helped establish
the company found himself on the cover of Time magazine, declared one of
a new breed of ‘instantaires’ (Collins, 1996).
While academics had been developing more sophisticated understandings
of internet users through the 1990s – as we will discuss in the next section –
John Hagel’s 1997 Net Gain attempted to explain how online communities
could be considered an important commercial resource. According to Hagel,
the aggregation of internet communities around certain areas of interest
provided an opportunity for so-called ‘info-mediaries’ to deliver audiences
to advertisers and marketers. This concept was engaged with literally by
some businesses, who then constructed web portals – sites that aim to
aggregate users around centralised content – in an attempt to concentrate
user attention.
Because of their importance for internet resource discovery, search engines
formed some of the earliest portals. America Online (AOL), and certain
other companies who provided dial-in access, sought to aggregate users by
channelling them into their sites as the user connected to the internet. But,
aggregation of users was only part of the problem. The more difficult goal
was to make money from those users in some way, either directly (through,
for example, subscription fees) or indirectly (for example, advertising).
Many of these so-called portal sites were to become emblematic of the folly
of initial attempts to commercialise the internet. The problem with portals
WEB 2.0
The Web has already become an almost iconic cultural reference – ubiquitous
and familiar. We think we know what it is by now. The Web we know now,
which loads into a window on our computer screens in essentially static
screenfuls, is an embryo of the Web as we will know it in not so many years …
The first glimmerings of Web 2.0 are now beginning to appear, and we can start
to see just how that embryo might develop … The Web will be understood, not
as screenfuls of text and graphics but as a transport mechanism, the ether
through which interactivity happens. It will still appear on your computer
screen, transformed by the video and other dynamic media made possible by
the speedy connection technologies now coming down the pike. It will also
appear, in different guises, on your TV set (interactive content woven seamlessly
into programming and commercials), your car dashboard (maps, yellow pages,
and other traveler info), your cell phone (news, stock quotes, flight info), hand-
held game machines (linking players with competitors over the Net), maybe
even your microwave oven (automatically finding cooking times for the latest
products). (DiNucci, 1999)
In DiNucci’s somewhat science fictional future gazing, she is linking Web 2.0
to the rise of ubiquitous computing. The spectres of Mark Weiser’s (1991)
prescient words about the importance of context-awareness and embed-
dedness within the constitution of ubiquitous technologies can be felt in her
vision. Weiser imagined a time in the future where computer technology
would vanish into the background as we moved beyond big, clunky
machines and into a world where ever-present (ubiquitous) but essentially
invisible computers became as commonplace as the written word. In his
vision, Weiser imagined a time when computer technologies were always
available and able to provide extra information about every conceivable
aspect of life – not too far different from the experience afforded by
today’s mobile technologies. DiNucci’s focus is, however, from a designer’s
point of view rather than, as the term later gets recruited, from a business
perspective.
some kind of turning point for the web, such that a call to action such as
“Web 2.0” might make sense?’
O’Reilly’s concept of Web 2.0 indicates that business in this new internet
age is tightly related to active, engaged internet users noting that ‘(n)etwork
effects from user contributions are the key to market dominance in the Web 2.0
era’ (O’Reilly 2005; original emphasis). Hints of this shift towards recognising
the importance of internet users – cast as customers – had begun appearing as
early as 1997. As previously mentioned, Hagel’s Net Gain first extolled the
virtues of engaging with users, and was followed in 1999 by a website (later
published as a book in 2000) called The Cluetrain Manifesto, which con-
sists of 95 short essays or ‘theses’ that argue that the real power of the web
is in connecting businesses with their clients. One of the key ideas from this
book is that:
Under the ‘read’ model of Web 1.0, if you wanted to provide information
online you would need to create your own website. This generally required
a lot of technical knowledge, and may have required you to run your own
web server, which was a fairly daunting technical task. The Web 2.0 model
sees the computer take over the task of managing technical details of for-
matting and presentation, allowing the user to focus on the production of
content. Blogs, for example, provide a way for users to publish informa-
tion online with few more technical skills than are required to use a web
browser and type at a keyboard. Thus, Web 2.0 makes creating content
vastly less complicated, and this, in turn, leads to much more content being
put online as the technical barriers to creation are removed. Once content
could readily be created by just about any user, the technological prerequisites
were met for the emergence of social network sites (SNSs; see Chapter 3
for details).
For Ethan Zuckerman, in his ETech paper entitled the Cute Cat Theory
of Digital Activism, ‘Web 1.0 was invented to allow physicists to share
research papers and Web 2.0 was created to allow people to share pictures
of cute cats’ (2008). While this somewhat facetious comment could be read
in light of the shift towards experiencing and conceptualising the internet
as a social space, it also highlights that, behind the oft-banal activities of
users, new forms of affective sharing and communities are emerging. As
Zuckerman observes, while Web 2.0 ‘was designed for mundane uses, it can
be extremely powerful in the hands of digital activists, especially those in
environments where free speech is limited’ (2008). But for others such as
Lovink, this ‘cute cat’ phenomenon is part of a simplification of the ‘social’
in which all journalism becomes a series of secondary references and divi-
sions (2012: 4).
However, we could say that the personalisation of technology had been
occurring long before Web 2.0, mobile media and the ‘cute cat’ phenomenon.
Indeed, countries such as Japan have excelled globally in their ability to
spearhead the ‘personal technologies’ revolution from the Sony Walkman
onwards. Mizuko Ito (2005), for example, argues that it is the notion of
the ‘personal’ – along with pedestrian and portable – that have characterised
Japanese technologies for decades (Fujimoto 2005; Okada 2005). Part of the
success has been their deployment of high-level customisation, particularly
apparent in what anthropologist Brian McVeigh has called ‘techno-cute’; that
is, the usage of the cute to make ‘warm’ and ‘friendly’ the coldness of new
technologies (2000).
Given that the kinds of behaviours attributed to Web 2.0 seem to have
their roots in older and well-established cultural uses of computer and
media technologies, one could argue that some formulations of Web 2.0
But as with many areas of Web 2.0, where the ‘2.0-ness’ is not something new,
but rather a fuller realization of the true potential of the web platform, this
phrase gives us a key insight into how to design applications and services for
the new platform. (O’Reilly 2005)
For Australian new media theorists Anna Munster and Andrew Murphie
(2009), the confusion around rhetoric to do with Web 2.0, and the type of
attendant agency it affords, is due to the fact that its semantics have been
misunderstood. They argue that O’Reilly got it wrong:
Web 2.0 is not an ‘is’, or not only this. Web 2.0 is also a verb or, as they taught
us in primary school, it’s a doing word. Here’s a list of some Web 2.0 things to
do: apping, blogging, mapping, mashing, geocaching, tagging, searching,
shopping, sharing, socialising and wikkiing. And the list goes on. Yet as the list
goes on it becomes apparent that part of what Web 2.0 does, while doing all
the things on this list and more, is colonise everything in the network. It seems
that there is no part of networked thought, activity or life that is not now
Web 2.0 … Anything can become or be 2.0 as long as it demonstrates or is
affiliated with a certain set of qualities. A list of typical Qualities 2.0 might look
something like this: dynamic, participatory, engaged, interoperable, user-centred,
open, collectively intelligent and so on. Clearly an ‘attitude’ can go a long way.
(Munster and Murphie 2009)
a platform for both control and freedom. This paradox and contestation is
at the heart of social media, and it is a topic we will further examine in the
second half of this chapter.
The term ‘user’ has two connotations: controller and controlled. In com-
puter parlance, the user is in charge of the machine. The user is in control
(at least apparently) of the computer’s operation – the computer seemingly
does nothing unless a user clicks a mouse button or presses a key. On the
other hand, within software development circles the user is often derided.
When we think about users in the context of social media, and particu-
larly within the construct of Web 2.0, which one of these categories is
most applicable? Are users the controllers, who are powerful because they
can create the content in stark contrast to the powerless audience of mass
media, or are users the subjects of control, as their personal information
and creative and cultural labour is monitored and commodified by social
media companies?
These questions are phrased here as binary opposites, and as such they
represent ideal positions at the extreme ends of a spectrum of different
possibilities. Certainly, social media can be seen as empowering or it can
be seen as a set of tools for commercialising the social, affective and crea-
tive efforts of the user. The ambient intimacy of everyday SNS practices
(the way the SNS can sit idle in the background while the user works on
something else, or the way that Twitter is always at hand on the mobile
device but not always actively engaged with) makes it hard to pin down
how much work the user is doing and how this translates to value for
the social media company. Work by Banks and Humphreys (2008) in the
area of game players, along with Bruns’ model of the ‘produser’ (see
Chapter 3 for detailed discussion), have attempted to provide more useful
models for conceptualising the often tacit labour that accompanies con-
temporary media practice today.
in the face of the greatest threat known to humankind helped to elevate the
internet – often referred to as cyberspace – to the sublime. This immunity to
nuclear war extended beyond physical attacks and into the political realm.
The internet, according to John Gilmore ‘treats censorship as damage, and
routes around it’ (Elmer-Dewitt 1993). In this light, the internet could not be
stopped, let alone tamed.
In the decade following these initial works, a great deal changed on the
internet. As described above, the demographic profile of internet users
changed as more and more people came online, and businesses started
colonising the internet. The dotcom crash was for some a kind of proof
that the internet was resistant to control. The emergence of Web 2.0 and the
near simultaneous emergence of mobile internet has again raised questions
about the ability of the internet to bypass conventional control and bring about
social change. Some of these themes, like citizen journalism and online activism,
are addressed in more detail in Chapter 3. Others are less visible, and are
represented in a variety of ways, but all focus on the ways that social
media empower the individual.
Certainly, it is difficult to ignore the way that social media has greatly
expanded the networked individual’s access to information. For the lucky
ones, answers to many questions are only a Google search away (though
veracity may be a little further afield). SNSs, as we will discuss in Chapter 4,
provide them with access to a wider social network, allowing them to find
employment or maintain social relationships that once would have died
owing to distance. Networked individuals have access to a large repository of
media almost at whim (YouTube, for example), and can make creative works
that can be enjoyed by thousands or millions of people where once they might
have been consigned to the back of a cupboard, to be discovered by relatives
sometime after they had died.
On a larger scale, social media has been implicated in regime change and
is playing an increasingly important role in the political: from unofficial
uprisings like the Arab Spring to political campaigns. With the uprising of
the Arab Spring in the Middle East, we have seen ways in which social and
mobile media can be used to help mobilise new forms of politics while at
the same time amplifying paradoxes around media effects and affects. For
example, the control/freedom paradox of the online addressed by Chun (2006)
can be seen in the recent ‘liberation technology’ (Diamond and Plattner 2012)
rhetoric of the Arab Spring in which media can both be a site for emancipa-
tion (in the case of Egypt and Tunisia) and a reinforcing authoritarian state
(Iran). Governments in some countries are becoming interested in social
media as a way to engage more directly with citizens, and citizens are using
social media to draw attention to local issues (Shirky 2009). We give this aspect
of user participation more treatment in Chapter 4.
[W]e must instead ask questions about how, and with what consequences, it has
come about that all social situations (whether at home or work, in public or
in private, at school or out shopping) are now, simultaneously, mediated spaces,
thereby constituting their participants inevitably as both family, workers, public
or communities and as audiences, consumers or users? (2005: 25–6)
Cast in this light, social media can be seen as a step in increasing the control
afforded by the information technologies. This point becomes clearer if we
compare social media with television, and consider how much more useful
consumer information can be gathered relatively easily. While television
provided an important means for product makers to connect with audience
through advertising, television also suffered from a number of shortcom-
ings. Under a broadcast model, for example, nobody can tell what television
station is being watched at any one time. Broadcast companies pay top
dollar to media ratings companies like ACNielsen who go to extraordinary
lengths to determine ratings for television programmes. Broadcasters simply
don’t know who is watching their channels without polling the audience.
Unless somebody is watching you from across the street, you can be com-
pletely certain that when you are watching a broadcast TV programme, you
are the only person who knows you’re watching it.
On the internet, however, every time you sit down at your computer and
access a website, your activity is instantly recorded by multiple sources – if
not your ISP, then at the very least, the website that is receiving your request,
and generally by a much more complex array of monitoring systems that help
website owners and search companies develop a profile of each individual’s
online habits. No matter how little information you provide to sites and ser-
vices (and many people provide quite a lot), the mere fact you are connected
to the internet immediately compromises your privacy. People who wish to
maintain their privacy online must go to significant lengths to do so, and
require a level of technical proficiency that eludes most internet users. When
seen in this light, the internet seems to be as much an advance in control as
an empowerment of the user.
The fact that everything is logged and available for analysis opens up a new
and valuable source of information for companies – very precise information
about the browsing habits of internet users, which in turn allows for a
much, much greater targeting of advertising and the prospect of direct sales
to the consumer, or simply the sale of collected information about users to
other parties. Instead of undermining central authority and power, this
seems to be doing the opposite. The processes of control are now beginning
to move further into our private lives, and users and their cultural and
interpersonal activities are being monitored, regulated and managed like
never before.
In Wendy Chun’s (2006) excellent book on the internet, the dichotomy of
control and freedom are presented as a paradox. Chun argues that the
meaning of freedom has gradually been shifted to incorporate control as an
implicit precondition. In other words, if you want freedom, then you have
to submit to control. This apparent contradiction makes a strange kind of
sense in a post-9/11 world where phantom terrorists lurk in every airport
terminal. According to this logic, there can be no greater threat to individual
freedom than death, and the only thing standing between us and death at
the hands of a terrorist is often control applied through surveillance – full-body
scanners, constant monitoring through security cameras, the tightly regimented
processing of people.
Conversely, our sense of freedom is realised through a sense of control
because the more control one has, the argument goes, the more freedom
you have to do what you want. Here we return to the earlier point men-
tioned near the beginning of this section about users, and the ambiguity
of the term. The user is understood as a powerful individual, and this
notion is reinforced in information technology and the internet all the time.
Microsoft’s slogan ‘Where do you want to go today?’ embraces the notion
of the all-powerful user who is in absolute control of his or her destiny
within the online environment.
As Chun points out, this draws upon earlier conceptions of cyberspace
as being a place beyond space (as we noted above), and also draws upon
popular representations of cyberspace from fiction which preceded and
accompanied the development of the internet. From William Gibson’s
cyberpunk novels and Neal Stephenson’s Snow Crash to a raft of themati-
cally similar books and films, images of cyberspace have been constructed
as a place where the individual is in control. From the utopian holodeck of
Star Trek: The Next Generation to the dystopian virtual reality that fea-
tured as a central plot device in the Wachowski Brothers’ Matrix, virtual
online cyberspaces have been represented as a place where the user exerts
control over his or her destiny by knowing or learning how to control their
online environment.
SNSs, as one example of Web 2.0 applications, also place the user at
the centre of their own universe, a platform on which to stand to engage
with and control their online space. As discussed above, underlying this
so-called user-centred media is the fact that the data is then mined and sold
to advertisers (Vaidhyanathan 2011; Lovink 2012; boyd and Crawford
2012). Consider that Facebook has some 800 million subscribers at the
time of writing, and yet each and every one of these subscribers (users) has
a network of friends in which they are the central nodes that organises
everything else. The user can switch off unwanted contacts, send messages
out to hundreds (or millions, just as easily), all the while developing the
illusion of freedom through control. This isn’t just any space that the user
is (apparently) controlling, it is not even cyberspace; it is their own ‘personal’
space. YouTube places you at the centre of the universe, and MySpace,
as this SNS so helpfully points out, creates an online space that is sup-
posedly mine.
However, as Chun argues, it is in the interests of the companies behind
these services to foster and develop the illusion of control. Providing users
with a certain kind of control (the ability to create profiles and interact
with others and produce cultural objects), all mediated within the company’s
platform, actually establishes broader economic and political controls
over the whole system. We have used the term ‘platform’ here a couple
of times to draw attention to another way of thinking about Web 2.0.
Tarleton Gillespie notes that a platform has a number of definitions in
English language, which together suggest ‘a progressive and egalitarian
arrangement, lifting up those who stand upon it’ (Gillespie 2010: 350).
When applied to Web 2.0 applications such as YouTube, Facebook or
Twitter, the term suggests that the role of the company is impartial – they
are just there to provide a platform that users can stand on and be treated
as equals.
The principles of the platform are enshrined in Facebook’s 10 principles
that, at the time of writing, contain the word ‘free’ or ‘freedom’ no less than
14 times (http://www.facebook.com/principles.php). Facebook’s role as an
open, free conduit for users to become empowered through their networked
agency is also reinforced through its mission statement, which reads in part:
‘to give people the power to share and make the world more open and con-
nected’ (www.facebook.com/facebook). Supposedly, Facebook is simply the
catalyst that makes these things possible. It is a wonder, then, that Facebook
should have been at the forefront of debates about privacy and intellectual
property. Given that many of its changes have given few rights to their users, and
in many cases have appeared to be self-serving rather than user-empowering;
arguments that social media are purely about putting the user in control
need to be taken with more than just a grain of salt.
While the platform metaphor seems to support the empowerment of the
user, it also plays another role, echoing Chun’s paradoxical alignment of
freedom and control. As we discussed above, Web 2.0 companies – those
that emerged after the dotcom bust, or which rode it out – are companies that
recognise the importance of and value of a user’s online practices, and use
those practices to enhance their businesses. This works for users, because it
means there are many services available that are cheap or free. Writing a blog,
putting videos online, developing software, creating or participating in online
communities – all these activities are free, given away by companies whose
actual motivations are less clear. What does an SNS like YouTube, Facebook
or a search engine company like Google get in return for their apparently
altruistic motives? The answer, of course, is control.
Google is a prime example of a company that has embraced (or helped
define) the Web 2.0 platform mantra. As a search engine company, Google’s
most prominent service is its almost universally known search page – a web
page that’s incredibly simple given the behemoth that lies behind it. Every
day Google’s computers index content on the internet, creating a massive
searchable database of most of the pages on the web. This vast database is
then provided to us via the Google homepage, or directly within our browser
software, and provides us with virtually instant access to many topics, and
with a little effort, many more beyond. Yet for all this, Google does not
charge its users a cent. The service is free.
Google’s revenue stream is drawn primarily from its advertising busi-
ness. When you type in a search term like ‘price of tulips’, Google’s
AdWords advertising engine will attempt to connect your search term
with terms that advertisers have purchased. If a match is found, you will
not only get your search results but also a list of results directing you to
advertisers – in this case, mainly florists. Google is at pains to keep this
part of their business separate from the search business. They don’t try to
integrate the advertisements into the actual search results, and they do not
allow anyone to pay money to have their sites appear at the top of search
results. Google’s search algorithms – the methods they use to locate and
present relevant information – are sacrosanct. The reason for this is quite
simple: if the search engine becomes less effective, users will start using
other engines, and that will cost Google market share. So, while Google’s
Adwords advertisements are presented in strategic places on search results
pages, Google uses various techniques to visually separate them from the
list of unpaid search results.
The main point here is that Google’s business model is ostensibly about
users. Google’s ‘about’ page says that the number one thing they know to
be true is ‘focus on the user and all else will follow’ (www.google.com/intl/
en/about/). However, as Siva Vaidhyanathan notes in The Googlization of
Everything, ‘we are not Google’s customers: we are its product’ (2011: 3).
Here Vaidhyanathan echoes Chun’s position – that the publicly touted
importance of Google as a platform for the user lies within a more funda-
mental Web 2.0 business model where users are actually the source of
value, not the information on the web that Google indexes. When we
search on Google, Google builds profiles that match search terms with
sites visited. Websites install Google Analytics, which allows them
to quickly and easily see who is visiting their pages, but also allows
Google to see where people are going. This is generally aggregated – Google
doesn’t care so much where you went today, but does care where ‘you all’
went today.
More than that, companies like Google are engaging in a process that
might be seen as horizontal integration. In traditional hierarchical mar-
kets (lemonade manufacturers, say), horizontal integration is where one
company buys out its competitors and by so doing is able to corner the
market for lemonade. Google’s purchases of YouTube and its integration
of other services like Gmail into one happy family doesn’t immediately
appear to be horizontal integration because all the companies it is buying
are all doing different things: YouTube serves video, Gmail is an online
email application, Google is a search engine and Google+ is a social net-
work tool. If, however, we accept Vaidhyanathan’s argument, that users
are Google’s product, then critically, the same could be said for the sites.
YouTube is a platform creating users, as is Gmail, as is Google+. Therefore,
Google’s purchase of YouTube allowed them to horizontally integrate,
dominating not the streaming video market but the user-as-commodity
market.
CONCLUSION
Web 2.0 is a notion that encapsulates a lot more than the idea that users are
important, or that markets are conversations. It is a philosophy of doing
business in the online environment and it is a response to the challenges of
control in a networked society where many of the structures established by
industrial societies are not always as effective. According to this philosophy,
Web 2.0 is the more advanced, updated, better version of Web 1.0. What’s
updated and improved here is not the technical architecture of the web
itself, but the way that business has come to think about the web, and most
importantly, the ability of business to exert control in an environment
which had previously been seemingly resistant to it.
The changes said to be part of Web 2.0 are sold to users as desirable
primarily because they apparently increase users’ control over their
environments: freedom through control. The changes improve the
agency of the networked individual, and through doing this apparently
give us all more freedom. However, while Web 2.0 can be regarded as
internet companies embracing the user and giving them more control
over what they can do online, it can also be regarded as a way for the
same companies to gain more control over their operating environ-
ments by building better knowledge of their users. The real revolution
encompassed by Web 2.0 is a revolution in thinking, where internet
companies have finally come up with a way of understanding the inter-
net and working out effective methods for using it as a technology of
control in the networked society.
Here we should go back to one of the points we made earlier. The tensions
between control and freedom should not be treated as absolute positions,
where you take a side and fight it out to the end. Instead, these represent
extreme ends of a spectrum in which complex interactions play out.
Sometimes social media is empowering, and may work very effectively to
increase a user’s agency and ability to control and interact with their envi-
ronment. Other times social media can be controlling, providing significant
financial benefits to the social media company but little or no compensation
to the user for their time and energy. Most often social media is both controlling
and empowering at the same time, in an uneasy relationship where a certain
amount of exploitation is negotiated as the price for a certain amount of
empowerment. In the following chapters we will explore some of these
themes further.
NOTES
1 The use of a number like 1.0 is something adopted from computer programming
practice. When a computer program was finished, it was given a version
number – version 1.0 for the first final, complete version of the software. As
the software was further developed other versions would be given a num-
ber like 1.1 or 1.2 to indicate that they represented minor developments. When
the software underwent a major overhaul, it might be given a whole new version
number to indicate the significance of the update – version 1.5 might be upgraded
to version 2.0, for example. During the 1980s, this quickly became part of
the marketing strategy for software, with companies offering up new versions
What makes social network sites unique is not that they allow individuals to
meet strangers, but rather that they enable users to articulate and make visible
their social networks. (boyd and Ellison 2007)
[S]ocial networking sites don’t publicise community, they privatise it. (Andrejevic
2011: 97)
Virtual communities
Back in 1993, Howard Rheingold popularised the idea of virtual communities
in his book by the same name (and subtitled, importantly, ‘Homesteading
on the Electronic Frontier’). Rheingold’s book examined his experiences
with an early online community called the WELL, a pre-internet community
based around Northern Californian new age ideologies. The WELL – an
acronym for Whole Earth Lectronic Link – was a computer bulletin
board maintained by a group of alternative lifestyle users who also pro-
duced the Whole Earth Catalog. Stewart Brand, editor of the catalogue
and founder of the WELL, coined the aphorism ‘information wants to be
free’ (Clarke 2000).
Rheingold’s work popularised the notion of online communities, and fed
into emerging media interest in the fledgling internet. On one side of what
Wellman and Gulia (1999) have described as a Manichaean and unschol-
arly debate were those who derided the idea of online communities as mere
escapism, and yet further evidence of the decay of society and social relations.
Here, the image that was constructed was of a socially awkward computer
nerd, sitting in his basement engaged in a fantasy world that further
removed him from reality and social connections. Castells points out that
these negative images of online communication fed into existing pessimistic
narratives about the loss of community in the modern suburb or megacity
(2001: 125).
Others, like Rheingold, saw potential in these online environments to
create new kinds of communities that could reinvigorate public discussion
and debate. Instead of seeing these networks as socially isolating, many
argued that the internet created a new space for social interaction and
democratic participation, establishing some of the basis for claims about the
Networked communities
As the number of people using the internet began to burgeon in the
mid-1990s, internet researchers had more opportunity to study online com-
munities. Researchers began to discuss and emphasise the continuity of
offline relationships and behaviours of users over discontinuity, amplifying
the importance of social context. While a great deal of research has been
done into online communities over the years, it is difficult to ignore the
contribution of certain key scholars. Wellman conducted some of the first
studies into the ways people used information technologies, and was one
of the first people to argue for the importance of offline factors in online
communication.
In one study Wellman, along with his colleagues, studied the ways that
computer scientists working in universities used computer networks as part
of both their work and social interaction. One of the key findings was that
people communicated more depending on how strong their offline ties were.
People who were already friends, or who had developed relationships with
each other through work, communicated with each other more often on
these networks (Haythornthwaite and Wellman 1998). The findings of this
and other studies led Wellman to argue for more robust models of under-
standing online communities where offline factors were recognised as having
an important role in online communication. Wellman and Haythornthwaite
brought these perspectives together with other research in their edited col-
lection The Internet in Everyday Life (2002).
The recognition that online experiences were grounded in real-world
settings led to what could be termed an ‘ethnographic shift’ in internet studies.
A good example of the ethnographic shift in internet studies is reflected in
Daniel Miller and Don Slater’s (2001) study, which looked at the use of the
internet by Trinidadians. Their focus went beyond the online behaviours of
Trinidadians to engage with the way in which internet use is contextualised
with other (offline) cultural activities. Rather than attempting to generalise
their study to describe all internet behaviour, Miller and Slater concerned
themselves only with explaining the specific instance of internet use that their
study focused on.
Unlike earlier studies of online communities, which typically started
by constructing the online environment as a novel communicative space,
Miller and Slater saw geographical place and the offline social world of
their users as an extremely important consideration in their attempts to
understand Trinidadian use of the internet. They describe their approach
as ‘one that sees it [the internet] embedded in a specific place, which it also
transforms’ (Miller and Slater 2001: 21). In this way, the internet shapes,
and is shaped by, the cultural context in which it is performed. Miller and
Slater found that being Trinidadian was an important factor in how and
why people in Trinidad went online. Furthermore, they discovered in some
cases that the online environment provided a space where people could be
Trinidadian.
A key facet of these new internet community studies was the recognition
that the internet is not one monolithic or homogeneous communication
technology. Instead, the internet is presented as an unbounded object, which
escapes a single all-encompassing definition. Unlike a mass media subject
such as television, the internet is not understood as representing a totality.
From this perspective, the internet is defined by an ongoing process of
meaning making, a process through which the internet is socially constructed
through its use. Moreover, in this understanding there is not one definition
of the internet but many, depending on the context of the people who use
the internet and the context of that use. Miller and Slater, for example,
argued that the internet must be ‘disaggregated’, emphasising that it is
important:
not to look at a monolithic medium called ‘the Internet’, but rather at a range
of practices, software and hardware technologies, modes of representation and
interaction that may or may not be interrelated by participants, machines or
programs (indeed they may not all take place at a computer). (Miller and Slater
2001: 14)
Miller and Slater describe the internet as both a ‘symbolic totality’ (as
people do refer to an entity called ‘the internet’) as well as a ‘practical
multiplicity’ – because one individual’s definition of the internet might be
radically different to another’s (2001: 16). Christine Hine, another leading
researcher who uses an ethnographic approach to the study of internet com-
munities, reinforces this by pointing out that while common parlance might
invoke the phrase ‘the internet’ as a single technological object, the actual
meaning of ‘internet’ can be quite different depending on who is speaking
and who is being spoken to. For example, she refers to the variety of different
attitudes and ideas about the internet reflected by the students in her under-
graduate classes (Hine 1998: 30).
Following her own interests in studying the internet from an ethnographic
perspective, Hine has argued that the internet can be treated as both culture
and a cultural artefact (1998). She points out that the notion of ‘the internet’
has meaning attached to it through a process of social negotiation. For
example, the parents of grown children may have internet access, but not
know what to do with it. However, when one of their children moves inter-
state or overseas, email may become an important method for maintaining
contact. And when a baby is born in a family, a family member sets up a
website with digital photographs of the new baby, and so the internet
acquires meaning again, this time as represented through the web.
Manuel Castells picks up this theme and connects it back to his well-known
overarching metaphor of the networked society. Castells points out that in
studies such as Wellman’s early work, and the Pew Internet and American Life
Project, internet use is revealed as instrumental to the activities of everyday life.
Earlier characterisations of ‘virtual communities’, then, needed to be reconsid-
ered to de-emphasise the virtual and emphasise the connectedness of activities
both online and offline.
Both Wellman and Castells argue that while the family still forms the
basis for many of the strongest social ties in people’s lives, other strong ties
are formed through activities like work or play, and these ties may not neces-
sarily be based on geographic proximity. We may work with people who live
hours away from us in the modern city, but we develop ties with them based
on shared knowledge and experience, and the internet allows us to maintain
these relationships over distance. These relationships take on the character
of networks in that each of us is connected to others by ties that, if mapped
out, would resemble a map of a computer or telephone network.
This does not mean that these ties between people are always strong, but
as Castells points out, just because a tie is weak does not mean that it is
not important. People coming together in an online forum to discuss a
topic of shared interest may come to know one another through their
posts, but never meeting in real life or knowing the real person means these
are weak ties. However, dismissing these ‘weak ties’ as unimportant is
clearly a mistake, as Clay Shirky demonstrates in telling the story of a lost
Motorola Razr phone (2008). In this example of the power of social net-
works, Shirky relates the story of how a lost phone that had been taken by
a passerby was recovered through the activities of an online community.
The links between the protagonist in this story and the community could
be characterised as weak – he didn’t know any of the people who helped
him recover the phone – but the weakness of the relationships did not
make the relationships ineffectual.
Wellman has pointed out that in many modern societies, a phenomenon
he calls ‘networked individualism’ has arisen; that is, individuals build net-
works to solve problems, make decisions or get support. The internet has
vastly extended these networks so that they are no longer constrained by
space. This change moves people away from traditional geographically
bounded social groups – neighbourhoods, for example – and towards
‘sparsely-knit and loosely-bounded networks’ (Wellman 2003). For Castells,
networked individualism is part of the networked society, rather than the
internet per se, but can be supported and augmented by the internet to pro-
duce ‘new patterns of sociability based on individualism’ (Castells 2001: 130).
To illustrate this, ask yourself a question: if you are thinking of buying
something – let’s say a new car – do you first get advice from a neighbour
or someone you work with, or do you Google it? If the answer is the latter,
then you’re engaging in networked individualism.
Networked publics
With the rise of the SNS, questions about the nature of online community
have again become a topic of interest. danah boyd has reworked the idea of
networked communities within the SNS to describe networked ‘publics’ as
an extension (but not necessarily an alternative) to the word ‘communities’.
When we speak of ‘the public’, we are in fact talking about a collection of
There are two fundamental components outlined here that are worth reiterating:
networked publics are both spaces and groups of people who are connected
through practice and technology. They are ‘simultaneously a space and a
collection of people’ (boyd 2011: 41). Importantly, boyd argues that these
publics are not just networked because they are linked together by the tech-
nology, but they are transformed and restructured by networked media.
SNSs are examples of online technologies that support the production and
reproduction of networked publics. As boyd notes, there are three key
dynamics in SNSs: invisible audiences, collapsed contexts, and the blurring
of public and private (2011: 49). In examining the transformation of publics,
she observes that ‘the affordances of networked publics rework publics more
generally and the dynamics that emerge leak from being factors in specific
settings to being core to everyday realities’ (p. 53). In the pervasiveness of
networked publics, boyd perceives erosions of physical barriers while, at the
same time, ‘many people feel unmotivated to interact with distant strangers’
(p. 53). In sum, in networked publics, ‘attention becomes a commodity’ (p. 53).
While the notion of networked publics has considerable overlap with
Castells’ concepts of a networked community, networked publics differs
primarily in its use of the idea of ‘publics’ rather than ‘communities’ as the
organising metaphor for conceptualising online users. This is a useful alter-
native, because it allows us to drop the cultural associations caused by that
term, a problem that Castells himself is keen to avoid (2001: 127).
NETWORKS OR COMMUNITIES?
There is still healthy debate in the scholarly discourse about the nature of
the social structures that are enabled by network technologies. While Castells
and others have moved away from the idea of online communities, and
embraced the network metaphor, others argue for the value of understand-
ing networked sociability as a kind of community.
The term ‘community’ is complicated and contested. There are a variety
of definitions that make it difficult to use without also accepting the intel-
lectual baggage that comes with them. This is why some, like Castells,
prefer to avoid the term altogether. Others have persevered, with interest-
ing results. Celia Pearce has attempted to sidestep some of the baggage by
drawing on German sociologist Ferdinand Tönnies’ definition of commu-
nity (Gemeinschaft) as ‘an association of individuals with a collective will
that is enacted through individual effort’ (2009: 5). Pearce notes that ‘a
community of practice is defined as a group of individuals who engage in
a process of collective learning and maintain a common identity defined by
a shared domain of interest or activity’ (2009: 5). One of the key factors
in making and maintaining a community is social capital. The concept of
social capital requires a few paragraphs to explain, but it is important so
it’s worth a minor detour.
The term ‘social capital’ is used by Pierre Bourdieu (1984 [1979]) in his
widely read work Distinction: A Social Critique of the Judgement of Taste.
Bourdieu was interested in taste and how taste becomes naturalised in
people. In other words, Bourdieu wanted to know what societal mechanisms
led to a person from one kind of background enjoying caviar and opera,
while someone else likes fried chips and hard rock music. These tastes are
naturalised so that people don’t even know why they like them, or how they
came to like them – it just seems ‘natural’ for them to feel such responses.
For Bourdieu, capital was a form of ‘knowledge’ that helped produce and
naturalise taste.
To explore this, Bourdieu interviewed 1,200 French people from varying
class backgrounds about their tastes in art, music and popular culture. As a
result of this research, Bourdieu deployed his concept of capital to discuss
what he saw as three significant kinds of capital that influenced people’s
taste: cultural (informed by education and upbringing), social (community
and networks), and economic. These factors, along with the individual’s own
‘habitus’ (the regulatory patterns of everyday life and everyday practices),
were the contributing factors in determining one’s identification with a par-
ticular lifestyle niche.
Bourdieu’s concept of ‘social capital’ took on new significance when it
was reworked by James Coleman (1988) to infer a more ego-centred con-
cept. Social capital was then redefined by Robert Putnam as part of his
savage exposé on the declining role of community and social welfare in the
US in Bowling Alone (2000). Here Putnam characterised social capital as
facilitates specific types of connections between people that can generate social
capital … SNSs are continuously reshaping our social networks and the
communication practices we use to maintain them, and thus constitute a
vibrant, important, and challenging context for studying communication
practices and their social capital outcomes. (2011: 141)
Center’s Internet & American Life SNS survey conducted in 2010 supports
this claim. In this study (which was conducted by phone) it was reported that
89 per cent of North American users’ Facebook friends are people they have
met more than once in person, and a high percentage of the total number of
friends on Facebook are people who are known to the user through school/
university, work, family or volunteer groups (Hampton et al. 2011).
This data suggests that SNS relationships are geographically and socially
oriented towards the local. Significantly, these findings suggest an important
point: it’s not just the size of the network that matters, but the quality of the
connections. More intimate connections seem to be more valuable and
more common for SNS users than large numbers of less intimate connections.
This weakens the importance of network effects and increases the relative
importance of intimacy, suggesting that rather than being characterised as
networked publics, SNS-constructed publics might alternatively be defined
as ‘intimate publics’.
Intimate publics
The idea behind intimate publics is that as social and mobile media become
more pervasive, different modes of using these media mean that increasingly
publics are defined by the strength of their relationships, rather than the total
number of network connections. The term ‘intimacy’ when used here not
only refers to the common-usage kinds of intimacies that exist between lov-
ers, family members or close friends (though these can and do play a role),
but also to intimacies that can exist at a social or cultural level. As Michael
Herzfeld observes, cultural intimacy describes the ‘social poetics’ of the
nation-state; it is ‘the recognition of those aspects of a cultural identity that
are considered a source of external embarrassment but that nevertheless
provide insiders with their assurance of common sociality’ (1997: 3). To put
it a different way, intimacy can be something that exists between strangers
because of the common bond they can share by virtue of them belonging to
the same cultural group, whether that be a town, city, nation or some other
sociological or political grouping. An example might be a small country town
where everyone who grew up there knows that the town has a reputation for
having the worst weather in the nation (or the most boring night-life, or the
most superstitious people in the province and so on). This shared knowledge,
even if it is potentially embarrassing, also acts as a kind of social bond – a
‘cultural intimacy’.
An example of when cultural intimacies come into play would be when
two people from similar cultural backgrounds but who are otherwise stran-
gers accidentally meet on a train in an unfamiliar country, and find that they
immediately have a connection. The way that SNSs have been developed or
simply picked up by different nations and cultures across the world is
also a tangible example of this kind of cultural intimacy. China’s QQ is not
just any SNS; it is the Chinese SNS, and to use QQ is to participate in a
community that shares a set of cultural intimacies. Facebook has become
popular in Korea, but it does not speak to and of Korea in the way Cyworld
minihompy does (Hjorth 2007). One of the first SNSs, Friendster, began in
the US but soon became widely adopted in the Philippines. Many Brazilians
use Google’s original SNS, Orkut. The list of SNSs which have become
associated with particular nationalities goes on. In these examples it can
be argued that a sense of community emerges through the performance of
personal intimacies and the aggregation and identification of public socio-
cultural intimacies.
At a more interpersonal level, SNS can be regarded as a technological tool
that mediates interpersonal intimacies. Esther Milne (2004) has suggested
that new media, such as SNSs, function socially as tools to mediate intimacy,
and should be historically contextualised with other technologies that have
filled this role. Far from being a new phenomenon, others have argued that
intimacy has always been mediated (Hjorth 2005), with examples of other
technological intimacy mediators including texting on mobile phones, the
telegraph (see, for example, Standage 1998), and written correspondence.
When seen in this light, the only thing that has changed with the arrival of
SNSs is that people have appropriated computer networks as yet another
technology that mediates intimacy.
Intimacy in SNSs is also represented by how people manage their online
details. All SNSs have a concept of a profile, or something similar, which
reveals something about the user. This profile, as we mentioned above, may
include images and other information, and can often be made public or
private through the software, with these two categories defined fairly rigidly:
sharing with everybody (public) or sharing with friends (private). Google+
introduced the idea of ‘circles’ that allow people to place friends into differ-
ent and potentially overlapping user-defined categories like ‘work mates’ or
‘school friends’. The amount of information about oneself that is revealed
through the profile is part of the performance of intimacy online.
For boyd (2011), US youth have responded to the growth in networked
media by creating networked publics that engage in various forms of semi-
public and semi-private modalities. Choosing what to share and who to
share it with allows people to control the privacy or publicness of their
information that goes beyond the relatively clumsy tools provided by social
networks. Instead, people use new kinds of strategies to control their infor-
mation, carefully assessing the social value of revealing information against
the potential costs (boyd and Marwick 2011). Privacy, in other words, is
not simply an on/off switch or a setting that is chosen and then ignored.
Rather, the boundaries between public and private are something that
people are constantly revising as a perpetual work in progress (Hjorth and
Arnold 2013). Rather than viewing all SNSs as ‘networked publics’, as
boyd does for the context of the West, we could characterise SNSs instead
as ‘intimate publics’ that are played out, and through, social media prac-
tices. Some of these practices are expanded and examined in a Korean
context in Chapter 7.
above (Hine 1998; Miller and Slater 2001), some researchers of SNSs have
used qualitative approaches (boyd 2004, 2009, 2011; Miller 2011) to focus
upon more detailed, local and personalised understandings of the intimate
and social dimensions of SNS use. Other approaches are broader, analysing
and criticising social media, or particular aspects of it (Shirky 2008). Studies
around SNSs as part of broader forms of twenty-first century media literacy
(Ito et al. 2008) have also begun to emerge.
SNSs are also implicated in other areas of research, like journalism, poli-
tics and law. As we will see in Chapter 4, social media are enabling changes
in the way people engage with politics via citizen journalism and online
activism, and SNSs are often the sites and technologies that support these
activities. The discussion is thus rapidly moving beyond Western and teen
contexts and expanding to engage with broader social issues that include
censorship, privacy and copyright. This means that study into SNSs is highly
multi-disciplinary, and thus methodologies and motives for research are
highly varied. In the remainder of this chapter we are going to paint an
impression with broad brush-strokes of some of the major areas of current
research that engages with SNSs. As we will see, SNS studies are maturing
and multiplying, reflecting the way that SNSs are occupying an increasingly
important role both within and across societies.
Non-Anglocentric studies
Initial research into SNSs focused upon contexts familiar with the researcher’s
own cultural context. Given the nature of English as the lingua franca in
‘global’ studies, many studies into SNSs came from Anglophonic researchers
(Goggin and McLelland 2009). Later, non-Anglophonic models began to
grow, relating the cross-cultural and global nature of SNSs (Yang et al. 2003;
Goggin and McLelland 2009).
This shift is important when we consider that many of the locations
for hardware and software manufacturing have been situated in non-
Anglophonic contexts. The Asia-Pacific region, which encompasses locations
such as India and China, is one such non-Anglophonic context that is home
to some of the oldest SNSs such as Cyworld (Hjorth and Kim 2005). Many
nations within the Asia-Pacific also have longer histories with mobile and
locative technologies (see Chapter 7) and demonstrate examples of new social
uses of the technologies in areas such as gaming, location-aware social media
and social networks (Hjorth and Chan 2009). This makes the region difficult
for researchers to ignore.
Studies of social media use in the Asia-Pacific region, for example, have
found that for youth, SNSs are not only a fundamental part of everyday life
and the exercise of their social capital (see below), but also a space that
helps to maintain intergenerational ties when geographic distance might be
involved – a point we will illustrate more clearly through a case study of
social media games in Chapter 6.
As studies in the US have found, in many countries in the Asia-Pacific it
is no longer just ‘youth’ that are using SNSs, as adult-to-adult and inter-
generational forms of dialogue and digital literacy expand with increased net
accessibility. Moreover, the demographics are shifting too, as internet access
becomes a tool not just for the rich or middle classes, but as an integral
part of a new mobile working class (Qiu 2008). Interestingly, in China it
is working-class use of the internet that is growing exponentially, mainly
through mobile media (CNNIC 2009). This gives us insight into the ways that
some of the largest societies on Earth are developing and integrating new
media into this development.
Cross-cultural approaches to the study of SNSs also provide us with
insights into the differences and similarities between new media practices in
different cultures. By moving the cultural frame to another context, we often
find we learn a lot about the way that social media works in our own culture,
as behaviours which are rendered invisible through familiarity become visible
in unfamiliar cultural contexts.
Privacy
One of the more significant and complex concerns associated with social
network sites is the issue of privacy and, not surprisingly, this is a burgeon-
ing area in the field of SNS research. This issue has received a great deal
of attention not just in popular media, but in critical literature as well.
Privacy is a highly popular topic precisely because it is so complex, and
because it provides a handhold for anchoring fears and anxieties about a
new technology.
Most new technologies are met with an initial social response that contains
elements of fear, paranoia and anxiety (Kember 1998). The media frequently
plays into these fears, often emphasising them out of proportion by focusing
on the exceptional and presenting it as the norm. SNSs are no exception to
this, and have been implicated as the culprits in a range of modern anxieties
from playground bullying to terrorism. The media is happy to offer a range
of anecdotes to support these fears (Marwick 2008).
For SNSs, concerns about privacy are tantamount. Parents worry that
their children are able to publish too much about themselves, attracting
unwanted attention, damaging their reputations or breaking laws that they
may not even realise they are breaking. For their part, teenagers often resent
their parents’ interest in their online activities, and see parental attempts at
monitoring their online behaviour as an invasion of their privacy (boyd and
Marwick 2011).
Some studies have examined the uses of SNSs and their built-in privacy
systems and have concluded that many users of SNSs – especially younger
users – do not take advantage of the privacy features (Gross and Acquisti
2005). Others suggest that to understand young people’s attitudes to pri-
vacy in the online environment requires a more thorough understanding of
online practices. boyd and Hargittai (2010), for example, show that teens
in their studies are very aware of the privacy implications of their online
activities and are very selective about what they share and who they share
it with.
Much of the work on privacy concerns the relationship between individu-
als and how people negotiate, public and private, between themselves and
other users of the SNS. However, another dimension of privacy that has been
less explored concerns the relationship between users and the SNS itself. The
issues associated with this go back to fundamental questions about user and
used, and control, that we examined in Chapter 2.
In a similar vein, Jer Thorp, a Canadian artist and educator, developed Just
Landed In ..., a visualisation of the location of Twitter posts that contain the
phrase ‘just landed in’. In this visualisation, lines appear superimposed across
the globe, representing people’s travel and providing a visual representation
of human movement through space and time and their connections with
social media (Thorp 2009).
Others are using social media sources and applying sophisticated infor-
mation processing techniques to explore the relationships and uses of SNSs
as a way of providing quantitative data that supports (or refutes) qualitative
research. Thelwall has conducted a number of such analyses of SNSs. In one
such study, Thelwall (2008) collected a random selection of around 20,000
profiles from MySpace users and analysed the information provided in
these profiles to develop a statistical overview of data such as age, religion,
frequency of access and so on. Thelwall found that users were younger than
previously reported (the median age was 21), and that the median number
of friends was 27. He also saw that there were groupings of this data; there
were a lot of people who only had 1 friend, a group who had 2–9 friends,
another that had 10–90 and a third group who had more than 90. The authors
noted that this was consistent with qualitative research that suggested that
people who use MySpace categorise friends into close friends, acquaintances
and strangers. Given that the median was 27 friends, the data suggested that
for MySpace users, most friends were acquaintances rather than close friends.
The suggestion here is that people made friends online, rather than using
the SNS as a way to maintain offline relationships. This is at odds with
some research, although it may also reflect the kinds of users that were
using MySpace.
Others still are using social media data to gain insights into how people
respond to events. For one example, Bruns et al. (2012) analysed social
media postings in the aftermath of the 2011 floods in Brisbane, Australia.
They followed the changing social media etiquette and the response by the
Queensland police to the disaster through large-scale data collection of
Twitter hashtags, and then created visualisations of the patterns of media
use and themes around this event. For another example, in Hjorth and
Kim’s (2011) case study of the role of social and mobile media in the crisis
of Japan’s earthquake, tsunami and Fukushima nuclear reactor disaster
(known as 3/11), they found that many victims of these events relied on older
media and communication methods. This seemed to be because the partici
patory nature of social media made understanding the event more confusing
than when it was encountered through ‘packaged’ media like television. One
respondent noted that prior to 3/11 he had viewed Twitter as ‘conversational’,
but during and after the crisis his opinion changed dramatically; he felt
bombarded by conflicting social media threads which then made him
redefine the discourse surrounding Twitter as more like a ‘conference’
than a ‘conversation’.
If we are to understand the impact of social media in times of crisis or
political upheaval we need to move our analysis from an effect-orientated
focus to affect-orientated one. We need more hybrid studies that combine
the micro and macro analysis, using such methods as Ken Anderson’s
‘ethno-mining’ in which ethnographic processes are used to analyse data-
mining (Anderson et al. 2009). Data-mining and visualisations may paint
pictures of media phenomenon during these times, but as anyone who has
lost an intimate will attest, they are abstract in the reality of grief’s texture.
Instead, there needs to be more ethnographies of media affect and mobile
intimacy to understand the micro, meso and macro levels of intimate publics
in times of trauma.
There is certainly a proliferation of data in current society, leading some
to call this moment the ‘era of Big Data’. However, as boyd and Crawford
note, ‘Big Data’ is ‘in many ways, a poor term’ (2012: 663). They argue
that:
like other socio-technical phenomena, Big Data triggers both utopian and
dystopian rhetoric. On one hand, Big Data is seen as a powerful tool to address
various societal ills, offering the potential of new insights into areas as diverse
as cancer research, terrorism, and climate change. On the other, Big Data is
seen as a troubling manifestation of Big Brother, enabling invasions of privacy,
decreased civil freedoms, and increased state and corporate control. As with all
socio-technical phenomena, the currents of hope and fear often obscure the
more nuanced and subtle shifts that are under way. (2012: 663–4)
CONCLUSION
In many ways, SNSs are the definitive social media technology. They are the
interface through which people engage with social media, and increasingly
they are the way that people engage with the internet. SNSs are shining
examples of the Web 2.0 ethos we discussed in Chapter 2 – they are user-
oriented, providing a space for people to make things, share things, communi-
cate and connect with each other, allowing for a wide range of empowering
practices from activism to creative production. However, we must not forget
that they are also commercial ventures, and as such can and do commercial-
ise users by collecting and using their data and details. SNSs are free, in that
we do not pay subscription fees to access them, but companies like Facebook
have multi-billion dollar valuations. The question we must ask here is, why?
If we are not paying in cash then are we paying in some other way, and are
we getting value?
In the first part of this chapter we have looked at the evolution of the inter-
net as a medium for sociality. From virtual communities we have explored
networks, networked publics and intimacy as structural features of the sociality
afforded by SNSs. Questions persist about whether SNSs and online inter
actions are better understood as a network or a community. We suggest that
SNSs exhibit properties of both.
SNSs are also both global and local. From Manchester to Manila, and from
Seoul to Sydney, people are using SNSs for similar reasons, and in this way
they are a powerful symbol of the way that communication technologies
really are spanning the globe, crossing cultures and encouraging research that
appreciates the diversity this represents. Yet SNSs are also intensely local,
emphasising rather than erasing geographical proximity. SNSs may span the
globe, but the relationships we maintain through them are generally those
that we also maintain through face-to-face contact.
There is, of course, a faddish element to SNSs, and it is likely that today’s
darlings of the digerati will be tomorrow’s old news, but it would be a mis-
take to dismiss SNSs as nothing more than a fashion, as they reflect social
practice as much as they create it. SNSs have not become popular because
they create social networks, but because they provide a space for social
networks to exist. These social networks, as we have seen throughout this
chapter, and indeed in other chapters, exist in both online and offline worlds.
They provide spaces for online relations, but also structure our offline
relationships.
I f there is one word that summarises the particular quality of social media,
it would be ‘participation’. Unlike the mass media before it, social media is
fundamentally a participative medium. Our online experience increasingly
involves methods of actively providing information about what we are doing,
or what we think of something. This might be as simple as a Facebook ‘like’
button, or as involved as maintaining a blog. Participation can take various
forms of agency from user generated content (UGC), in which users forward
content made by others, to user created content (UCC), in which the content
is made by the user. Every time we participate we partake in various forms of
labour sharing – from creative and social to emotional and affective labour.
In each cultural context, what it means to participate takes on different
dimensions. For example, in China where the internet is highly regulated by
the government, participation can often take the form of what in the West
might be called ‘lurking’. After all, imagine if the 457 million internet users
(CIW 2012) all spoke online at once. If everyone were talking, who would be
listening? Indeed, listening as a form of participation has only recently gained
attention in critical work concerning social media (Crawford 2009).
The emergence of social media and its emphasis on participative modes
of use has many significant implications for the study of media and society
more broadly. In each location, the implications of ‘click-activism’ are play-
ing out with different results (Nugroho and Syarief 2012). For example, we
can see changes in the fabric of activism in the emergence of social uprisings
like the Occupy Wall Street movement, or the Arab Spring uprisings of 2011
(Diamond and Plattner 2012). New dimensions are added to crisis manage-
ment and response as seen during the New Zealand earthquake of 2011
(Bruns et al. 2012) and the events of Japan’s 3/11 tsunami (Hjorth and Kim
2011). Perhaps most significantly for journalists, social media is throwing
up challenges to the privileged position of journalists and the news media
as the sole arbiters of reportage. As participative forms of media like blogs
and SNSs become more mainstream, we are seeing the rise of the citizen
journalist – a phenomenon that is challenging conventions around press
First, we will expand upon one of the threads explored in Chapter 2 and
engage more deeply with how academic scholars have conceptualised the
idea of audiences/users as producers. We also deploy Bruns’ term ‘produser’,
which we think is a very helpful term for describing the kinds of productive
behaviours to be addressed in this chapter. After we have defined what we
mean by produser, we will tackle another slightly more straightforward and
related theme: user created content (UCC). We will define what we mean by
these terms, contextualise them within the literature and then go on to look
at some of the implications of produsage and UCC.
Once we have defined the ideas of users as co-producers and of UCC, we
will move on to look at how these concepts play out in more generalised
and grounded practice. First, we will consider crowd sourcing as one way
that combined user production can be utilised to great effect in an online
environment filled with literally millions of users. We will then examine the
phenomenon of citizen journalism, before going on to examine online activism,
both quite practical examples of the kinds of behaviours that participatory
media enable. We will limit our discussion of journalism in light of the
book’s focus upon social media. Entire books have been written about citizen
journalism (Gillmor 2006; Meikle and Redden 2010) and online activism
(Zuckerman 2008; Pickerill 2010; Lovink 2012), and where relevant, we
encourage you to refer to these works if you want to develop a more thor-
ough understanding of the phenomena.
their lives into characters and storyworlds of Star Trek. In other words,
Trekkers might visit Star Trek whereas Trekkies completely inhabit the
world of Star Trek. Thus, engagement with the TV series was much more
than passive watching; it was highly creative and active with audiences
participating in the making of meanings and interpretations. Indeed, it
was in these fan communities – both those of gamers (we will come back
to this in Chapter 5) and fans of popular culture – that the first kinds of
production emerged, well before anyone coined the term ‘Web 2.0’, and well
before the first social media sites appeared. In reflecting on these observa-
tions, Jenkins describes something he called a ‘participatory culture’, which
he defines as:
a culture with relatively low barriers to artistic expression and civic engagement,
strong support for creating and sharing one’s creations, and some type of
informal mentorship whereby what is known by the most experienced is passed
along to novices. A participatory culture is also one in which members believe their
contributions matter, and feel some degree of social connection with one another
(at the least they care what other people think about what they have created).
(2006: 3)
While Jenkins was concerned with the way that audiences produced, others
pointed to the emergence of a new class of worker – the professional ama-
teur, or ‘pro-am’ – whose production did not fit into ideas of either amateur
production or professional production, but occupied a territory somewhere
between (Leadbeater et al. 2004). The professional amateur was someone
who worked at their interest like a professional, spending as many hours on
their endeavour as they might in their day job, treating it like it was a task
that earned money, and yet was not a professional since they were not part
of a professional community and did not get paid for their work.
Other scholars looked more intently at the way in which users who pro-
duced undertook their work. Australian academic Bruns has engaged with
the idea of the user who produces through his book Blogs, Wikipedia, Second
Life and Beyond (2008), in which he describes the term ‘produser’ as a
conflation of the words producer and user. This coining of a new phrase is
not just about coming up with a word to mark academic territory; it is also
about developing a word that is simply lacking in the English language to
describe something that has become so ubiquitous that it is simpler to use a
new word than to continue using a phrase.
Other terms have also been coined, but as yet none have stuck. The idea
and practices are quite new, and it will be some time before the phrases are
resolved. Perhaps in the future we may simply return to a term like ‘user’,
and simply incorporate the understanding that users are also producers into
this term. In any event, the term ‘produser’ is useful in the context of this
discussion, and you will find that throughout this book we use Bruns’ term
when we are referring to users who produce.
data and a carefully crafted blog post that is uploaded to the internet for the
express purpose of being read by other users, or a short film that may rep-
resent many hours of production work that is then uploaded to YouTube. In
this book we draw a distinction between content that has been created pur-
posefully by a user expressly for exhibition to others, and content that is
generated by users as a result of using social media. We use the term ‘UCC’
to more precisely refer to the kinds of content produced intentionally by
users, usually for the purpose of consumption by other users.
There have been a number of criticisms of UCC, some of which we
explore in more detail in other chapters (see Chapter 5, for example). One
of these criticisms is that (often) amateur UCC is displacing established
forms of content creation where the content creator is a professional who
has significant training and experience in their field. This is a pronounced
criticism in some fields, such as journalism and the arts, as we will see below.
In seeking to move the discussion of user creativity beyond the professional/
amateur dichotomy, Jean Burgess (2007) has used the term ‘vernacular
creativity’ to identify UCC as something that is characterised by the vernacular
and everyday. She points out that while creativity is often seen as the exclu-
sive domain of trained elites like artists or design professionals, creativity has
always been an activity that everybody engages in, even if in the past it was
not always visible.
Scrapbooking, writing of short stories, and family histories, home crafts
and decoration are all examples of vernacular creativity that are no less
creative just because they are not produced by professionals or widely acces-
sible in the public domain. For Burgess these activities are about cultural
citizenship, a concept that expands and redefines classical notions of citizen-
ship that are based on participation in political activity. Here, the production
of creative works acts as a way of asserting and defining one’s citizenship,
which is ‘practised as much through everyday life, leisure, critical consump-
tion and popular entertainment as it is through debate and engagement with
capital “P” politics’ (Burgess et al. 2006: 1). We will return to some of these
issues around UCC in Chapter 5 when we come to look at art and cultural
production in the age of social media.
While some of the material about user participation explores cases where
individual produsers can participate in activities that were once beyond them,
others focus on the action of groups of users working together to produce
materials or solve problems. James Surowiecki’s Wisdom of Crowds (2004)
explored the idea that large groups of people can often solve problems that
individuals within the crowd cannot. Tim O’Reilly picked up this idea and
worked it into his formulation of Web 2.0 (2005).
Extrapolating from this idea, Amazon launched their Mechanical Turk ser-
vice in 2005, which allowed people to either pay a fee to have some problem
worked on by a group, or on the flip-side, to be paid a small fee for participation in
helping to solve a larger problem. Other online services provide facilities that
allow users to associate tags (single words or short phrases) with content,
thus allowing large numbers of people to build up keyword indexes based on
human-entered information.
The National Library of Australia has also utilised crowd sourcing to
help fix text from scanned newspapers for their online Trove service. Trove
is a digital repository of Australian media, including print, images and
audio. Part of their archive consists of newspapers dating back to 1803,
resulting in millions of pages of newsprint available online. These pages
were passed through an optical character recognition (OCR) process that
automatically turned the printed text into electronic text, which allows the
text to be indexed and searched. This is clearly a very valuable resource for
historians. However, OCR is not perfect and newsprint – especially material
that is smudged or damaged – does not always scan correctly. To fix this a
person needs to read the original text, compare it to the OCR text and make
any corrections manually. For a collection the size of Trove, this is an
intimidating amount of work.
In order to tackle the sheer size of this task, the National Library of
Australia developed an interface that allowed anyone on the internet to
register and edit the text themselves. This has proved a very successful way
of using crowd-sourcing techniques to help produce a publicly accessible,
searchable archive. The success can be measured in part by metrics. An early
version of the service was released in July 2008 and as of February 2009,
2.2 million lines and 104,000 articles had been corrected by internet users
(Holley 2009). While this is only a small percentage of the total 3.5 million
articles in Trove (which is planned to increase to 40 million), the strong
engagement of the community so rapidly suggests that there is a great deal
of value in crowd-sourced applications, especially when people perceive
they are helping the community. As one person noted about her motivation
for fixing material on Trove:
‘I enjoy the correction – it’s a great way to learn more about past history and
things of interest whilst doing a “service to the community” by correcting text
for the benefit of others’ but also, her motivation for continuing to change
pages was driven by ‘the knowledge that you are doing something that will
benefit future people that wish to access articles on their family history’.
(Holley 2009: 17)
This gives us a small insight into the kinds of motivations that drive user
participation. If Mechanical Turk and the Trove experience demonstrate the
power of crowds, then Wikipedia provides a compelling case for the power
of utilising millions of online users to create an online knowledge repository.
Wikipedia is a portmanteau of two words – ‘wiki’ and ‘encyclopaedia’. The
word ‘wiki’ was developed from a Hawaiian word that means ‘quick’. In
practical terms, a wiki is a web-based system, developed by Ward Cunningham
in 1994, that allows people to write and edit a shared document, which can
be quickly linked to other documents. This allows a group of people to
collaborate on the production of documents that contain a hierarchical and
inter-linked arrangement of content. A perfect use for this kind of service is
the production of documents that consist of small chunks of self-contained
information that have relationships with other chunks of information – a user
manual for a piece of software, a technical manual for a mechanical device,
or an encyclopaedia.
Realising this potential, the founders of Wikipedia sought to create the
world’s largest repository of knowledge on just about any topic imagi-
nable. To facilitate this, they created a wiki that they then opened to
everyone to contribute to. Wikipedia has quickly become the world’s largest
source of knowledge on a variety of topics, from the history of the
Roman Empire to biographies of actors who starred in obscure cult TV
shows. Anyone can create pages in Wikipedia, and anyone can also edit or
amend information on Wikipedia – allowing not only for the creation of
a wide variety of information, but also for an iterative process of correc-
tion and amendment, towards the goal that over time the quality of the
entire source will improve.
Some of the major issues with crowd-sourcing come from a failure to rec-
ognise its limitations. The main limitation is the inherent fallibility of crowds
and the disproportionate ability of a committed individual or small group to
bias results. As soon as the Mechanical Turk services started, for example,
people quickly found ways to maximise the amount they could earn by pro-
viding random results or by automating their input. This means that certain
kinds of activities that might utilise the service are open to abuse. Wikipedia,
while offering far more information than traditional encyclopaedias, cannot
assert the same level of quality that an encyclopaedia with a tightly controlled
editorial process can. While the crowd will tend to correct errors and omis-
sions, individuals and groups who have strong views will attempt to sway
articles to reflect their points of view, which has led to some significant
disagreements over certain contentious subjects.
CITIZEN JOURNALISM
Citizen journalism is fundamentally about the collision between traditional
news reporting and participative media. In the online environment, users can
take an active role in the production of content, and when this extends to
reporting on events, it constitutes citizen journalism (although as we will see
this definition makes the issue seem more clear-cut than it is). Citizen journalism
has appeared at the intersection between the challenges faced by traditional
news reporting in the internet age, the emergence of social media and the
growing ubiquity of devices like mobile phones that can capture images and
video. For some like Gerard Goggin, the rise of mobile media such as the
camera phone, along with personal but broadcast media like Twitter, has
made messages, contexts and content more intimate (2011). As part of a
broader movement of intimacy into the public realm (Berlant 1998), the
role of mobile media – as one of the most personal and intimate devices
(Fortunati 2002) – has had an impact upon journalism. With ‘amateur’
images taken by the mobile phone having more of a raw and unpolished feel,
the texture of visuality in journalism has changed. It is not uncommon for a
journalist to evoke that amateur feel to give the news a more intimate and
trustworthy affect. The use of camera phones to shift notions of intimacy
and place is discussed further in Chapter 7.
In the contemporary media landscape, one of the areas that has come
under the most pressure is news reporting, and in particular, newspapers.
Newspapers, which have long been supported by advertising revenue,
have steadily lost this revenue to the internet as advertisers follow user
attention online. Compounding this, many online news sources offer con-
tent for free and directly compete with traditional newspapers for their
readers. This loss in revenue for newspapers results in increased pressure
for newspapers and related news organisations to find cheaper ways to
produce and present the news. Social media offers an interesting alter-
native for the collection of news stories, allowing news organisations to
crowd-source content which not only gives them access to content that
would not have been possible to get in years past but also to get it for very
little money.
have better access to people and have honed their skills in interviewing and
research. In effect, this argues that there is more to journalism than having a
publication platform, and that participative media, in providing people with a
platform, is not enough to make them journalists. This argument does not
necessarily dismiss the value of citizen journalists, but emphasises the need for
professionally trained journalists even in an era of participative media.
Another serious criticism of citizen journalism is that it lacks the trans-
parency of traditional news media. Citizen journalists do not need to follow
professional codes of conduct. Furthermore, while a professional journalist
is kept in check by editorial processes, citizen journalists are free to write
whatever they want. Any political affiliations or bias on the part of the
citizen journalist is therefore harder to determine, bringing the impartiality
and accuracy of their work into question, and placing more responsibility
on the reader to determine the quality of the source. Trust becomes a major
issue that was in the past mitigated to some degree by the reputation of the
news source.
Citizen journalists are also far more vulnerable than journalists working
for news organisations because they lack the protection that is often extended
to them as employees. Some of the best journalism provokes strong responses
from people and opens journalists to legal (and sometimes physical) retal-
iation. Journalists working for news companies enjoy a certain amount of
protection, especially legal protection against civil litigation (for example,
being sued for libel). Because citizen journalists are working for themselves
they have no overarching protection, which in some circumstances seri-
ously compromises their ability to report on provocative issues. As Dan
Gillmor asks:
Who would have exposed the Watergate crimes in the absence of powerful
publishers, especially the Washington Post’s Katharine Graham, who had the
financial and moral fortitude to stand up to Richard Nixon and his henchmen?
(Gillmor 2006: xxvii)
The speed with which social media allows news to be distributed also raises
issues. Traditionally, the professional journalist’s job was not simply to relay
pieces of information to their readers, but to sort rumour from fact, to ana-
lyse and synthesise a story from multiple, often-contradictory sources. With
instant messaging services like Twitter, news can be disseminated so rapidly
that it bypasses normal editorial control, potentially leading to incorrect and
misleading reports being released which are then left to readers to analyse
and evaluate. The short 140 characters or less format of Twitter in many
ways remediates its technological predecessor, SMS (short messaging system).
Twitter thus borrows the etiquette of SMS as both compressed and seemingly
fleeting. While the information arrives faster, the quality of that information
may be lower and less considered. Increasingly stories are noted as either
Twitterable or not.
Some have also pointed out that the same factors that allow citizen journal-
ism to challenge Big Media and established power structures can be utilised
by these same groups to move Big Media into the internet age. Politicians,
celebrities and large organisations are using blogs and social media in the
same way as citizen journalists, presenting their own perspectives and views
with the same level of apparent openness as any other citizen journalist.
While this may not be a problem in and of itself, it undermines the conten-
tion that digital media is fundamentally empowering – it can also be used to
reinforce, strengthen and deepen existing power structures.
As social media becomes more pervasive, it is also influencing the way
we think about citizen journalism. Following their citizen journalism pro-
ject based around the 2007 Australian election, Bruns et al. noted that
there is a role for sites that facilitate ‘communities of news and content
makers’, which suggests a new form of journalism that they tentatively
refer to as ‘journalism as social networking’ (2009: 205). For them, this
construction helps get around the unfruitful professional-versus-amateur
issue that lies at the heart of much criticism of citizen journalism. They argue
that an emergent ‘networked journalism’ would incorporate both trained jour-
nalists and citizen journalists, enjoying the strengths of both forms of news
gathering and reportage.
ONLINE ACTIVISM
Cyberspace has become a global electronic agora where the diversity of human
disaffection explodes in a cacophony of accents. (Castells 2001: 138)
the key texts are. As Garrett (2006) points out, the literature also comes
at the same issue from different perspectives and fields, further complicating
study. Our goal in this section is not to try to cover the entire gamut of
online activism, because to do so would take us well away from social
media and into other subjects that are only tangentially related. However,
in order to examine the importance of participation, it would be remiss of
us to ignore the importance of online activism and particularly its intersec-
tion with social media. By way of contextualising this section, let us briefly
outline some of the fundamental points about online activism before
going on to examine some examples and how social media has influenced
online debates.
One of the earliest and best-known groups to use the internet for social
activism was the Zapatistas, a revolutionary movement based in the
Chiapas state of Mexico. In late 1993 the Zapatistas occupied a number
of towns in southern Mexico, and gained attention in the West for their use of
the internet as a means to communicate with the rest of the world. Although,
as Turner (2005) points out, the Zapatistas come from an impoverished area
of Mexico, they relied upon non-governmental organisations to place their
hand-written materials onto the internet. In particular, they were supported
by the San Francisco Institute of Global Communication, who Castells
describes as ‘an NGO of socially responsible “techies’’ ’ (2001: 138). They
helped establish an internet network in Mexico called La Neta, which in
turn supported the Zapatistas as well as a number of other activist groups
in Mexico.
While the reality of how the Zapatistas got their messages online may
dispel romantic images of revolutionaries writing emails from satellite-linked
laptops in remote caves or jungle hideouts, it emphasises the growing impact
of the internet for activists – even sometimes those working outside wired
environments – to increase awareness about their cause. For Cleaver, the
Zapatistas’ approach to activism, including their use of the internet, ‘has
inspired and stimulated a wide variety of grassroots political efforts in many
other countries’, a phenomenon which Cleaver refers to as the ‘Zapatista
effect’ (Cleaver 1998). As the quote at the beginning of this section suggests,
Castells sees the internet as becoming a place where social disaffection is
engaged on a global scale.
then moved into the surrounding countries. In 2005, Kony was indicted by the
International Criminal Court for crimes against humanity. The idea
behind the video was to bring popular attention to Kony’s crimes and to
drum up popular support for continued US military support of the Ugandan
government’s attempts to capture Kony.
The video was immediately shared by millions of people via Facebook and
Twitter, and the YouTube video achieved over 40 million hits within the first
few days of its launch, with the number continuing to climb past 84 million.
Critical responses to the video followed, with a number of journalists, aca-
demics, Ugandans and organisations criticising the way the video simplified
a complex problem, and, many claimed, contained misleading facts. The film
was seen by some as a kind of neo-colonialism that depicted Ugandans as
powerless, while others questioned the financial transparency of the Invisible
Children organisation. These criticisms prompted Invisible Children Inc. to
publish a response to the critiques.2
Questions about the finances and backing of online activists, as raised
in the criticisms of Kony 2012, suggest more fundamental issues with the
transparency of online activist organisations. As we identified earlier, social
media allows campaigns to be organised very quickly, and the combination
of affective social networks and well-designed rich media (for example,
video materials) can evoke action (even if it is in the form of making a finan-
cial donation to a cause) before people have had time to properly assess
the cause. This was the case with Kony 2012, and will no doubt be the
case for similar causes into the future. This is not to say that such causes
are not worthy of support, but merely to point to some of the problems that
stem from slick marketing-inspired campaigns that encourage action without
thought.
The Kony 2012 video is a powerful demonstration of how effective social
media can be for groups with the right skills and knowledge to get a message
out to millions of people very quickly and raise the profile of a cause. However,
the immediate controversy around the video and its use of social media raise
some very salient questions about the role of social media in social activ-
ism. In particular, it raises serious questions about the value of social media
in creating a new kind of activism in which participants feel good because
they have taken a stance, but in fact may have done very little, and may
actually do harm. The problems facing central Africa are significant, and
cannot easily be reduced to a 24-minute YouTube video, or a trite one-line
campaign message. Support for the video is not the same thing as support for
a movement.
This kind of activism has pejoratively been described as ‘slacktivism’ to
denote activism that is lazy, half-hearted and generally ‘slack’. Others point
to the emergence of online activist groups who use social media and online
marketing techniques as a way of engaging social media users with political
movements. The word ‘clicktivism’ has been coined to describe ‘the pollu-
tion of activism with the logic of consumerism’ (www.clicktivism.org).
Critics argue that this undermines traditional modes of activism and consti-
tutes a threat to movements that require more substantial engagement from
activists.
In Australia, for example, GetUp was established as an online site that bills
itself as ‘[a]n independent movement to build a progressive Australia and
bring participation back into our democracy’ (www.getup.org.au). The site
allows people to vote for causes that GetUp will back if they achieve sufficient
popularity. Thus, ‘ordinary Australians’ can feel they are engaging politically
and making a difference by signing up to digital petitions or voting for causes
that matter to them. Whether this constitutes true activism or not is open to
debate, but it does suggest the evolution of new avenues for political activism
with unusual topologies (Flew and Wilson 2010).
CONCLUSION
In this chapter we have looked at participation as a central concept that
underlies social media. The internet has always been a two-way medium
that supports the production of digital content by anyone with internet
access. However, social media has enabled and encouraged participation by
making the production, distribution and storage of content less challenging
and, in many cases, all but free. Emergent from this phenomenon is the
user who produces – the phenomenon that Bruns (2005, 2008) has help-
fully described as the ‘produser’. However, it is important to ask how much
‘produsing’ is repurposed for profit by the companies such as Google and
Facebook, and at what point does this repurposing become problematic,
especially when weighed against the potential value of the material being
produced?
The examples of produsage illustrated in this chapter – crowd sourcing,
Wikipedia, citizen journalism and online activism – demonstrate some of
the ways that produsage is impacting on how we engage not only with the
internet but also with society itself. This provides yet another context
for both of our central themes. Participation reinforces the importance
of offline realities in online behaviour. Far from developing new forms of
expression that are disconnected from the real world, people’s online produc-
tive behaviour is strongly anchored in real-world concerns. For Salam Pax,
blogging from inside Iraq during the Gulf War, participative media allowed
his very local and contextualised perspective to transcend the informational
controls of both the Hussein regime and the mainstream Western media
to provide readers from around the world with a strong and personal
connection to something very real that had ramifications well beyond the
digital world.
While produsage – particularly in forms such as online activism and citizen
journalism – appears to be a very empowering use of social media, there are
always issues around exploitation, as we saw in Chapter 3. We must be careful
to bring a critical eye to any claims of revolutionary change. Although citizen
journalism promises empowerment through a more open press that avoids
the bottlenecks and gatekeeping of Big Media, it also raises issues about trans-
parency, trust and quality. Online activism provides exciting opportunities
for democratic participation and change even in places where political dissent
is treated harshly, but it also opens up potentials for the abuse of good intentions.
The participative dimensions of social media are subject to local conditions,
highlighting that what constitutes ‘participation’ is defined by the forces of the
local environment.
NOTES
1 David after the Dentist became a hit on the internet when it was posted in 2009.
See www.youtube.com/watch?v=txqiwrbYGrs.
2 www.invisiblechildren.com.s3-website-us-east-1.amazonaws.com/critiques.html.
CULTURAL PRODUCTION
To begin, it is important to be clear about what we mean by the term ‘cultural
production’. The term ‘culture’ is highly loaded in academic literature and
has been studied to a great degree from many different perspectives. Rather
than trying to present a single unified definition of culture here, we will take
a particular perspective on culture that we draw from the academic traditions
of critical theory and cultural studies. These traditions are particularly
relevant for an engagement with cultural production concerning the media
(and here we include the internet) because the media has often been a focus
of their study.
Critical theory is a branch of sociology that is associated with the Frankfurt
School, a group of leftist scholars who worked in Germany prior to the Second
World War. These scholars, writing in the 1920s and 1930s, included people
like Max Horkheimer, Theodor Adorno and Walter Benjamin, although
the tradition extends past the War and into the 1960s to include others like
Herbert Marcuse and Jürgen Habermas. These scholars wrote on a number
of issues, but their most well-known works concern the relationship between
culture and capitalist society. Writing from the 1930s, before television and
during the heyday of radio and the early days of commercial cinema, the
Frankfurt School scholars were witnessing a great deal of social upheaval
in Germany as Adolf Hitler and the Nazi party began making their move into
power. The importance of radio and its use by the Nazi party for propaganda
was significant at this time. As Jews and intellectuals, many of the Frankfurt
School scholars fled Nazi Germany and moved to the US where they contin-
ued their writings.
Much could (and has) been written about the Frankfurt School, but for
our purposes here, we want to focus on one of the most famous essays that
came out of the Frankfurt School, titled ‘The Culture Industry’, written by
Max Horkheimer and Theodore Adorno in 1944. In this essay, Adorno and
Horkheimer described the culture industries as being akin to heavy, factory-
based industry. Commercial radio, newspapers and cinema were factories
that produced culture for the purposes of manipulation of the masses in the
pursuit of consumerism and/or fascism. They distinguished between high
culture (like fine art, for example), which they felt encouraged thinking and
critical engagement, and low culture, which they saw as base-level enter-
tainment pumped out by the culture industries. This low culture was easy
for people to access and enjoy, but also acted as a kind of tranquiliser that
provided people with distraction and built demand for goods that they didn’t
really need.
This notion of the media as all-powerful cultural industries remained
prominent until the late 1970s. As discussed in Chapter 4, Hall’s seminal
Encoding and Decoding in the Television Discourse (1973), in which he
argued that the meaning in a media text (a TV show or a film, for example)
is not necessarily fixed but is open to active interpretation by the viewer,
started a radical rethinking of this model. Hall’s restructuring of dynamics
between the reader/audience and text can also been seen earlier in the highly
influential work of Roland Barthes, particularly in his concept of the ‘death
of the author’ (1966 [1977]) in which he pronounced the ‘birth of the reader’
as an active participant in the process of making meaning. The encoding/
decoding nexus opened a new chapter on the study of the culture industries,
because although the culture industries could produce culture with an inten-
tion, the reader/viewer/listener was free to interpret that content as they
wished. Later cultural theorists developed a circular model – such as the
circuit of culture (Du Gay et al. 1997) – where the culture industries were
themselves subject to the meanings constructed by audiences. In other words,
through things like TV ratings and audience analysis, the culture industries
modified their production of culture to better suit their target markets. This
general approach to studying cultural production (one in which the audience
plays a crucial role in the construction of meaning) has become known as
‘cultural studies’, and now includes a range of different ideas and theories,
and forms an important scholarly basis for many internet studies.
From this perspective, cultural production is about the way that culture is
produced and reproduced within modern societies. Culture here can be in
the form of media texts (television programmes, podcasts, websites, films) or
cultural objects (iPods, mobile phones and so on). Horkheimer and Adorno
saw cultural production as a mass industry, with factories producing easy-
to-swallow culture to be consumed by the masses in order to keep them
controlled (and consuming). For Hall and cultural studies, the focus has
shifted from production to consumption as the point at which meaning is
made. With this shift in focus also comes a shift in the focus of study: from
the political economy of large organisations and how they make cultural
products, to the uses of cultural products and how people construct meaning
in everyday life.
With the rise of the internet and especially more recently with the rise of
Web 2.0 and participative media, cultural-studies scholars have found
themselves in familiar territory. As discussed in Chapter 4, people are now
producing and distributing of their own original media texts that are in turn
being consumed by other people. The dominance of centralised production
Both museums and galleries are committed to the mystification of the objects
that they display, holding to the fiction of a distinct realm of high art that
stands above the bureaucratised world of work and the complementary vulgar
blandishments of mass culture. (Stallabrass 2010: 7)
The term ‘cultural institution’ has been used to describe a wide range of
organisations that produce or are involved in the production of culture
under a broad definition that includes the cultural industries (such as radio,
television and film), schools and education. Here, though, we are mainly
concerned with public institutions that have traditionally been the arbiters
of culture – museums, galleries and libraries. When we use the term ‘cultural
institution’ in this chapter, we are referring to these organisations. Keep in
mind, however, that many of the points we make about museum, galleries
and libraries also apply to other cultural industries.
Cultural institutions, whether they be galleries, museums or libraries,
have traditionally mediated the relationship between the art/artefact and
the public. Whether public or commercial, the role of the gallery is to act
as a filter, choosing a subset of works and presenting them to the visitor/
viewer/reader. The idea of selection is vitally important here – what gets
selected (and what gets excluded) by an institution makes explicit claims
about the cultural values of the institution, which in turn reflects and poten-
tially reinforces or constructs the culture of a society. Traditionally, art and
culture that is selected and included in this way can be referred to as ‘high
culture’, whereas cultural productions that are common, everyday or folk
art are considered ‘low culture’.
Art institutions have played a key role in normalising tastes and aesthet-
ics associated with ‘art’. Even when the content of art brings traditionally
non-art content or values into the gallery (like avant-garde movements such
as ‘relational aesthetics’), ultimately the role of these spaces in orchestrating
taste cultures only seems to be reinforced. Relational aesthetics is defined
by the French art critic Nicolas Bourriaud as ‘a set of artistic practices
which take as their theoretical and practical point of departure the whole
of human relations and their social context, rather than an independent and
private space’ (Bourriaud 2002: 113). Artists who engage in this kind of art
practice will stage an event in which the experience of the event is the art-
work, rather than objects hanging in a gallery. The point is to displace the
gallery and even the artist from the art. An example is Rirkrit Tiravanija’s
untitled show in 1992 where he invited people into a gallery and cooked
them Thai food. The art here was the environment and experience of that
environment created by the food, the people and the cooking: the art is the
Now the ‘active participant’ that Huhtamo referred to is being recast as the
‘interactive participant’ as various factors come to push for a change in the
way that cultural institutions engage with the public. Chief among these con-
cerns is the ongoing democratisation of cultural institutions that, as we have
described, has been an ongoing concern throughout the twentieth century. The
second is that since the 1980s, as governments started placing more faith in
market forces, cultural institutions have been required to justify their funding
by showing how effectively and efficiently they are engaging with the public.
There is really no way any longer for public museums and libraries to separate
their bricks and mortar business from their virtual business. They have to
integrate with each other, and it is no longer optional to do these things. It is
now required. (Broun 2007)
neatly integrating the online experience into the social network. This is not
just about making the gallery available online; it is also about utilising
social media in order to encourage people to engage with the galleries.
Despite their technological ‘wow’ factor, the value of the virtual gallery
is debatable. While they provide access to the images of the gallery, they
do not necessarily engage the viewer in the same way as a physical space.
For Australian artist Anastasia Klose, the physicality of the gallery space is
important:
YouTube, MySpace and the Internet in general is no substitute for a gallery. But
it can offer a good resource for people wanting to research an artist … The
gallery space is transformative, powerful and singular. Being able to physically
experience an artwork (video or otherwise), i.e. ‘see it’ in all its non-compressed
glory, is paramount. Seeing documentation, or video excerpts online, is no
substitute. (Klose, 2011)
Virtual galleries are, however, only one of the ways through which museums
are utilising social media. Many museums and galleries are using social
media in order to encourage a kind of crowd-sourced grass-roots curatorial
role for users. Here, the idea of curators as experts who can put together a
single unified exhibition of (say) sixteenth-century tapestry is replaced with
a model where the general public are encouraged to build their own path-
ways through a collection. The role of the curator does not disappear, but
does change. In the examples of virtual galleries above, social media plays an
important role.
Many museums are now making their collection databases available
online for public access, and in many cases are providing interactive ways
for people to engage with the collections through the web. The use of folk-
sonomies is one example of how exposing databases on the web allows
people to become more involved with public art institutions – in this case,
encouraging people to become curators. Folksonomies, as the name sug-
gests, are crowd-sourced taxonomies, essentially asking large numbers of
people to associate keywords or ‘tags’ with objects. Over time, and with
enough people tagging objects, new ways of presenting and organising
objects within a collection can be determined through a crowd-sourced
index. This allows people to ‘assert their own connections and associations
between objects in ways that reflect personal perspectives and interests’
(Trant 2006: 85).
The Commons on Flickr, for example, is a collection of images uploaded
by different museums and galleries from around the world. Users can
browse the image collections, which, by virtue of being in the Flickr SNS,
can be tagged and marked, shared by users and even accessed through the
Square Best Buy (a convenience store) where he didn’t buy anything. His
tweets told people what he was doing, thinking and looking at, and for the
24 hours he remained engaged in thoughts about buying. While Bartlett has
attracted a great deal of attention for his use of social media in his perfor-
mance pieces, his work raises questions about the nature of internet art in
the age of social media. For example, how important is interaction with the
audience to the performance, and if it is important, then what is the role of
the artist, and to what extent does the work become a collaboration with
Bartlett as just one of the participants?
An Xiao (2010) is a social media artist who engaged with some of these
questions in a round-table she convened with some other social media artists
in 2009. One of the questions she asked them was, What is social media art?
As a result of the discussion, she defined social media art as consisting of four
‘rules of thumb’. She says that, first, social media art is art where the web
plays a key role in the expression of the art; it is not enough for the work to
be marketed on the web, nor that it is sourced from the web. Second, the art
must involve the audience somehow because social media is a social medium.
Third, she says that the work must be accessible to an audience outside the
art world, but must still be conceptually rich. Fourth, social media art is about
intent – the artist must be able to articulate a reason or purpose of the art-
work that then permits it to be examined and validated by others.
What’s interesting in Xiao’s conceptualisation of social media art is that
a clear notion of a divide between the artist and the audience must exist,
suggesting tensions between the privileged position of the artist as a specialist,
and the untrained. However, as we saw in Chapters 2 and 4, the rise in UCC
practices is having profound effects on traditional modes of knowledge
and cultural production and is blurring the lines between the producer and
consumer. The question is whether artists can maintain their privileged
position, and if not, what does this mean for the symbiosis of art, culture
and technology?
As the internet continues to become more mainstream, the concept of
‘internet art’ as a separate art form may well decline. From some perspec-
tives the Candy Factory (Styled as *CANDY FACTORY PROJECTS), a
Japanese group consisting of collaborations with artists from elsewhere,
encapsulates the spirit of this by deploying collaboration and the visual
economy of repetition often associated with the visuality of the internet.
Much like their frequent collaborative partners Young-Hae Chang Heavy
Industries, whose visual aesthetics seem to be a reminder of early web ani-
mated graphics, the Candy Factory deploys an aesthetic and ideological
position that is a cornerstone of Web 2.0. For the Candy Factory’s Takuji
Kogo, the web-inspired aesthetics of the work bleed into the gallery space,
and the relationship between online art and offline art is blurred:
More recently I often present images as looped and mirrored or still images
animated through close-ups and pan shots. I see them as looped video sculptures
that function both for the exhibition space and online. I have also been
engaged in several collaborative projects working with different subject matter
and material using the same techniques. I’ve been trying to utilize *CANDY
FACTORY PROJECTS as a sort of software which can be used as a platform
of diverse collaborations. (Kogo, 2011)
The Candy Factory both mirrors the composition of new media as well as
adapting it back into the traditional context of art, the gallery. Kogo’s use of
the Candy Factory Project as a platform for collaboration echoes the way
that SNSs are utilised as a platform for sociality, as discussed in Chapter 3. This
adaptation of their work into traditional forms and offline spaces empha-
sises that the Candy Factory artists do not see themselves as internet artists,
but as artists who use the internet. For Kogo, the importance of the internet
needs to be understood in relation to offline considerations:
The internet here is seen mainly as a way to engage with audiences over a
wider geographical area. This is not to diminish the extra affordances that
the internet presents to artists as the various contexts, content and genres
afforded by this medium provide not only bigger audiences but also feed-
back into the collaborative nature of *CANDY FACTORY’S PROJECTS.
The value of using the internet for art, however, is not seen by Kogo as
remarkable or extraordinary. Indeed, Candy Factory artists feel uncomfort-
able with titles such as ‘internet artists’, preferring instead to see themselves
as just ‘artists’. Their use of internet themes and aesthetics reminds us that
creativity and new media are both frequently borrowed from older modes
of visuality as they simultaneously expand into new ones.
Beijing-based artist Cao Fei is another internet artist whose concerns and
intentions are firmly anchored in the offline, but who uses the internet as a
medium for engagement, performance and display. Cao Fei’s art is embedded
within the political reality of modern China and the Chinese government’s
complex and deeply conflicted relationship with the internet. China’s govern-
ment recognises the internet as being important for national development,
but also as presenting a threat to the tight centralised control over the media
and communications that they currently exercise. Thus the internet is not
banned in China, although it is highly regulated. This tension has led to the
implementation of the so-called Great Firewall of China, a technical and
human system designed to block and censor the parts of the internet that
are seen by the government as destabilising. For example, in 2009, with the
twentieth anniversary of the Tiananmen Square massacre and the subse-
quent ethnic riots taking place in China, ‘Western’ media such as Facebook
and Twitter were banned. Alternative social media services – carefully
sanitised versions of SNSs like Facebook (Renren) and YouTube (Douban) –
have been created, supported and controlled by the government. In this
environment, the use of social media tools by artists is often blocked or
hampered.
It is into this internet context that Cao Fei’s work is presented. For her,
the internet is another space for popular cultures in which local and global
images and ideologies are up for reappropriation. Cao Fei’s practice draws
upon the various popular cultural references (hip hop, karaoke, cosplaying
and so on) from Taiwan, Japan, Hong Kong and Hollywood. However, her
perspective is informed by the way in which technologies and the internet have
functioned in China as it shifted from communism to capitalism. Indeed,
the visual culture of the internet in China is one informed by the particular
politics of the local and governmental. She says:
China’s Great Fire Wall, or ‘Great Firewall’, for Internet information control is
escalating. It is due to China’s specific national conditions. What we can get is
all ‘restricted’, limited, and incomplete. As Chinese, we are forced to accept the
reality. Fortunately, we can use Second Life … Currently, the Internet provides
a new public space for exchanging our feelings. The Internet plays a good role
in venting and consoling. Increasingly people hope this personal, intimate space
can permeate the public life. The problem in the current Internet age is, it is a
society under siege and every separated individual needs sympathy in a broader
level. (Cao Fei 2011)
While Cao Fei’s media practice would be perhaps less politically remarkable
in other countries, within the Chinese technoscape her work takes on a
much more politically charged element, something that is highlighted by her
Second Life work, RMB City. Second Life is an online virtual world that is
loosely based on the imaginings of science fiction writers such as William
Gibson and Neal Stephenson and their respective concepts of cyberspace and
the metaverse. Within Second Life, people, represented by a 3-D avatar, can
Within RMB City, Cao Fei takes on the form of her avatar, China Tracy,
who presents us with a pastiche of contemporary Chinese popular cul-
ture. In the virtual space of Second Life, RMB City presents as a playful and
performative world where Pandas mix with MTV references in a space that
mingles the popular with Cao Fei’s own offline life and history. Cao Fei notes
Figure 5.3 Cao Fei/China Tracy, RMB City: A Second Life City Planning (2009)
that as an artist, the internet intensifies and complicates the artist’s relation-
ship with his or her audience:
The power for connection is not only infinite and creative, but also subversive
and destructive. I think that is a paradox. When finishing an artistic creation,
the artist wishes it to be independent and subjective. At the same time, they also
hope to receive public attention and response. The Internet as a medium
provides a good model for interaction. But it depends on how the artist
understands and handles its so-called ‘open borders’. (Cao Fei 2011)
For Cao Fei, there is the dissolution of the barrier between visual art and
new media practice, so that using new media and making art become similar
things. Her position de-privileges the artist, or makes artists of us all:
The world is multiplying and becoming cheaper to access. It is impossible for one
culture to dominate another anymore. This is an information age. During this
period, art can be communicated, copied and connected immediately. So it is
more important that art can maintain its openness and sharing. I have uploaded
a lot of video on to YouTube while many artists still only agree to put their works
in the gallery or keep their works in limited access and not easily reproduced.
RMB is in Flickr. You can find a lot of people to do some recording and sharing.
I think these are the aesthetic characteristics of this era: ‘Communication, sharing,
created by a lot of people.’ (Cao Fei 2011)
deviantART
In a great deal of the discussion within the arts community, art and artists are
seen as a privileged elite, who produce art which is consumed by people – an
audience – in much the same way that mass media produce content for
audiences. As we have seen, for some artists (particularly those engaging
with social media and the internet) the notion of the audience is becoming
unclear as the boundaries between practice and art blur. However, in the
mainstream art world, the clear division between artist and audience remains.
The US National Endowment for the Arts report’s title Audience 2.0 sets the
tone. This report tells us how people – ‘the audience’ – use the internet to
participate in the arts. Participation here is generally not conceived as active
construction of art works. Art is created by artists and presented to audiences
in venues like galleries and theatres (National Endowment for the Arts 2010).
Participation in this context implies watching or listening, but rarely active
creation.
Yet as we have seen, social media is raising questions about the role and
nature of the professional artist as a privileged source of culturally significant
creative production. One of the consequences of the shifts in media paradigms
Almila Akdag Salah (2010) compares deviantART to the Salon des Refusés,
the gallery established by Napoleon III in 1863 for art works that were
refused by French Academy Salons. For Salah, deviantART provides a tan-
gible site that engages with some of the key questions in art today: the
identity of the artist, the role of institutions and the status of collaborative
art. The identity of the artist refers to how an artist is defined within the art
community, both in terms of the kind of artist they are and in terms of their
status as an artist. Salah identifies the ‘identity politics’ movement in the
1980s art world in which any difference (such as, gender and race) resulted
in artists being labelled and pigeon-holed as ‘minority’ artists. This labelling
profited some artists, but for many it operated as a handicap and limited
readings of their work. Thus, an artist who is defined as an Australian Abo-
riginal artist is expected to produce works of art that are consistent with
their aboriginality. This ‘exoticisation’ led to debates around the changing
relationship between art and anthropology (Marcus and Myers 1995)
whereby art critic Hal Foster (1996) proposed the artist as ‘ethnographer’. In
the online environment of deviantART, artists define themselves through a
peer community, and so are freed somewhat from this form of discrimina-
tion and vertical labelling through curators and art buyers.
The second point that Salah makes about identity is perhaps more funda-
mental in the context of this chapter. She notes that identity also applies to
the identity of the artist as an artist. This argument goes to a more funda-
mental question about what it is that defines an artist and the separation
between artist and non-artist, a direct representation of the amateur/profes-
sional division that troubles other professions like journalists, as we saw in
Chapter 4.
In an SNS such as deviantART, it becomes possible to develop quantitative
measures of the relative popularity of individual users by tracking the number
of people visiting their pages, making comments on their works and watching
them. deviantART also regularly highlights the work of individual members
of the community through the ‘daily deviation’, where anyone in the deviantART
community can nominate a work to be featured. The daily deviations are then
selected by deviantART and ‘official’ volunteers to create presentations for
the homepage. These processes provide a way for artists to have their work
recognised, and be promoted by the community.
Traditionally, artists produce works that gain attention through exhibitions
or the forwarding by arbiters of taste like curators, gallery owners or art
buyers. Exhibition curators are responsible for selecting the works that are
to be presented at exhibitions, and thus act as the arbiters of what is and
is not considered art, or at least what is considered high culture, worthy of
display at a gallery or in an exhibition, and what is considered low culture.
Online art sites like deviantART provide new ways for artists and amateur
artists to present their work and have it judged by other users, rather than by
the arts establishment alone. Furthermore, as Salah (2010) points out, if we
judge museums and galleries in terms of visitor numbers to their websites,
we find that deviantART has a much more significant web presence than
some of the leading cultural institutions, including the Museum of Modern
Art, the Metropolitan Museum of Art, the Louvre and the Tate.
What is striking about this function of deviantART is that in providing a
platform for artists to gain recognition, it is providing a role that is very
similar to that traditionally provided by galleries. The major difference here
is that all works are potentially visible, and the selection of works that are
presented is a function of community selection, rather than selection by an
expert such as an exhibition curator. SNSs like deviantART therefore raise
questions not only about the status of the artist, but also about the institu-
tions that make choices about what is and is not considered art.
Dan Perkel argues in his dissertation ‘Making art, creating infrastructure:
deviantART and the production of the web’ (2012) that Web 2.0 isn’t creat-
ing new forms of art or artistic practices, but rather the fabric of the web is
a ‘multi-faceted form of infrastructure’ that forces old tensions in art to col-
lide with Web 2.0 ideals. This then, in turn, produces new tensions. Reflect-
ing upon the changing nature of IP and art theft, Perkel considers ‘two common-
place notions: (1) that a new generation of Internet savvy creators are upend-
ing old ideals about art as property and (2) that the Internet is a medium for
sharing rather than control’ (2012). This leads Perkel to conclude that as well
as creating new tensions between sharing and theft, there is a new balance
between sharing and control. Far from disrupting ‘[r]omantic conceptions of
art and creativity’, Perkel argues that the web uneasily accommodates these
multiple and often conflicting ideologies (2012: 1).
VERNACULAR CREATIVITY
If amateur art questions the boundaries of art, then vernacular creativity
stretches them even further. As we have seen, websites like deviantART sup-
port communities of people who create and share creative works. Outside of
arts-focused SNSs like deviantART is a world of ordinary, everyday creative
production. This everyday creativity includes a range of creative activities
that are rarely commercialised and which are very unlikely ever to appear in
museums. It includes things such as travel photographs, family photographs,
photo album arrangement, needle stitch, quilting and so on – all often prac-
tical or instrumental pursuits, but also undeniably creative.
Jean Burgess (2007) has called this kind of creativity ‘vernacular creativity’,
drawing upon the word ‘vernacular’, which implies language that is ordi-
nary and everyday. This kind of creativity clearly predates modern tech-
nologies by hundreds if not thousands of years, but the rise of social media
means that these forms of everyday, ordinary creativity can be circulated
and shared. Furthermore, the growing ubiquity of devices like mobile cam-
eras and the creative potential of modern computers for video and digital
art have meant that there has been an explosion in the production of ver-
nacular works.
A good example of this kind of everyday vernacular creativity is the pho-
tographic vernacular, which refers to images of the everyday and banal that
started to appear, as photography became more accessible to a mass market.
Photography moved out of the realm of the professional and into the ama-
teur and domestic. This was brought about by the development of Eastman’s
Kodak camera. As Burgess discusses, the Kodak camera made photography
an everyday activity by making it available to everybody. Vernacular photog-
raphy emerged as people began to take images in the course of everyday life,
capturing glimpses into the personal and private. Travel and holiday photos,
family photos and a miscellaneous array of images of people, animals, archi-
tecture and anything else that took a person’s fancy soon began to become
commonplace.
If the Kodak camera made photography an everyday activity, then
the digital camera has democratised the technology, and the advent of the
internet and photo-sharing websites such as Flickr has pushed this even
further. The emergence of the digital camera, especially with its ubiquitous
appearance on mobile phones, suggests a democratisation of media. Those
who previously couldn’t afford cameras or film processing costs suddenly
have access to cheap and convenient alternatives, which has opened up
the world of photography to people who would formerly have not had the
time, technical knowledge and/or money. For example, in Seoul, the ready
availability of digital cameras has seen some women develop a love of
photography that has sometimes ultimately led to them wanting to become
professional photographers (Lee 2005). We discuss some of the key emer
ging visualities around camera-phone images in social and locative media
spaces in Chapter 7.
However, for Burgess, the term ‘vernacular creativity’ embraces two
important concepts: the everyday or ordinary, but also a concept of situated
creativity, because vernacular also implies locality and context dependence.
Burgess cites Batchen (2001), who describes a range of ways in which people
used photographic prints: arrangement in photo albums, as grids hung on
the wall with an implied narrative, using photographs in ornaments, and
the image and give it specific local meanings within a community. This dynamic
environment in which strangers and intimates are sharing images and com-
ments (both aesthetic and technical) suggest that new visualities are
ordered by a ‘situated creativity’ (Burgess 2008) and also new geosocial
cartographies (Hjorth and Gu 2012).
As camera phones become more commonplace with the explosion of
smartphones – along with new contexts for image distribution like micro-
blogging and location-based services (LBSs) – we are witnessing emergent
types of visuality. In particular, through LBSs such as Facebook Places, we see
ways in which users create new contexts for the overlay between place, ambient
images and geographic locations. While, globally, camera-phone genres like
self-portraiture have blossomed, we are also witnessing the flourishing of
vernacular visualities that reflect a localised notion of place-, social- and
identity-making practices (Lee, D-H 2009; Hjorth and Gu 2012). We dis-
cuss this in greater detail in Chapter 7.
CONCLUSION
This chapter has looked at the impact of social media on art as a way of
examining the broader questions about the emerging relationships between
social media and cultural production. As we saw at the beginning of the
chapter, cultural production has come to be understood as something that
involves the consumer as well as the producer. Social media invites new
modes of cultural production where the user pays an even more active role
by actually producing cultural objects rather than just interpreting them.
The result of this is not only that more people can now produce and dissemi-
nate culture, but also that institutions and individuals who have traditionally
been a part of this production process are having to adapt to these new
modes of production.
This can be seen in the way that cultural institutions – galleries and
museums – have begun to focus more heavily on their audiences. While there
are other factors in play that are only tangentially related to the evolution of
the internet, social networks have nonetheless provided a way for museums
and galleries to engage with audiences. Consequently, we are now seeing
people being included in cultural processes that were once solely the domain
of curators and other experts.
For artists, the response to social media has been more mixed. Some artists,
like Man Bartlett, have engaged with social media, working it into their artistic
practice. However, as artists engage with social media and the internet, they
also raise questions about the nature of the artist and his or her relationship
with their audience. Other artists and art collectives, such as the Candy Factory,
prefer not to see themselves as social media or internet artists at all, but are
instead artists who use the internet. For other artists still, such as Cao Fei,
social media and the internet open up spaces for critical analysis that engages
with audiences and again complicates the role of the artist and his or her
audience.
The complexities that these artists are grappling with belie the still-
dominant concept of the artist as a member of a privileged elite. However,
as we saw in the example of deviantART, art and cultural production is not
the domain only of those with the ‘right’ background, but increasingly it is
becoming the domain of people who produce creative works but whose
work would rarely be considered by the art establishment. Instead, sites like
Flickr and deviantART are providing alternative forums for the presentation
and critical assessment of creative works and promoting works based on
community consensus.
This trend is exacerbated by the increasingly important and visible role of
everyday creativity in which vernacular forms of cultural production, such as
mobile-phone imaging, have become part of the structure of social networks.
While not new, vernacular creativity is revealed within social media as playing
an important role in mediating social relationships. It does this through con-
textually nuanced performances that often engage with and adopt the codes
and themes of both popular media and art.
In the final analysis, it seems clear that the shifts we have identified in other
chapters, from mass audience to individual user, and from customisation
to personalisation, are also being felt in the sphere of cultural production.
These changes promise to have significant implications not just for how
we conceptualise and understand art, but also for a whole range of creative
practices and their role in society more generally.
S ocial media games are games that are played within social network sites
(SNSs). These games, with names like Happy Farm and Mafia Wars, are
a familiar and integral part of using SNSs for many people. They are both a
source of entertainment and a way to maintain relationships with friends
and family. Thus, for people who study social media, social media games
become an important part of the overall experience of social media, and
demand attention. For games-studies scholars, the emergence of social media
games (which are generally termed ‘casual games’ within the industry) is also
significant. Social media games are being played by people with different
motivations and different demographics than those who play games on consoles
or those who play conventional computer games, and they are presenting a
new economics of game production, which has seen new players enter the
market. For some, like Juul (2009), the emergence of social media games
constitutes a ‘gamification’ of culture, a casual revolution.
In this chapter, we are going to examine the emergence of social media
games from both a social media and a game-studies perspective. In the first
part of this chapter we look at the idea that games are social, which is at
odds with the popular image of computer games in Western culture. Follow-
ing this we consider social media games more specifically, paying particular
attention to their role within social networking sites and identifying some of
the unique features of these games and the associated implications for
gaming and social media more broadly. As part of this analysis we look at the
way in which social media games provide socially valuable spaces, but also
acknowledge that these spaces are not provided for purely altruistic reasons;
behind the games are SNS and games companies that rely on social gaming
as an important source of revenue. The final part of this chapter presents a
case study that brings the conceptual ideas we have discussed in the chapter
together. The case study focuses on the uses of social media games in China
and the way that the games are mediating relationships between generations
as young people are increasingly moving away from their home and families
for study and work.
in which they interviewed participants and played the games, taking notes
and recording the way that other players were engaging with the game.
In one of the studies, the researchers found a sample of 750 people who did
not play the game Asheron’s Call; they gave copies of the game to one half
of the group, and used the other group as a control. Using these method-
ologies, the researchers were able to gain some insights into the way that
people socialised on MMOGs and on the differences between playing and
not playing. The results of the two studies were convergent (they came to
the same conclusions), both agreeing that MMOGs represented a sig-
nificant social environment. For Steinkuehler and Williams, people’s use
of MMOGs had a social dimension that fits within Oldenburg’s concept
of a ‘third place’ (the social space beyond work and home, discussed fur-
ther in Chapter 2). The studies provide further tangible evidence of the
social nature of games.
Beyond the games themselves, gaming has been a source of sociality as
players build forums to discuss games and develop user created content that
ranges from fan fiction to game ‘mods’ that see players become creators and
actively hack games in order to customise them. Far from being an isolating
or solitary experience, game-playing is often a highly social experience, espe-
cially when games move into online multiplayer environments. If games can
be social, then it might also be fair to say that games are frequently found
alongside social activities, and in this respect SNSs are certainly no excep-
tion. SNS-based games are extremely popular, as we will see below, and have
played an important role in building the subscription bases of social media
as players seek friends to play with.
Yet games have often been disconnected from other forms of analysis
because they are frequently conceptualised, especially in Western culture, as
something that exists outside of everyday life, in a kind of sealed-off space
of their own. This idea of games as being separated from other social activi-
ties reflects early debates about cyberspace as a place that is different from
the offline (we touched on some of these arguments back in Chapter 2).
However, as the internet grew, it became apparent that the exceptionalism
that dominated early writings about the internet hid the fact that the same
inequalities that existed offline were duplicated online (Nakamura 2002).
And, as we saw in Chapter 3, there is also strong evidence to suggest that
offline relationships have a very important bearing on online relationships
amongst users of SNSs.
In the literature on games, the idea that games are separate to other
parts of life was explored as far back as 1938 by Johann Huizinga, who
examined games in his book Homo Ludens (1938 [1970]). Huizinga was
fascinated by play and wanted to understand its function in culture. One of
the concepts described by Huizinga in his book was the idea of the ‘magic
circle’, a concept that was picked up and popularised by Katie Salen and
Eric Zimmerman in their 2003 book Rules of Play. For Huizinga, Salen
and Zimmerman, the magic circle acted as a site for play that was separate and
removed from the real world. Inside the magic circle was the game world which
was associated with play rather than reality. Thus inside the game world you
are completely free to play, to try things out and to fail without fear of there
being any real-world repercussions. There is no sense that the barrier
between the game world and the real world is at all permeable.
Thomas Malaby offers a useful counter position in which he critiques
what he sees as the exceptionalism in much game and play scholarship that
continues to separate play from everyday activity (2007). One of the critical
points that Malaby makes here is the idea that play can be usefully thought
of as a mode of experiencing reality, rather than a separate activity that has
no bearing on the real world. For Malaby, games are a set of processes that
are linked to experience; many non-game activities have playful elements,
just as many games have strong relationships with the real world. Using
examples of both digital and non-digital games, Malaby provides empirical
evidence to show that play is not always separate, safe or necessarily pleas-
urable, and that these notions of play are culturally nuanced and socially
constructed. He also suggests that the division of work and play into sepa-
rate spheres of experience may be a cultural artefact, rather than a universal
distinction. Here Malaby points to work by scholars like Sherry Ortner,
who defines the distinction between work and play/leisure as a ‘modernist
affectation’ that results from the nineteenth-century construction of the idea
of work (Malaby 2007: 8).
Astute readers will already have noted the parallels between the idea of
the magic circle and early constructions of cyberspace in internet theory.
Both have been seen as separate places that are disconnected from the real
world. Yet, just as internet theorists have come to see the importance
of the offline in the online, so too game theorists have begun to question
this rigid separation of games and other aspects of social life. These insights
into the position of games with respect to the other human activities become
more important as we consider social media games which, like SNSs, are
sites for social activities and where online and playful activities intersect
with the social. As we will see below, social media games play an important
role in SNSs, helping to recruit new users, and also providing a way for
SNS users to maintain and develop contacts with their friend networks.
Before delving into social games, however, we need to contextualise the
development of social media games within the broader phenomenon of
casual games.
Casual games
Casual games, as the name suggests, are games that can be enjoyably played
without the high level of attention that is associated with non-casual games.
These are games that do not require the same investment of time, and so
appeal to a broader market than traditional console or PC-based games. The
games are typically said to be easy to learn (but perhaps difficult to master).
However, as Juul (2009) points out, these stereotypical definitions of casual
games often do not withstand close scrutiny.
Many social media games fall into the general category of casual games
by virtue of their design and their mode of play. While many social media
games are casual games, it is not true to say that all casual games are
social media games (many casual games are designed to be played on
smartphones or other mobile devices, and often have little if any integration
with social networks). However, to understand the role and importance of
social media games, it is useful to understand them in the broader context
of casual games.
As noted above, possibly the most distinctive feature of casual games is
their light-attention mode of engagement, or their ‘interruptibility’ as Juul
terms it (2009). Where traditional computer games encourage players to set
aside hours for dedicated play, and reward players accordingly, casual games
allow players to engage with the games for minutes at a time, dipping in and
out of the game as time and interest permit. This different mode of engage-
ment has a number of significant implications for game producers. First,
casual games appeal to a broader range of people who do not have the time
or inclination to play games that require them to put aside hours of dedi-
cated playing time. Second, this low-attention mode of play means that casual
games are well suited to environments in which the user is likely to be engaged
in more than one task. For example, casual games are well suited to mobile
devices, where players can play the games during brief periods of leisure
throughout the day: during lunch breaks, waiting in a queue, or even while they
are working, switching between work tasks and social media as time permits.
For some, social media games allow people to be present in two social places
at once, ushering in forms of ‘presence bleed’ (Gregg 2011) whereby boundaries
between work and leisure blur.
Casual games also provide different incentives for play. As Hou (2011)
points out, games designed for dedicated play focus on mastery of difficult
tasks, demand full attention, and punish players for failure. For example,
MMOGs have traditionally been the preoccupation of ‘hardcore’ gamers,
demanding full attention and the investment of many hours of continuous
gameplay. In MMOGs like World of Warcraft, failure at tasks is often treated
casual games has been around for more than ten years, there was no consist-
ent way for game developers to get their games to large groups of users and
to receive suitable financial compensation in order to make casual game
development an ongoing reality. Furthermore, the wide variety of different
mobile technologies, each with their own nuances and technical limitations,
meant that game development for the mobile market was extremely labour
intensive.
For casual games on mobile devices, the important market development
has been the emergence of smartphones, which are unified by their operating
systems and associated app stores (the big names at the time of writing are
Apple iOS, Google Android, RIM BlackBerry and Microsoft Windows
Mobile). These operating systems, along with their online stores, provide a
means for game developers to produce a game for a single platform (rather
than a hundred different phones) and, significantly, to distribute the game
within a system that allows them to profit from each game sold. For many
game developers (especially smaller companies), the royalties that can be
garnered from selling a game through an app store are significantly higher
than would normally flow back to them if the game were published in a
traditional retail outlet.
Web-based casual games have also come of age as the platforms to support
them have developed. Again, the technical capacity for web browsers to sup-
port games has been around for a long time. A key technology is a software
application from Adobe Inc. called Flash. Flash is a tool that allows the
production of interactive content that can be delivered via a web page,
and for many years Flash provided the only consistent way to develop and
distribute web-based games. Many games have been developed, and advertising
supported web gaming sites like Newgrounds, FlashPortal and Kongregate
encourage Flash game developers to upload their games to the sites. These
Flash games are often supported by advertising, which is facilitated by
the sites (with game makers getting a royalty based on how popular their
games are), or on in-game advertising, which also sees royalties flow to games
based upon a variety of metrics.
stores and operating systems offer benefits for mobile game developers. By
developing a casual game for Facebook, for example, a developer can gain
access to the user’s friends’ lists and add a social dimension to the game. This
allows successful game developers to access many more people than if their
game was placed on a standard Flash game site, thus increasing the total
number of players, which in turn translates to a more profitable game. To
illustrate the relationship between SNSs and game developers, we will examine
the relationship between two of the largest – Facebook and Zynga.
There has been a lot of interest in social media games, driven partly by the
popularity of SNSs, but also because the success of some social media games
has demonstrated that these games can be just as profitable (and in some
cases, more so) than console or PC-based games. The two factors that feed
into this are the lower cost of developing a social media game, and the poten-
tially broader audience of users that the game will attract. One of the major
game production companies, Electronic Arts, acknowledged this when they
purchased the UK-based social game development company Playfish for
US$300 million in 2009 (Playfish 2009).
SNS-based games can collect revenue from advertising, from in-game
purchases or from promotional marketing (Shin and Shin 2011: 854).
In-game purchases are small payments made by players as they play a
game to enhance their utility within that game. For example, a player who
is playing Happy Farm can use real money to purchase virtual goods, such
as seeds for the farm or decorations. To a non-player, the idea of spending
money on virtual goods may seem ridiculous, but this kind of purchasing
represents an important source of revenue for companies like Zynga, who
have said that they derive 90 per cent of their revenue through in-game
purchases of virtual goods. To illustrate the kind of money involved here,
consider that in July 2012 Zynga reported that they had 306 million players
per month and revenue from these users of US$332.5 million, of which
US$291.5 million (close to 90 per cent) came from online games, the remain-
ing amount coming from advertising. So, game and in-game purchases
make up the bulk of Zynga’s income, and although Zynga also sells games
directly on mobile platforms, revenue from Facebook games continues to
make up the lion’s share of Zynga’s profits (Zynga 2012).
While the SNS provides a valuable platform for social media games,
it would be a mistake to think that the relationship between SNSs and social
media games is one-way. Symbiosis would be a more accurate characteris
ation of the relationship between SNSs and SNS-based games. There are a
number of reasons for this. SNS games are a method of recruiting new users
to the SNS. Many games reward players with game bonuses, in-game gifts
and/or currency for recruiting friends. For the committed player, the incentive
to get all their friends and acquaintances involved in the game is significant.
Once a player has exhausted their online friends, there is an incentive to get
even more people involved by encouraging friends and family to join the
SNS. Beyond the recruitment of new users, and perhaps more significantly
for established SNSs with already large user bases, games also provide a
revenue stream.
The importance of the game revenue stream is highlighted by the symbiotic
relationship that has developed between Zynga and Facebook. Zynga, as
mentioned above, is one of the most successful SNS game developers. In
2011 when Facebook made its initial public offering (meaning it would list
on the stock market so people could buy shares in it), it revealed that Zynga
accounted for 12 per cent of Facebook’s revenue (Geron 2012). Zynga and
Facebook had entered into a five-year partnership in 2010 in which Facebook
guaranteed Zynga a steady increase in the number of users, and Zynga guaran-
teed Facebook that it would make its web games exclusively for Facebook.
Another indication of the importance of games (and Facebook apps gen-
erally) for Facebook can be seen in the deployment of Facebook Credits.
Facebook Credits is a system designed to allow Facebook users to pay for
content within Facebook. The idea is that you use a payment system, like a
credit card or PayPal, to buy Facebook credits. Then, if you’re playing a
game like FarmVille, instead of paying for in-game purchases with your
credit card, you use some of your Facebook credits. This system has one
major advantage for Facebook: it allows Facebook to take a cut of all in-game
purchases. The amount that Facebook takes is 30 per cent, so if someone buys
a virtual chicken coop for US$3.00, the game-maker gets US$2.10 and
Facebook gets 90 cents. Multiply this by every transaction on Facebook
and it can add up to a lot of money. Even if Facebook decides to phase this
system out, it will probably continue some kind of ‘pay by Facebook’ scheme,
where users can opt to make online payments with a Facebook account
rather than a credit card.
While the examples we have presented here have focused on Facebook
and Zynga, they are provided here only as examples to illustrate the kind
of business that is going on behind the scenes of social media games. In the
future perhaps neither of these companies will exist, having ridden a wave
of investor enthusiasm that ended up being little more than another dot-
com-style bubble. They might both exist in a different form; the future is a
dangerous thing to predict. The key point to take from this is that the strategies
that underlie Facebook and Zynga’s approach to the online environment are
representative of the business of social media games. From the above, it is fair
to say that to Facebook and Zynga, the value of social media games is that
they provide another way for SNSs to commodify users’ sociality. As discussed
in Chapter 3, SNSs are businesses, and thus need to make money. They do
this by deriving profit from their users, primarily though analysis of user’s
submitted data, and selling highly targeted consumers to advertisers. Here,
SNS games provide a kind of platform-within-a-platform for deriving profit
from the sociality of online games.
To this point, we have focused on the economic aspects of social media
games. This is important because it outlines the environment in which social
media games exist. Understanding the economics of social media games gives
us important insights into how the games are designed. For example, once we
know that the major source of revenue for Zynga is the purchase of virtual
goods, it becomes clear that the economic imperative for the designers of the
game is to make virtual goods a central design feature of the game.
does not necessarily always translate to social media games. MMOGs such
as Everquest and World of Warcraft are games that are hardly casual. In
fact, these games are often associated with time-consuming attention-
demanding play that rewards players for time investment and, in the case
of World of Warcraft at least, virtually requires that players form groups
at higher levels to continue playing the game and progressing. Casual
games, by comparison, do not demand the same kind of time investment,
and so it seems reasonable to hypothesise that a different level of engage-
ment would yield different kinds of social interaction. It is also possible that
games that are designed to operate within an SNS provide different environ-
ments which may influence the way people socialise within them.
One approach that begins to develop insights into these questions
compares games that are designed to work within an SNS with stand-
alone games to determine what (if any) effect social networking has upon the
structure of games. Kirman et al. (2009) conducted one such study and
have suggested that one of the reasons for the popularity and rapid growth of
SNS games when compared to stand-alone games is because of the intimate
nature of the game request. For a stand-alone game to pick up new players,
it needs to recommend itself either through advertising or word of mouth.
For social games on SNSs, the word-of-mouth method becomes very power-
ful not only because it is easy, but also because the recommendation to
play the game inevitably comes from someone in your social network. In
other words, if a good friend invites you to play a game with them you are
more likely to say yes.
Other studies have analysed the social networks constructed by social
media games, or have asked questions of the players directly. In her study of
social media games, Hou (2011) found that players do not seem to be motivated
by challenge and competition to the same degree that players of conventional
games are, but instead play social media games for relaxation and diversion.
This, she suggested, reflects the casual nature of social media games. Specifi-
cally, they do not require extreme expenditure of time or attention and do not
punish failure. Importantly, however, Hou also found that the social dimen-
sion of social media games was important for players. According to Hou,
‘respondents played social games more frequently, spent more time on the
game, and got more engaged in game activities for the purpose of social inter-
action’ (Hou 2011). Hou’s research suggested that social interaction was the
most significant reason for playing social media games. This is somewhat at
odds with the ISG (2010) report mentioned above which found that only
around 25 per cent of their respondents identified social interaction as the
major reason for playing social media games. Respondents to that study
nominated relaxation and enjoyment as their primary motivation.
It is not clear why there is a disparity between the ISG study and Hou’s,
but the methods used for analysis, including the way in which the respond-
ents were selected, may play a role. One of the key issues here is that there
is a tendency in some research to generalise the player population. Players
are not one homogeneous group, but are diverse, and different people play
different games in different ways. Not all players of social media games
approach them with the same casual attitude, and some social media games
promote different kinds of game play than others. Some players do invest
serious amounts of time and resources in social media games, and could be
termed ‘hardcore’ players, even though they play casual games. Referring
again to the ISG (2010) report, 12 per cent of US and UK social media game
players play their games for more than ten hours a week, and 9 per cent play
social media games for more than three hours at a sitting. It is reasonable to
hypothesise that the experience and motivations of social media games for
these hardcore players is somewhat different to the average player who,
according to the ISG report, plays social media games for one to five hours
a week. The more hardcore players can also have a stronger influence on the
game than more casual players. That is, a small number of players can be
responsible for a very large proportion of interactions within a game (Kirman
et al. 2009). These players, while in a minority, may be responsible for a dis-
proportionate amount of apparent social activity within the game.
If we consider that different players may play social media games differ-
ently, then it is just as important to consider how different kinds of social
media games can affect player sociality. Rossi (2009) suggests that social
media games can be grouped into two broad categories based upon how the
user’s friend network is utilised by the game’s design. The first type is ‘skill/
knowledge’ games, which essentially challenge the player to accomplish a
given task or answer a question. In these games, SNS friends are fellow
competitors, to be played against and/or ranked against in a score table. The
second kind of social media game identified by Rossi are managerial or, as
Rossi terms them, ‘truly social’ games, in which the object of the game is not
to win so much as to build and develop a virtual space populated with virtual
goods. This would include games such as Happy Farm, FarmVille and Pet
Society. In these games, the goal is to maintain and perhaps develop your
virtual space through regular maintenance.
These different kinds of game would tend to suggest different modes of
play and different ways to involve friends. The skill/knowledge games require
more attention over shorter periods of time (depending on the game). Friends
in these games are competitors, and are more likely to be drawn from a closer
group of people you know, people with whom you have strong ties in a social
network. Truly social games are played over long periods, often having no
defined end-point or winning strategy, and require the user to dip in for
relatively short periods to perform basic maintenance tasks. Friends in these
games take on a new dimension as the game rewards players with virtual
goods for recruiting new players and sharing or helping out other players.
In these games, friends become game resources where the more friends you
have connected with, the more agency you have in-game. Even with such
broad categories of social media games, and rough/confusing delineations
of users into ‘casual’ and ‘hardcore’, it becomes clear that there is no one
way to play social media games, and that the levels and kinds of social
interaction are going to vary from user to user and game to game.
It is perhaps not surprising that some have argued that social media games
may not be as social as they may seem. While a game like Happy Farm
appears ostensibly to be a hot-bed of social activity as users increase the size
of their social networks, more often than not this behaviour is purely
instrumental. As Shin and Shin put it, ‘[i]n social games friends are not
really friends, they are mere resources’ (2011: 853). In other words, the
game rewards players for inviting new players, and so players comply, but
this does not necessarily lead to those players going on to be sociable in that
environment. For Rossi (2009) this then leads to the idea that there can be
at least two kinds of friends in an SNS gamer’s network: real friends (that
is, people who they socialise with), and instrumental friends (who are on the
user’s friend list within the SNS, but who might otherwise have only a very
weak relationship with that user).
This raises an important point about sociality within the social media
game. The game can act as a kind of catalyst for other social behaviours, and
may spawn further kinds of sociality and creative production from players.
Rao (2008) has previously argued that social games are inherently social,
and may constitute a third place for sociality (in much the same way that
Steinkuehler and Williams argued that MMOGs construct a third place).
For Rao, social media games can foster relationships by providing a forum
for socialisation that is separated from the mainstream way that this is
performed in an SNS. An example might be a mother and daughter who play
an online game together regularly. The game provides a scene for online socia-
bility that probably would not have existed in the user’s conventional SNS
activities, which is reserved for other kinds of (non-family) contact.
The remainder of this chapter is dedicated to a case study that examines
some ways that social media games are played in China. In this case study
Happy Farm (an adaptation of FarmVille) features prominently, although
other games are also mentioned. Many of the themes discussed above are
played out in the case study, demonstrating that social media games do play
an important and sometimes surprising role in online social interaction.
In particular, the case study provides qualitative insights into how social
media games are being used socially to bridge social gaps that have opened
up in a country that is experiencing rapid development and social change.
This serves two purposes. First, it highlights how some of the conceptual
ideas about social media games work in practice; and second, it demonstrates
that social media games are not just Anglophonic phenomena, and are not
just used by the young.
Cultures … [do] not hold still for their portraits. (Clifford and Marcus
1986: 10)
The rise of social games such as Happy Farm, played through SNSs such as
Renren and Kaixin, have seen millions of people young and old participating
in their communities of practice. According to iResearch, a consulting group
specialising in internet research, around 50 per cent of the 26 million daily
users of one of the main SNSs, renren.com, play online games (Cheng 2010).
These games generate around half of the website’s annual income. Whilst the
number of traditional network games (like MMOGs) has remained relatively
constant at 50 million, SNS users have burgeoned from nothing to tens of
millions in a few years (Cheng 2010). This phenomenon, whilst highly social,
also demonstrates changing attitudes to both the online world and gaming.
The success of Happy Farm in China may have something to do with the
context of Chinese society more generally (Hjorth and Arnold 2012). China
is going through a process of rapid development and industrialisation and is
transitioning from a communist to a capitalist economy. The communist
underpinnings of Chinese society are still very much a part of the educational
system, and all educated Chinese children graduate from school with a firm
understanding of Marxist and Maoist philosophy. Happy Farm allows people
to play with the basic economic model of capitalism, replete with the morally
questionable practices of theft to increase one’s personal wealth. In addition
to this socio-economic backdrop, rapidly increasing real-estate prices in
major population centres like Shanghai and Beijing means that for many, the
thought of owning one’s own home is an elusive dream. Against this back-
drop Happy Farm is a nostalgic fantasy that rewards hard work (represented
in the game by hours online) with a successful farm.
Happy Farm is in many ways the epitome of the casual social game, and
the central game mechanic is a simplified model of capitalism. The object of
the game is to start with a very basic farm with few resources and to develop
the farm over time by acquiring, raising and selling produce for a profit. One
of the key – albeit subversive – factors of this game is to steal other people’s
produce when they are offline. While this is a slightly damning condem
nation of the morals (or lack thereof) of capitalism, it serves an important
function for the game in keeping players engaged.
Many Happy Farm players keep the game open on their desktop whilst
doing other activities (such as work) to avoid being robbed. Some have
been known to set their alarms for late in the night so they can go online
when everyone is asleep in order to steal. Interaction with friends is built
into the mechanics of the game through a number of features. Players can
gain experience points by helping their friends develop their farms, and are
rewarded with in-game gifts (that help players improve their farms) for
encouraging their friends to play the game. Hou (2011), although acknowl-
edging the limitations of her study and warning against generalising her
results, found that the elements of social interaction in Happy Farm
encouraged people to play the games more frequently for longer and with
higher levels of engagement.
While the casual game connotes a less attention-intensive mode of game-
play, it would be wrong to think that this translates to less engagement.
Unlike their ‘hardcore’ predecessors, SNS games are much more casual in
their demands for engagement. However, behind this casualness is a play
architecture that is often just as time-consuming – but just in the form of
distracted micro-engagement. Users often have the game open on their desk-
top behind other screens like word documents, email, instant messaging
and so on whilst their virtual plants mature and grow. Leave the vegetables
too long and they will wither and rot, and weeds will grow, so the pace of the
game is slower and requires less intense attention than hardcore games, but
it still requires attention.
Interestingly, the demographic of players migrating to these games don’t
consist of the obvious demographic of the young student. Instead, it seems
that the fastest growing demographic is parents and even grandparents. They
are often being taught how to use the internet by their children, who are living
away from home for study or work. This cross-generational new media literacy
emerging in China’s increasingly mobile population (that is, migrating to cities
like Shanghai for work or study) sees social media such as QQ (the largest
and longest-running SNS in China) and online games as helping to alleviate
the negative effects of cross-generational class mobility by maintaining kinship
relations.
In China, the generation of children born in the post-1980s period are
called the ba ling hou (literally, the ‘after 1980s’: ba ling is 1980 and hou is
after). The term has been used in a similar way and with similar connotations
to the use of the Anglophonic term ‘generation Y’. Children from the ba
ling hou generation grew up in a rapidly modernising, post-cultural revolu-
tion, one-child family China. As a result, these children are characterised
as spoiled and self-centred but also technologically adept and more open-
minded than older generations. Reflecting the rapid urbanisation of China,
many people from this generation move to find work or take up study in
urban centres, so it’s not unusual to find ba ling hou living away from their
familial support structures. Like all generational generalisations, of course,
the idea of the ba ling hou is more an embodiment of social anxieties and
hopes about the present and the future than it is an accurate description of
real people, so we can see that the characterisation of the ba ling hou repre-
sents a generation at the forefront of significant social and economic change
in China.
Interviews with young adult users of Chinese social games provides us
with some valuable and surprising insights into the ways that the issues sur-
rounding the uses of social games and inter-generational politics are being
played out in China. In a study conducted by Hjorth and Arnold (2012),
they saw that it was a desire to maintain inter-generational connection as
students left home for work or study that encouraged communication via
social media and especially social media games.
Maintaining these relationships requires work that, for a time-poor student,
was sometimes regarded as laborious. For example, one female respondent
aged 25 played online games such as mahjong (an online version of the
traditional Chinese board game) because her mother liked to play them
with her. She said, ‘The more I play, the happier my mother is. I like to make
her happy.’ Another female respondent aged 20 complained that she believed
her father to be addicted to playing games. She said, ‘He has so much time
on his hands, he just wastes it on gaming. Our generation don’t have time.’
This is an interesting role reversal in the stereotypes associated with age and
media practice. Rather than children being accused of wasting their time
with new media, instead parents are often called out for excessive online
media use.
Another female respondent (aged 19) said her parents were still learning
to play games and use social media like QQ. Her parents were retired and
they now had time for the student to teach them new media skills. This
education was not always a great success:
They have a lot of time to stay at home, so they will play the computer games
and want to surf online. But my father and mother are not good at it yet, so
I continue to teach them and with the help of QQ, I can contact them more
often. For example, when I come back home, I find that my father’s mobile
phone has something wrong – it always happens and he can’t receive my short
message. I said he’s a little old for it. He has played games in QQ and also,
Happy Farm. He liked stealing vegetables. But my uncle is more of a social
media user. Several years ago, we taught him how to use the internet – how to
connect and talk with others by QQ or something like that. And now he uses
it all the time. He even makes friends with strangers. And so, every time I come
back to my home and we can talk a lot about QQ and games. I don’t know if
the technology is a very good thing for him. I don’t know, because I think
maybe he has spent a lot of time on this new technology, maybe too much.
Personally, I have no doubt he has a very, very young heart from his technology
use. And I think he is enjoying his life very much. (Hjorth and Arnold 2012)
This respondent invested time and energy in getting her family online and
bringing them up to a level where she could include them in the social
activities, like playing Happy Farm. While her parents seemed to be slow in
the uptake of the technology, her uncle took to it with alacrity, which the
respondent found a little surprising. Like the previous respondent who dis-
approved of the amount of time her father spent online, this respondent
disapproved of her uncle’s use and noted with great surprise how willing
her uncle seemed to play with strangers online. She viewed her uncle’s atti-
tude as demonstrative of a type of youth or youthful attitude, or what could
be dubbed a type of ‘kidults’ (adults adopting kid-type attitudes to lifestyle
objects like new media). Here we see that knowledge of new media is
equated with a type of youthfulness, but not knowing the often-tacit limits
of media practice seemed to be the faux pas for older users (according to ba
ling hou respondents).
Another respondent, this time a 23-year-old female, also indicated that
there was effort required to maintain these social connections, effort that was
sometimes more than she was willing to expend. This respondent liked play-
ing a variety of games that reflected two worlds – the casual and the hardcore.
She liked both casual games like Happy Farm and hardcore games like World
of Warcraft. However, she found that she has too many friends on her SNS,
Renren, which means that once she starts playing one game she gets caught
up and finds it hard to leave and get work done. So now she mainly plays the
online game on the local university server. As she notes:
I liked playing Happy Farm very much. I think the plants I planted are very
beautiful and I like to keep my farm looking neat. I also like visiting other
people’s farms and stealing some plants and flowers from friends. Often I’m
playing in the same [physical] space as my friends. It adds to the enjoyment of
the game … I think it is very good for socialising, but then I became too busy
and it was hard for me to stop playing when friends were logging on. I now use
a local game developed at Fudan University. It is good because I just play that
one with my friends and roommates at the university. We play when we have
all finished or need a break studying. (Hjorth and Arnold 2012)
For her, playing online social games is a social activity, and one that’s sometimes
carried out with her friends in the same physical space. But the intensity of
this activity became too much, and began impinging upon other aspects
of her life, so she moved her activities into another space where she has less
social connections to maintain. For this respondent, the labour associated
with maintaining too many connections was too great to justify continuing
to play the game.
For the ba ling hou, social games are not only about play, they are also
about the work of maintaining relationships. This is a full-time intimate
labour that requires effort for the user to maintain. Like playbour (forms
of labour carried out in and around computer games), and the work that
modders (those who modify games’ software) do, this labour is essential to
the economics of social games and social network sites. As Kücklich’s
(2005) example of the modder illustrates, the modder has to call upon their
various forms of labour (social, creative, affective, emotional) that reflect
their community and the associated modes of knowledge. This process is an
integral part of the gaming communities – a labour of love that is supported
and then turned into profit by the industry (Andrejevic 2011). Media such
as SNSs and social games can be seen to operate to exploit a type of full-time
intimacy in which work and life boundaries continue to blur (Wacjman
et al. 2009).
In this case study we have seen how SNS games amplify the local. Through
the case study of Chinese social media games we see the growing tensions
around forms of socio-economic mobility for the ba ling hou as they try to
be both at home and away; a process in which intimacies are negotiated
across private and public, work and life spheres. The constant contact
afforded by social and mobile media means that users can be operating across
various forms of co-presence. In SNS games like Happy Farm, we see new
forms of cross-generational media practices that are indicative, and sympto-
matic, of China’s transformations.
CONCLUSION
Games are social activities, and it is no accident that some of the earliest as
a well as some of the most current forms of online sociality take place within
the framework of games. While there is a tendency to regard games as some-
thing that is separate to real life, games must, as Malaby (2007) points out,
be understood as a mode of experiencing reality, one that is no less valid than
other modes. So, when we come to look at social media, it is hard to ignore
the great degree to which games have found their niche. As we have seen,
social games – as games that are played within the constructs of SNSs –
have attracted great numbers of players, brought new subscribers to SNS
sites and have opened up both gaming and social media to demographics
that were not otherwise engaged with them.
Social games are social because of the fundamental role that social
engagement plays in them. All social games, in some way or another, utilise
people’s friends, or the potential for people to make new connections in the
online environment. This makes these games more enjoyable, and provides
playful ways for people to socialise in online environments. Playing social
games requires players to engage with the game in different ways that are
often more socially oriented, and thus embedded in offline contexts. This,
in turn, has given birth to new forms of player practices. These games are
no longer about hardcore, subcultural practice, but are casual, in that they
require less intensive attention than hardcore games – they are interruptible
(Juul 2009) – and have various mechanics that make them appeal to broader
audiences.
This is not to say that all social media games are the same, or that all
players interact with these games in the same way. While some users play
social media games competitively with their friends, others tend virtual
gardens for hours a week, this time accumulated from many small moments
snatched throughout the day. Some only play games with friends, and others
play games with family as a way to maintain contact. Still others collect SNS
friends (who may not be actual friends, but friends only as defined by the SNS)
as resources that give them greater utility within their virtual space. The uses
of social media games are almost as varied as the people who use them, but as
we saw in the case study, there are familiar motivations even across language
and cultural boundaries. Social media games are playing a central role in
the ways that people socialise within SNSs.
While Hyunjin waited for Soohyun in a café in Shinchon, South Korea, she
toyed with her iPhone. Having downloaded some of the numerous photo apps,
she began to experiment. Finally she was happy with the Hipstamatic lens that
made her coffee look like it was out of some old analogue photo shoot. She then
quickly uploaded it, along with the caption ‘Waiting’, to a few social media sites
with location-based services (LBSs) like Facebook Places and Cyworld mini-
hompy. While the relevance of the caption might be lost on many of her friends,
for Soohyun it served as a reminder that she was keeping her friend waiting as
she dashed from the train station. Another friend of Hyunjin, Joon, was in the
area when she saw her friend’s photo message and her location via her phone.
She quickly made a detour to the café and sneaked up behind Hyunjin. Both
girls laughed and shared a coffee while Soohyun raced to get to the café in busy
peak hour traffic.
Toshi had never understood why people used the LBS game Foursquare until
after the earthquake, tsunami and nuclear disaster of March 2011 in Tohoku
(called ‘3/11’). During 3/11 he tried in vain to make contact with his parents,
finally making contact hours later and after much distress. Access to social
network sites like Mixi was difficult with mobile phone (keitai) networks jam-
ming. Moreover, with loss of electricity, Toshi found himself in an unfamiliar
situation – no working keitai. So when some sort of normality was established
after 3/11, Toshi began to constantly use Foursquare not to play the game but
as a way to give his friends and family a way to always know where he was.
Over in Shanghai, a mother, Jia, misses her daughter, Yuewen, who has just
moved away for university work placement. Thoughtfully, before she left,
Yuewen installed the most up-to-date software on her mum’s pirated smart-
phone (shanzhai). As instructed by her mum, Yuewen installed all the most
popular games, especially social media games like the farm simulation Happy
Farm, for even though the game had its heyday in 2009 it is still played by mil-
lions today. So Jia logs onto China’s version of Facebook, Renren, and signs
onto Happy Farm. She sees Yuewen is not online and so not protecting her
crops. So Jia uses the most popular instant messaging (IM) service, QQ, to send
a message to Yuewen to see if she has time for a quick play in Happy Farm. For
Yuewen, typical of China’s Generation Y (ba ling hou), LBS mobile games like
Jiepang, along with microblog Weibo, are her favourite media. However, when
she reads her mum’s message, she quickly signs into Happy Farm to reconnect
and catch up on gossip with her mum.
Grandmother Amy had never been a technology kind of person. She grew up in
an age when technology was a male domain. On the other hand, her husband
was a passionate adopter of new technologies and always had the newest Apple
Mac item on hand. But one day, Amy’s husband brought home an iPad with
some of her favourite books uploaded. The next minute, Amy was hooked.
Before long she had numerous games and social media apps downloaded,
thanks to her enthusiastic grandchildren. Before long, Amy, like so many grand-
parents in Melbourne, had joined Facebook. And with that membership, her
relationship with her grandchildren took on new forms of intimacy. While Amy
was a little unsure about the ways in which her grandchildren seemed to impul-
sively upload pictures and make comments, she enjoyed seeing this other side
to them and their friends – their ‘mobile publics’ if you will.
These examples are but a few of the millions of intimate vignettes that describe
the ways in which mobile social media is being deployed. Across numerous
technical platforms, personal and cultural contexts, and through a wide vari-
ety of social media, people young and old are using social, locative and mobile
media to rehearse earlier forms of ritual and, at the same time, create new
forms of intimacy and different contexts for the expression of intimacy. While
locations like Seoul and Tokyo have long been centres of innovation in the
invention and popularisation of mobile media, the relationship between per-
sonal, social, locative and mobile media is quotidian and, for the most part,
tacit in its familiarity. In other locations like Australia or the US, convergence
in the form of smartphones is nascent. Mobile social media is a global phe-
nomenon, but also local at every point (Hjorth and Arnold 2013).
Having explored the convergence of social network sites and games in
the last chapter, this chapter looks at the way in which mobile media has
increasingly become the key portal for social and locative media as smart-
phones increasingly, and unevenly, are adopted around the world. Devices
such as the iPhone have become synonymous with this media evolution.
Through this growth, we have witnessed a shift from the device being analysed
as communication medium to being understood as a networked media tool
in which social media, games and various forms of everyday creativity can
be found (Hjorth et al. 2012). In the short few decades in which mobile
phones have been readily available, this technology has changed from being
a mere extension of the landline to being a sophisticated and convergent
online mobile media portal (Goggin and Hjorth 2009). For many, mobile
media is the key device and context for online and social media, with loca-
tions like Japan having more than a decade of mainstream mobile internet.
In China, three-quarters of its 485 million online users (318 million) access
the internet via mobile media (CNNIC 2011). In locations like Singapore
and Melbourne (Australia), the rapid adoption of smartphones has seen a
rise in cross-generational social media usage – much to the disgust of some
younger users who fear their parents won’t understand much of the tacit
etiquette.
Over the few decades of mobile phone use, the analysis of mobile com-
munication has rapidly expanded to encompass many disciplinary and inter-
disciplinary approaches such as sociology, anthropology, internet studies,
games studies and new media studies. While not everyone has a computer,
the ubiquity of mobile phones in places like China and India have provided
many with access to a version of the online. But while mobile media is a
global phenomenon, it is also reflective of the local at every level. The growth
in mobile media has served to amplify the various complex dimensions of
locality, rather than eroding the importance of place. This interdisciplinary
context has afforded the study of mobile communication a rich history of
mixed methods and conceptual paradigms. But it also means that studying
mobile communication has become the battleground for some of the various
disciplines’ boundaries.
All over the globe, location-based services such as the global positioning
system (GPS), geotagging and Google Maps have become a pervasive part of
everyday life through platforms and devices such as smartphones, Android
devices, tablets and portable gaming devices. Moving beyond printed maps,
mobile digital devices now frame and mediate our ability to traverse, expe-
rience, share and conceptualise place. This shift appears to have a range of
consequences for our relationships to place, intimacy, privacy, time and pres-
ence. Locative media shapes, and is shaped by, a variety of factors such as
culture, age and temporal differences.
In this chapter we reflect upon the journey of the mobile phone as it has
unevenly developed into social, locative and mobile media. In particular we
look at the development of LBSs, often integrated with social media, that
have converged with mobile technologies in the smartphone. Acknowledging
the increasing use of these converged devices, we look at how mobility has
become about more than the ability to take your social media with you.
Specifically, there are two results of media mobility: the expansion of car-
tographies enabled by LBS devices and mobile apps, and the development
of location-based social apps that blend social relationships with geography.
These changes reflect broader shifts in the relationships between identity,
place and community and raise important issues about privacy, but also how
we narrate and attach meaning to place. In the next section we reflect upon
the nature of the convergence of mobile, social and locative media. This is then
followed by a discussion of the changing role that camera phones play in
our understandings and visualisations of place, especially as they become
entangled in locative media practices.
media; it houses many media and apps that are a hybrid of new and old media.
Along with housing remediation, mobile media also operates as a vehicle for
‘presence bleed’ – a term Gregg uses to describe the bleeding of one’s personal
presence across platforms, contexts and media (2011). We have previously
examined this term in Chapter 6 as a part of our discussion of the ‘always
on’ nature of social media and casual games.
Along with the changes in the technology have come changes in the profile
of the people who use the technology. Associations of the mobile phone with
ostentatious wealth that accompanied the first generation of mobile phones
in the Anglophonic world are now almost completely forgotten as cheap
phones are plentiful and coverage is wide, even on a global scale. While
for many of the world’s poor countries, mobile phones are 2G (second gen-
eration, that is, non-internet), according to the International Telecom-
munication Union (ITU) (ITU 2011), this is rapidly changing as markets for
pirated smartphones have seen phenomenal growth. Some statistics suggest
that as of March 2012 China’s mobile phone subscriptions topped 1 billion
(CNNIC 2012).
The increasing use of mobile technologies worldwide is also seeing the
growth in the use of mobile internet, which is enabled in part through
the increased accessibility of smartphones. With prices being driven down
by demand and competition in some of the most populous countries in the
world, smartphones are becoming increasingly accessible even in develop-
ing countries. Cheap smartphones and shanzhai (imitations of brand-name
phones like Apple’s iPhone) are now readily available, and prices keep falling
as companies develop low-cost phones for the burgeoning markets of China
and India.
Until recently in many countries – especially the Anglophonic world –
the evolution of the mobile phone has appeared in the form of a transition
from a mobile communication technology to a mobile multimedia technol-
ogy. But, as social media becomes mobile, it is also beginning to converge
with other technologies, such as LBSs, that continue to redefine the uses of
the mobile device. LBSs utilise various features in smartphones (including
GPS, and various methods of triangulating position based on proximity
to cell-phone towers and wireless networks) in order to determine the
location of the user in geographic space. LBSs in current smartphones allow
the phone to determine its position to an accuracy of within 100 metres,
usually much less.
Although LBSs have been available in mobile devices since the early
1990s, it is only fairly recently that LBSs have become a feature of smartphones,
and so have started to become available to people who would not otherwise
have gone out to purchase a separate device such as a GPS unit. While
locative media, like the Internet, has its history in the military, GPS was
quickly adapted for commercial use. However, locative-based mobile games
were developed from an experimental and creative context (de Souza e Silva
and Hjorth 2009) that has later taken the form of commercial games such as
Foursquare.
The transition of LBSs can be thought of in terms of generations. The first
generation of LBSs were available through custom devices which provided
a single-use device, often seen in countries such as the US and Australia only
in higher-end motor vehicles. The use of first-generation LBSs saw some
innovative experiments with play, but was constrained largely to experi-
mental uses by early adopters. Second-generation LBSs have emerged as
GPS and GPS-like services that are embedded in consumer devices as just
one of many features on those devices. With the more general accessibility
of LBSs, the experimental uses of the technology have been commodified
and are moving beyond gaming and into other applications. The most
immediate impact of these second-generation LBSs for users of smartphones
is through services like Google Maps, where an interactive map can pinpoint
a user’s location and calculate the fastest route to almost any destination.
Although the navigational capabilities this affords are important, the feature
only represents a fraction of the implications of LBSs, particularly when
they converge with networked media. It should also be noted that while this
convergence of mobile, locative and social media is quite new in some coun-
tries (particularly in the Anglophonic world), in other countries, like Japan,
the mobile phone (keitai) has been associated with social and locative media
for over a decade (Hjorth 2003; Ito 2005). Now, with the increasingly wide-
spread use of smartphones and the convergence of mobile, social and locative
technologies in these devices, the implications of convergence are being seen
in many places.
For obvious reasons, place has always played an important role in mobile
media (Ito 2002; Hjorth 2005), but the ramifications of these changes go
beyond the immediate promise of being able to access the internet and
social media anywhere, anytime, or of mapping one’s position in the world
(as useful as that may be). Mobile media highlights the various, often-tacit
notions of place as something that is lived and imagined, psychological and
geographic (Hjorth 2012). As mobile media converges with social and loca-
tive technologies, new forms and practices are emerging that are especially
focused on developing social connections. These technologies can be seen as
increasingly overlaying space with digital information in order to create new
places that are mediated in part by the technology itself. These new places are
not entirely online, since they are fundamentally rooted in geographic space,
but neither are they entirely offline – they sit somewhere in between.
RETHINKING PLACE:
IMAGES OF GEOSOCIAL MAPS
The contestation of the term ‘place’ is further magnified through mobile media.
Increasingly, urban spaces are being mediated by technologies. Mobile media
not only mediates intimate relations; it also mediate how we experience
and think about spaces and places. Place is not only a space with geographic
contours, it is a space that operates across many levels: imagined and lived,
social and physical. While nineteenth-century narrations of the urban were
symbolised by the visual wanderer of the flâneur, the twenty-first-century
wanderer of the informational city has been rendered into what Robert Luke
calls the ‘phoneur’ (2006). The flâneur was both a voyeur and part of the urban
spectacle of the growing bourgeoisie: the phoneur sees the city transformed
into an informational circuit in which the person is just a mere node with
little agency.
The impact of networked and locational media on place cannot be under-
estimated (Farman 2011). Digital, mobile maps change how we navigate
and conceptualise place. Consider how the act of looking up a journey
via printed paper road maps is a distinctively different experience and
conceptualisation of space when compared to the automated and digitally
co-present experience of LBSs. LBS games like Foursquare and Jiepang high-
light how place cannot be mapped just as a geographic or physical location,
but also reflects cultural, emotional and psychological dimensions.
Through the convergence of social, locative and mobile media we are seeing
the contested notion of place becoming even more complicated. Michel de
Certeau famously defined place as a ‘“proper” and distinct location’, whereas
‘space is a practised place’ (1984: 117). This is a little confusing, because it is
counter to the common-sensical way in which most people use these two
words, in which space is merely a geographical location and place is a space
invested with emotional meaning. In this chapter we, like many other cultural
critics, will use the terms ‘space’ and ‘place’ in their colloquial sense. For exam-
ple, Tim Cresswell defines place as ‘a meaningful site that combines location,
locale, and sense of place’ (2009: 1), and Doreen Massey sees places as entangle-
ments and combinations of the ‘stories-so-far’ (2005: 130).
While maps give us one sense of space, they are incomplete in conveying the
complex and often competing cartographies of place. As Massey notes:
One way of seeing ‘places’ is as on the surface of maps … But to escape from
an imagination of space as surface is to abandon also that view of place. If
space is rather a simultaneity of stories-so-far, then places are collections of
those stories, articulations within the wider power-geometries of space. Their
character will be a product of these intersections within that wider setting, and
of what is made of them … And, too, of the non-meetings-up, the disconnections
and the relations not established, the exclusions. All this contributes to the
specificity of place. (2005: 130)
In Wilken and Goggin’s timely Mobile Technology and Place (2012), the
various authors reflect upon the conceptual currents and controversies facing
notions of place in the face of increasingly ubiquitous mobile and location-
aware devices. With the ‘spatial turn’ in media studies (Falkheimer and Jansson
2006), entanglements between space, mobility and sociality have, for good
reason, been further complicated. For Wilken and Goggin:
daily lives and habits of billions of people worldwide – and for the manifold
ways that they mutually inform and shape each other. Place is a notion that is
of enduring relevance – one worth mobilising – if we are to comprehend fully
how we think about and experience who we are, where we are, and the ways
we interact and relate with one another. In other words, place is considered a
vital notion in that it represents a ‘weaving together’ of social and human-
environment interaction in ways that situate place as central to how embodied,
technologically mediated mobile social practice is understood. (2012: 18)
These authors argue that the ‘meshing of located place and networked space’
creates crucial questions, especially around whether mobile media ‘collapse
the space-place distinction, or enable “space” and “place” to be simultaneously
present’ (2012: 185). They argue that mobile media devices create a different
dynamic around perpetually interrupted and distracted body–screen–place
relations (2012: 194) in which ‘placing’ and ‘presencing’ are entangled.
Building on Gordon and de Souza e Silva, Richardson and Wilken argue
that mobile media practices involve a series of overlapping ‘presences’
(telepresence, co-presence, located presence and net-local presences) in every-
day spaces (2012: 195).
These ‘presences’ and their intimate, social and networked affects construct
a rich fabric of emotional, social, technological, electronic and geographic
overlays that inform, and are formed by, existing social and cultural prac-
tices. To understand the implications of these presences on notions of place
and intimacy, locative media needs to be studied as part of everyday social
and mobile media practices. Case studies of educational, creative and experi-
mental projects by researchers such as Gordon, de Souza e Silva and Christian
Licoppe (2004; Licoppe and Inada 2006) are useful, but are limited because
they are not studies of locative media in the context of the everyday. In par-
ticular, what has been missing from the conversation has been the role of
camera phones in the use and adaptation of locative media in everyday contexts.
As this chapter’s opening vignettes described, with locative media there is more
impetus for users to take and share camera-phone images as a place-making
exercise that interweaves the social with the geographic. In the next section
we consider the changing role of the camera phone as part of locative, social
and mobile media networks.
in question as well as view the image captured in the place in question, thus
overlaying another context. Finally, the social distribution of the images cre-
ates a social public for those images, thus overlaying another context, and the
image tags entered by the public overlay yet another context.
CONCLUSION
We began this chapter with an overview of the development of mobile
technology, noting the development of LBSs and their integration into
mobile media. An important point we made was that smartphones are emerg-
ing as converged technologies that bring together other technologies like
wireless internet, LBSs, camera phones and so on. Mobile devices do not
just extend the number of places that you can use social media; they bring
social media to those places and, through the LBS, contribute to the con-
struction of new cartographies of space. In other words, they provide us
with new ways of mapping meaning to space and creating new places.
The growth of locative media is having significant impact upon cul-
tural practice, place-making and relationships in ways that are shifting,
ongoing and emergent (Cincotta et al. 2011; Farman 2011; Gordon and
de Souza e Silva 2011; de Souza e Silva and Frith 2012). This phenom
enon has both positive and negative impacts upon localised notions of
privacy and surveillance across both micro and macro, and individual and
collective, levels (Michael and Michael 2010; Gazzard 2011). While much
analysis has been conducted into experimental forms of locative media/
art (de Souza e Silva and Sutko 2009; de Souza e Silva and Hjorth 2009),
the increased ubiquity of locative media through devices such as the smart-
phone will undoubtedly transform the way in which place and mobility
are articulated.
feature of modern societies, just as the steam engine and the factory were the
defining features of industrial societies. However, when theorists like Manuel
Castells (1996) declare that we live in a networked society, there is another
subtle element being introduced. Here Castells is using the idea of the
networked society as a metaphor for how the structures in society are being
organised in a way that is similar to a computer network. For others, the
analogy of the network seems to neglect the importance of the journey and
movement within these spaces as part of broader entanglements of place
(Pink 2011).
In Chapter 3 we examined some of the practical applications of this. Barry
Wellman’s notion of the networked individual represents a person who is
able to control their social environment better because of their networked
connections. People like boyd (2008, 2011) have argued that such net-
worked-media forms as the SNS provide a place in which people can
come together to form networked publics – publics that are mediated by the
technology. To summarise boyd’s argument as presented in Chapter 3, this
mediation of the network plays an important role in defining what is different
about these ‘networked’ publics. Networks extend the reach of publics
geographically and temporally; for example, think of a special-interest
internet forum. Members are from all over the world and while some might
constantly monitor the forums, many log in just once a week. As boyd
observes, ‘as social networking sites and other genres of social media
become increasingly widespread, the distinctions between networked pub-
lics and publics will become increasingly blurry’ (2011: 55). In these
dynamics, which some define as an intimate turn, the boundaries between
social media and everyday life further erode.
While the networked qualities of networked publics are important, there
is some evidence to suggest that something else is going on. As we saw in
Chapter 3, studies in the US have found that people’s online social networks
often consisted of people who they knew well and who lived within a rela-
tively small distance. Then, in Chapter 6, we looked at the way youth in
China were using SNS games to stay in touch with their family as they moved
away from home for work or study.
This brings us to a third central theme of the book: intimacy. Through
their work looking at the uses of mobile and social technologies in the Asia-
Pacific region, Hjorth and Arnold (2013) conclude that intimacy plays a very
important role in social networking. They suggest the notion of ‘intimate
publics’ to emphasise the role that intimacy plays in the construction of
online publics. They are not arguing that networks do not also structure
publics, but that intimacy is a primary structural factor. Intimacy here is
not only romantic or familial intimacy, but also includes other kinds of
intimacy, like the connection that two people have by virtue of the fact they
went to the same school or come from the same country. As we have noted
previously, as intimacy becomes increasingly public (Berlant 1998), we see
various forms of ‘presence bleed’ (Gregg 2011) across publics, networks,
SNSs, platforms and media. So too, the ways in which cultures are imagined
and experienced rely on particular forms of cultural intimacies.
This movement of intimacy into the context of the social returns us to
questions about empowerment and control. Instead, we can reframe the
empowerment/control dialectic in terms of a shift in social dynamics in
which displays of intimacy begin to play an increasingly important role in
the way we socialise and interact with the media. For example, even jour-
nalism has been affected by the intimacy turn through utilising an intimate
mode of address online. As Goggin has pointed out, much of Twitter’s eti-
quette borrows from more intimate media like SMSs (Goggin 2011). This
movement of intimacy into the public is not simply a phenomenon of social
media, it is a social phenomenon that is shaping, and being shaped by, con-
temporary media.
In the final analysis, then, we would like to think that Understanding Social
Media is not about understanding the specific technologies and politics that
are only associated with this month’s stock market favourites. Far more
importantly, Understanding Social Media is about comprehending the way in
which this new medium is both affecting and reflecting social developments
more broadly, and as a result of understanding social media, we will come
to develop a better understanding of the world in which we live.
Big Media – large companies that treat the news as a commodity, where
costs of making news are pushed down while profits are maximised; this is
rarely aligned with good journalism.
Casual games – social media games that allow players to engage with the
games for minutes at a time, as opposed to the more demanding MMOGs.
Computational turn – a term that points to the way that computer technology
appears to be changing processes and structures of existing organisations.
For example, the computational turn in journalism relates to the way that
news research and publication has been altered by the emergence of the
networked computer and the internet.
Killer app – a software application that is so successful that it sells the plat-
form that it runs on.
Local, the – a term that describes things that are close to an individual, not
just physically (the local shops) but also culturally and socially (close friends,
for example, may not be physically close).
Mobile publics – associations between people (publics) that are made possible
owing to or through the use of mobile technologies.
Networked publics – public groupings that are structured by the logic and
reality of computer networks.
Produser – a user who produces (rather than just consuming) internet con-
tent. The term emphasises the highly interactive nature of much social
media use, wherein the user often contributes their own creative products
to websites. Produsage includes activities such as uploading videos and
photos to web services like Flickr and building and maintaining blogs.
Remediation – a term used by Jay David Bolter and Richard Grusin (1999)
to describe the way that new media tends to call upon the conventions of
earlier forms of media. So, for example, photography remediates perspec-
tive painting, and YouTube remediates television.
Smart mobs – a large group of people who use mobile technologies as a way
of connecting with each other, thus allowing the group to act with a kind
of collective intelligence.
SMS (short messaging system) – the technical term for mobile phone
texting; that is, sending short text messages from one mobile phone handset
to another.
SNS (social network site) – websites that support or extend social networks.
According to danah boyd, the basic requirements of an SNS are: 1) the
ability to create a profile; 2) a ‘friends’ list or similar; 3) exploration of friends
lists and perhaps others in the system (boyd and Ellison 2007). Typically,
most online relationships are with people who are already part of the user’s
offline social network.
Social, the – a general term used to describe the social world that constantly
surrounds us and in which we live. The term reminds us that although we
live in a physical world of things, we also live in a social world, which has
a very large influence on how we act and behave.
UGC (user generated content) – this is similar to UCC but covers a broader
range of user-produced material. In this book we define UGC as material
that is produced as a by-product of another activity, possibly without any
knowledge or intent on the part of the creator. An example of UGC could
be forum posts, in which the content of the website is text created by users,
and read by other users.
Web portals – sites that aim to aggregate users around centralised content.
Alden, C. (2005) ‘Looking back on the crash’, The Guardian, 10 March. Retrieved
from http://technology.guardian.co.uk/online/story/0,3605,1433697,00.html.
Allen, M. (2009) ‘Tim O’Reilly and Web 2.0: The economics of memetic liberty and
control’, Communication, Politics & Culture, 2 (2): 6–23.
Anderson, K., Nafus, D., Rattenbury, T. and Aipperspach, R. (2009) ‘Numbers have
qualities too: Experiences with ethno-mining’, Ethnographic Praxis in Industry
Conference Proceedings, (1): 123–40.
Anderson, L. (2011) ‘Demystifying the Arab Spring’, Foreign Affairs, May/June.
Retrieved from www.ssrresourcecentre.org/wp-content/uploads/2011/06/Anderson-
Demystifying-the-Arab-Spring.pdf.
Andrejevic, M. (2011) ‘Social network exploitation’, in Z. Paparcharissi (ed.),
A Networked Self. New York: Routledge. pp. 82–101.
Bamman, D., O’Connor, B. and Smith, N. (2012) ‘Censorship and deletion practices
in Chinese social media’, First Monday, 17 (3). Retrieved from www.uic.edu/
htbin/cgiwrap/bin/ojs/index.php/fm/article/view/3943/3169.
Banks, J. and Humphreys, S. (2008) ‘The labour of user co-creators’, Convergence,
14 (4): 401–418.
Barlow, J. P. (1996) A Cyberspace Independence Declaration. Retrieved from http://
w2.eff.org/Censorship/Internet_censorship_bills/barlow_0296.declaration.
Barthes, R. (1966 [1977]) ‘The death of the author’, in S. Heath (ed.), Image – Music –
Text. New York: Hill and Wang. pp. 142–48.
Batchen, G. (2001). Each Wild Idea: Writing, Photography, History. Cambridge,
MA: MIT Press.
Baym, N. (1998) ‘The emergence of on-line community’, in S. Jones (ed.), CyberSociety
2.0: Revisiting Computer-mediated Communication and Community. Thousand
Oaks, CA: Sage. pp. 35–68.
BBC Editorial (2002) ‘Amazon turns its first profit’, BBC News, 22 January.
Retrieved from http://news.bbc.co.uk/1/hi/business/1775294.stm.
Bell, D. and Kennedy, B. (2000) The Cybercultures Reader. London: Routledge.
Beniger, J. R. (1986) The Control Revolution: Technological and Economic Origins
of the Information Society. Cambridge, MA: Harvard University Press.
Bennett, W. L. (2008) ‘Changing citizenship in the digital age’, in W. L. Bennett (ed.),
Civic Life Online: Learning How Digital Media Can Engage Youth. Cambridge,
MA: MIT Press. pp. 1–24.
Bennett, W. L., Wells, C. and Rank, A. (2009) ‘Young citizens and civic learning: Two
paradigms of citizenship in the digital age’, Citizenship Studies, 13 (2): 105–120.
Berlant, L. (1998) ‘Intimacy’: A Special Issue. Critical Inquiry, 24 (2): 281–88.
Hampton, K. N., Sessions Goulet, L., Rainie, L. and Purcell, K. (2011) Social
Networking Sites and Our Lives. 16 June. Retrieved from http://pewinternet.org/~/
media//Files/Reports/2011/PIP%20-%20Social%20networking%20sites%20
and%20our%20lives.pdf.
Haythornthwaite, C. and Wellman, B. (1998) ‘Work, friendship, and media use for
information exchange in a networked organization’, Journal of the American Society
for Information Science, 49 (12): 1101–1114.
Herzfeld, M. (1997) Cultural Intimacy: Social Poetics in the Nation State. New York:
Routledge.
Hine, C. (1998) Virtual Ethnography. Paper From Internet Research And Information
For Social Scientists Conference, 25–27 March, University Of Bristol, UK.
Hinton, S. and Whitelaw, M. (2010) Exploring the Digital Commons: An Approach
to the Visualization of Large Heritage Datasets. Paper presented at Electronic
Visualisation and the Arts Conference, 5–7 July, London.
Hjorth, L. (2003) Kawaii@keitai’, in N. Gottlieb and M. McLelland (eds), Japanese
Cybercultures. New York: Routledge. pp. 50–59.
Hjorth, L. (2005) ‘Locating mobility: Practices of co-presence and the persistence of
the postal metaphor in SMS/MMS mobile phone customization in Melbourne’,
Fibreculture Journal, (6). Retrieved from http://journal.fibreculture.org/issue6/
issue6_hjorth.html.
Hjorth, L. (2007) ‘Snapshots of almost contact: A case study on South Korea’,
Special Issue, G. Goggin (ed.), Continuum, 21 (2): 227–38.
Hjorth, L. (2012) ‘iPersonal: A case study of the politics of the personal’, in L. Hjorth,
J. Burgess and I. Richardson (eds), Studying Mobile Media: Cultural
Technologies, Mobile Communication, and the iPhone. New York: Routledge.
pp. 190–212.
Hjorth, L. and Arnold M. (2012) ‘Home and away: A case study of students and
social media in Shanghai’, in P. Law (ed.), New Connectivities in China: Virtual,
Actual and Local Interactions. Dordrecht: Springer. pp. 171–83.
Hjorth, L. and Arnold, M. (2013) Online@AsiaPacific: Creativity, Literacy and
Politics in the Asia–Pacific Region. London: Asia’s Transformation series,
Routledge.
Hjorth, L. and Chan, D. (2009) Gaming Cultures and Place in Asia-Pacific. New York:
Routledge.
Hjorth, L. and Gu, K. (2012) ‘Placing, emplacing and embodied visualities: A case
study of smartphone visuality and location-based social media in Shanghai, China’,
Continuum, 26 (5): 699–713.
Hjorth, L. and Kim, H. (2005) ‘Being there and being here: Gendered customis-
ing of mobile 3G practices through a case study in Seoul’, Convergence, 11:
49–55.
Hjorth, L. and Kim, Y. (2011) ‘The mourning after: A commentary on crisis manage-
ment in Japan post 3.11’, Television & New Media Journal, 12 (6): 552–59.
Hjorth, L. Burgess J. and Richardson I. (eds) (2012) Studying Mobile Media: Cultural
Technologies, Mobile Communication, and the iPhone. New York: Routledge.
Hjorth, L., Wilken, R. and Gu, K. (2012) ‘Ambient intimacy: A case study of the
iPhone, presence, and location-based social networking in Shanghai, China’, in
L. Hjorth, J. Burgess and I. Richardson (eds), Studying Mobile Media: Cultural
Marwick, A. (2008) ‘To catch a predator? The MySpace moral panic’, First
Monday, 13 (6). Retrieved from http://firstmonday.org/htbin/cgiwrap/bin/ojs/index.
php/fm/article/view/2152.
Massey, D. (2005) For Space. London: Sage.
McVeigh, B. (2000) ‘How Hello Kitty commodifies the cute, cool and camp:
“Consumutopia” versus “control” in Japan’, Journal Of Material Culture, 5 (2):
291–312.
Meikle, G. and Redden, G. (2010) News Online: Transformations and Continuities.
London: Palgrave.
Michael, M. G. and Michael, K. (2010) ‘Towards a state of überveillence’, IEEE
Technology and Society Magazine, 29 (2): 9–16.
Miller, D. (2011) Tales from Facebook. Cambridge: Polity.
Miller, D. and Slater, D. (2001) The Internet: An Ethnographic Approach. London: Berg.
Milne, E. (2004) ‘Magic bits of paste-board’, M/C: A Journal of Media and Culture, 7.
Retrieved from www.media-culture.org.au/0401/02-milne.php.
Mørk Petersen, S. (2008) ‘Common banality: The affective character of photo sharing,
everyday life and produsage cultures’. Doctoral dissertation, IT University of
Copenhagen.
Munster, A. and Murphie, A. (2009) ‘Web 2.0 is a doing word’, Fibreculture Journal,
14. Retrieved from http://fourteen.fibreculturejournal.org/.
Nakamura, L. (2002) Cybertypes: Race, Ethnicity, and Identity on the Internet. New
York: Routledge.
Nardi, B. A. (2010) My Life as a Night Elf Priest: An Anthropological Account of
World of Warcraft. Ann Arbor, MI: University of Michigan Press.
National Endowment for the Arts (2010) Audience 2.0: How Technology Influences
Arts Participation, June. Retrieved from www.nea.gov/research/new-media-report/
index.html.
Nugroho, Y. and Syarief, S. S. (2012) Beyond Click-Activism? New Media and
Political Processes in Contemporary Indonesia. Jakarta: Friedrich-Ebert-Stiftung, fes-
media Asia series. Retrieved from www.fes.de/cgi-bin/gbv.cgi?id=09240&ty=pdf.
Okabe, D. and Ito, M. (2003) ‘Camera phones changing the definition of picture-
worthy’, Japan Media Review, 28 August. Retrieved from www.ojr.org/japan/
wireless/1062208524.php.
Okada, T. (2005) ‘Youth culture and the shaping of Japanese mobile media:
Personalization and the Keitai Internet as Multimedia’, in M. Ito, D. Okabe and
M. Matsuda (eds), Personal, Portable, Pedestrian: Mobile Phones in Japanese Life.
Cambridge, MA: MIT Press. pp. 41–60.
Oldenburg, R. (1989) The Great Good Place: Cafes, Coffee Shops, Community
Centers, Beauty Parlors, General Stores, Bars, Hangouts, and How They Get You
Through the Day. New York: Paragon.
O’Reilly, T. (2005) What is Web 2.0: Design Patterns and Business Models for the
Next Generation of Software. Retrieved from http://oreilly.com/web2/archive/
what-is-web-20.html.
Palmer, D. (2012) ‘iPhone photography: Mediating visions of social space’, in L. Hjorth,
J. Burgess and I. Richardson (eds), Studying Mobile Media: Cultural Technologies,
Mobile Communication, and the iPhone. New York: Routledge. pp. 85–97.
Parks, M. R. (2011) ‘Social network sites as virtual communities’, in Z. Papacharissi
(ed.), A Networked Self: Identity, Community and Culture on Social Network Sites.
New York: Routledge. pp. 105–123.
SEO Sydney Blog (2009) Facebook Australia: User Statistics & Demographics. www.
seosydneyblog.com/2009/08/facebook-australia-user-statistics.html.
Shin, D.-H. and Shin, Y.-J. (2011) ‘Why do people play social network games?’,
Computers in Human Behavior, 27: 852–61.
Shirky, C. (2008) Here Comes Everybody. The Aspen Ideas Festival, 30 June–8 July,
Aspen Colorado. Retrieved from www.channels.com/episodes/show/12772338/
Clay-Shirky-Here-Comes-Everybody?page=5.
Shirky, C. (2009) Here Comes Everybody: The Power of Organizing Without
Organizations. New York: Penguin.
Simon, N. (2012) Museum 2.0. Retrieved from http://museumtwo.blogspot.com.au/.
Sonnenfeld, B. (Director) (1997) Men in Black [Film]. United States: Amblin
Entertainment.
Soukup, C. (2006) ‘Computer-mediated communication as a virtual third place:
Building Oldenburg’s great good places on the World Wide Web’, New Media &
Society, 8 (3): 421–40.
Stallabrass, J. (2010) ‘Can art history digest net art?’, in D. Daniels and G. Reisinger
(eds), Net Pioneers 1.0: Contextualizing Early Net-based Art. New York: Sternberg.
Standage, T. (1998) The Victorian Internet: The Remarkable Story of the
Telegraph and the Nineteenth Century’s On-line Pioneers. London: Weidenfeld &
Nicolson.
Steinkuehler, C. and Williams, D. (2006) ‘Where everybody knows your (screen) name:
Online games as “third places”’, Journal of Computer-Mediated Communication,
11 (4). Retrieved from http://jcmc.indiana.edu/vol11/issue4/steinkuehler.html.
Stephenson, N. (1992) Snow Crash. London: Penguin.
Stuart, A. (2007) ‘Citizen journalism and the rise of “mass self-communication”:
Reporting the London bombings’, Global Media Journal, 1 (1). Retrieved from
www.commarts.uws.edu.au/gmjau/iss1_2007/pdf/HC_FINAL_Stuart%20
Allan.pdf.
Surowiecki, J. (2004) The Wisdom of Crowds: Why the Many Are Smarter Than
the Few and How Collective Wisdom Shapes Business, Economies, Societies, and
Nations. New York: Doubleday.
Taylor, T. L. (2006) Play Between Worlds Exploring Online Game Culture.
Cambridge, MA: MIT Press.
Thelwall, M. (2008) ‘Social networks, gender, and friending: An analysis of MySpace
member profiles’, Journal of American Social Information Sciences, 59: 1321–30.
Thorp, J. (2009) ‘Just landed: Processing, Twitter, MetaCarta and hidden data’,
blprnt.blg. 11 May. Retrieved from http://blog.blprnt.com/blog/blprnt/just-landed-
processing-twitter-metacarta-hidden-data.
Trant, J. (2006) ‘Exploring the potential for social tagging and folksonomy in art
museums: Proof of concept’, New Review of Hypermedia and Multimedia, 12 (1):
83–105.
Turkle, S. (1984) The Second Self: Computers and the Human Spirit. New York:
Simon and Schuster.
Turkle, S. (1995) Life on the Screen: Identity in the Age of the Internet. New York:
Simon and Schuster.
Turkle, S. (2011) Alone Together: Why We Expect More from Technology and Less
from Each Other. New York: Basic Books.