ORIGINAL PAPER
The reality of friendship within immersive virtual worlds
Nicholas John Munn
Published online: 21 May 2011
� Springer Science+Business Media B.V. 2011
Abstract In this article I examine a recent development
in online communication, the immersive virtual worlds
of Massively Multiplayer Online Role-Playing Games
(MMORPGs). I argue that these environments provide a
distinct form of online experience from the experience
available through earlier generation forms of online com-
munication such as newsgroups, chat rooms, email and
instant messaging. The experience available to participants
in MMORPGs is founded on shared activity, while the
experience of earlier generation online communication is
largely if not wholly dependent on the communication itself.
This difference, I argue, makes interaction in immersive
virtual worlds such as MMORPGs relevantly similar to
interaction in the physical world, and distinguishes both
physical world and immersive virtual world interaction from
other forms of online communication. I argue that to the
extent that shared activity is a core element in the formation
of friendships, friendships can form in immersive virtual
worlds as they do in the physical world, and that this possi-
bility was unavailable in earlier forms of online interaction. I
do, however, note that earlier forms of online interaction are
capable of sustaining friendships formed through either
physical or immersive virtual world interaction. I conclude
that we cannot any longer make a sharp distinction between
the physical and the virtual world, as the characteristics of
friendship are able to be developed in each.
Keywords Virtual worlds � Friendship � MMORPGs �Interaction
Introduction
Many differences have been claimed between the rela-
tionships developed online and those developed in the
physical world. Dean Cocking and Steve Matthews have
argued that ‘net friends’ are relevantly distinct from
friends, and do not fulfil all the characteristics of friendship
(2000). More recently, Adam Briggle has critiqued this
position, arguing that online communication can (but often
does not) result in close friendships (2008). In this article I
discuss the possibility of friendship formation in the
immersive virtual worlds of massively multiplayer online
role-playing games (MMORPGs), focussing on the concept
of shared activity as a requirement of the development of
friendship. This concept has most effectively been articu-
lated by Bennett Helm (2008). I argue that MMORPGs
facilitate friendship development through shared activity in
a way parallel to that offered by physical world interaction,
and that both the immersive virtual worlds of MMORPGs
and the physical world can be distinguished from prior
generations of online interaction in virtue of their ability to
provide this medium for shared activity. As MMORPGs
are a relatively recent phenomenon, I spend some time
examining the fundamental shift that has occurred in
moving to this kind of online interaction from previous
generation forms of online interaction such as email, chat
rooms, instant messaging and newsgroups. I also discuss
the status of newly emerging immersive social environ-
ments such as Facebook, arguing that these are more
similar to the prior generation online communication than
to either physical interaction or immersive virtual worlds.
N. J. Munn (&)
School of Philosophical, Historical and International Studies,
Monash University, Melbourne, VIC 3800, Australia
e-mail: [email protected]
123
Ethics Inf Technol (2012) 14:1–10
DOI 10.1007/s10676-011-9274-6
In order to achieve this, I argue that the purpose and use of
Facebook and similar online social networks is importantly
distinct from the purpose and use of immersive virtual
worlds such as World of Warcraft, drawing on a distinction
between using these services for communication (as occurs
in Facebook use) and engaging in shared activity within
virtual worlds.
The possibility of friendship formation within immer-
sive virtual worlds is important as it enables the benefits
of friendship to be shared more broadly. In particular, it is
commonly held that special obligations arise between
friends (Scheffler 1997; Mason 1997; Leib 2007) and the
ability to form relationships incurring these obligations
remotely has the potential to raise important issues of
responsibility and expectation arising from virtual rela-
tionships. Friendships elicit responsibility, such that when
faced with a choice between acting so as to benefit a
friend or to benefit a stranger, the fact that one possible
beneficiary is a friend gives a reason to act in their
favour. If online relationships can generate friendships,
then they similarly generate these kinds of obligations,
held by us to those we have never physically met. It is
this feature of online friendships that is controversial, as it
implies duties to act in particular ways to preserve the
online friendship, including acting to the detriment of
non-friends with whom one does have physical contact,
when so acting is necessary to avoid similar detriment
accruing to the online friend. If friendships can be formed
online, then all the special obligations triggered by
friendship generally are triggered by these friendships,
and our accounts of special responsibility must be able to
take this into account. A second important consequence of
the possibility of online friendship development arises
from the ability of friendships to enhance our knowledge,
particularly in this case our knowledge of the world and
those within it. Elizabeth Telfer argues that friendship
itself is knowledge enhancing (1971), and as such the
possibility of developing true friendships online opens the
opportunity for friendship with a wider range of persons
than are generally available through physical world social
networks. People from divergent backgrounds, societies
and status are available as ‘potential friends’ who would
not be available without the medium of the immersive
virtual world. This second argument is probabilistic.
Online friendships are distinctly valuable because they
provide the opportunity for those that have them to
interact with and gain knowledge of people in social
settings distinct from their own, more easily than is the
case without online friendship. This knowledge could be
gained in other ways (as for example when you befriend
new arrivals from abroad), but the possibility of real
friendship formation online makes such friendships more
feasible for more people, more often.
The structure of the article is as follows. Firstly, I make
the case for the centrality of shared activity in the forma-
tion of friendships. Secondly, I apply the shared activity
criterion to the four kinds of activity identified above: older
generation online communications; social media; immer-
sive virtual worlds; and the physical world. Thirdly, I argue
that neither older generation online communications nor
social media have the capacity to develop close friend-
ships, while both immersive virtual worlds and the physical
world share this capacity. Fourthly and finally, I argue that
all four kinds of activity share an ability to maintain
existing friendships.
The characteristics of friendship
Aristotle identifies three kinds of friendship, the imperfect
friendships of utility and of pleasure, and the perfect
friendship of virtue, in which each participant wishes well
for the other for their own sake, rather than as a means to
either the utility or pleasure of the lesser friendships. (1998,
NE 8.3–4) It is this last, perfect friendship with which I am
concerned in this article, as imperfect friendships do not
have the same strong positive outcomes as perfect friend-
ships, and also are taken by many commentators to be more
feasibly established online. Cocking and Matthews, for
example, confine their criticism of ‘net friends’ to close
friendships, acknowledging that some of the characteristics
of friendships can and do develop through online interac-
tion (2000).
Certain characteristics of friendships are accepted by all
who work in the area. Dean Cocking and Jeanette Kennett
identify affection and well-wishing as components accep-
ted by all accounts of friendship (1998), while Bennett
Helm argues that the concepts of mutual caring, intimacy
and shared activity are shared by the majority of philo-
sophical accounts of friendship (2010). Mutual caring is the
idea that each friend cares for the other and does so for the
sake of the other, not themselves; intimacy is the notion of
a deeper relationship than mere collegiality or acquain-
tance, and shared activity is the idea that friends will do
things together, as they each enjoy the thing in question,
and, further, they enjoy doing this thing in the company of
friends.1 In this article I focus predominantly on the shared
activity criterion of friendship, which originated with
Aristotle who claimed that friends will share their activities
and in doing so improve themselves and their friendship.
(1998, NE 9.12) I do so as I consider this criterion to be
1 Discussion of these three criteria is, as Helm suggests, widespread
in the literature on friendship. In addition to Helm (2010), discussion
can be found in Cooper (1977a, b), Sherman (1987), Telfer (1971),
Thomas (1987). Helm (2009) provides many further discussions for
those interested.
2 N. J. Munn
123
foundational. It is (usually, if not always) through shared
activity that intimacy and mutual caring develop.2 Nancy
Sherman for example claims that it is the ‘‘capacity to
share and co-ordinate activities over an extended period of
time’’ that is constitutive of friendship (1987). A possible
exception arises in familial relationships, in which inti-
macy and mutual caring between parents and child exist
prior to any engagement in shared activity. However, it is
standard to draw a distinction between friendships in
general and familial relationships such that familial rela-
tionships are not friendships. (Helm 2010) Such a distinc-
tion begins with Aristotle, who distinguishes ‘‘both the
friendship of kindred and that of comrades’’ from his
general description of friendships. (1998, NE 8.12) I follow
that convention here.3 I take it that outside of the familial
environment, shared activity is the best available contender
for providing the kind of contact which is required for the
development of mutual caring and affection. If a particular
mode of interaction does not provide meaningful oppor-
tunities for shared activity, then this mode will also not be
able to cause a relationship of intimacy and mutual caring
to develop. As Aristotle says, ‘‘friendship requires time and
familiarity’’ and men cannot ‘‘admit each other to friend-
ship or be friends till each has been found lovable and been
trusted by each.’’ (1998, NE 8.3) To illustrate this point,
consider the development of a friendship between Jesse
and Kelly. They meet during a multi-day bike ride, in line
for the evening meal. Sitting together over dinner, they
already know they share an enthusiasm for bicycling.
Conversation reveals that they each work at a university,
one as a librarian, the other an academic. This provides a
further background of shared experience. Over the course
of the bike ride, they have more opportunities to converse.
It may transpire that they enjoy each others company, and
have further interests in common. From here, intimacy and
mutual caring can develop. While a chance encounter with
a stranger may in principle lead to the same kind of out-
come, it is at the least more likely that a foundation of
shared activity will provide a platform for the development
of a friendship than do situations which lack this
foundation.
Simply engaging in a mutually liked activity cannot
however suffice for that activity to be relevantly shared.
Two people may each enjoy bicycling, whilst having no
preference to bicycling with others. They may each enjoy
bicycling with others, without wishing to bicycle with a
particular other person. For the shared activity to be a
foundational component of friendship, it is also necessary
that each friend enjoys engaging in the activity with the
other. In this way, the pleasure of the activity is increased
by the company in which the activity is enjoyed. This
conception of mutual activity is articulated by Nancy
Sherman who follows Aristotle in claiming that ‘‘the best
sort of friendship provides us with companions with whom
we can share goods and interests in a jointly pursued life’’
(1987). Not every person is a candidate for friendship. I will
argue in the following section that there is an important
difference between, on the one hand, earlier forms of online
interaction and the current generation social media para-
digm of online interaction, and on the other hand, interac-
tion in immersive virtual worlds and the physical world, in
the way that these realms facilitate shared activity.
Shared activity
In the four subsections below I address distinct kinds of
interaction. I argue that the early forms of online interac-
tion discussed in ‘‘Early forms of online interaction’’ and
the social media paradigm of interaction discussed in ‘‘The
social media paradigm’’ share a characteristic of not pro-
viding an independent means of engaging in shared activ-
ity, while immersive virtual worlds (‘‘Immersive virtual
worlds’’) and the physical world (‘‘The physical world’’)
both do provide such a forum. Before beginning this dis-
cussion, I must briefly describe the content of the shared
activity under consideration. I take it that the shared
activity component of friendship requires friends to coop-
eratively engage in activity, whether in pursuit of the
experience of doing so or of some greater goal, and to do so
not only for the sake of the activity, but in order to engage
in the activity with their friends. This means that, as dis-
cussed in Sect. ‘‘The characteristics of friendship’’, one is
not engaged in shared activity just because they like
bicycling with others, and have found someone to bicycle
with. That other person must also wish to bicycle with
others, and each must want, specifically, to bicycle with the
other person, rather than simply to enjoy bicycling with
some unspecified other.
This concept of shared activity is demanding. It can be
contrasted with the accounts of social action theorists who
are concerned with examining what it means for groups of
agents to behave ‘‘in a way that is coordinated through
planning and deliberation’’ (Helm 2008). These accounts
are less demanding. A representative example is Michael
Bratman’s account of shared cooperative activity which
does not require participants to have an interest in engaging
in the activity with specified others (1992).4 It is enough for
2 For examination of Intimacy and Mutual Caring, see Cocking and
Kennett (1998), White (2001). I do not address these criteria in depth
in this article.3 While some commentators, such as Rorty (1993) explicitly include
familial relationships within the realm of friendship, I follow the
majority in excluding them. 4 Other accounts include Velleman (1997), Gilbert (2000).
The reality of friendship within immersive virtual worlds 3
123
the participants to cooperatively pursue designated ends,
where each participant has a desire to achieve that end.
Under my account of shared activity, friends engaged in
such activity jointly pursue a goal when all of them not
only desire a particular outcome, but also desire that the
outcome be the product of the combined activity of the
group, as it is composed. As friends, they may be willing to
reduce the likelihood of achieving the desired outcome, in
order to ensure that if it is achieved, the group which
achieves it is composed of the friends. Bratman’s account,
and those of action theorists who are concerned with social
activity more broadly, rather than shared activity in the
context of friendship, does not require this final step. This
can be illustrated by considering the case of a social
football league. A group of friends who share an interest in
football may form a team, with a goal of winning the local
league. In order for their actions to constitute shared
activity, the goal must be further specified, to winning the
league as part of a group of friends. Bratman’s agents could
attempt to achieve their goal by bringing in ringers who
shared the goal of winning the league and were willing to
co-operatively engage in attempting to do so, but they
would achieve this at the cost of sacrificing the goal of
some members of the group, those excluded in order to
increase the chance of winning. My agents, as friends,
could not do this, as doing so would undermine the
important process, of winning as a group.
Further, I distinguish between communication and
activity. I take communication to be the planning of
activity, the sharing of ideas, the development of proce-
dures and so on, while activity involves putting the things
discussed into practice. In making this division I follow the
usage common to social action theorists, whose examples
of shared activity are consistently of this kind (Tuomela
1993; Bratman 1992). While in some instances communi-
cation could arguably extend to activity, as in for example
the orchestration of a letter writing campaign as a political
protest, such instances are both rare and contentious, and
for the moment I leave them aside.
Early forms of online interaction
While the landscape of the modern internet is dominsated
by the social media paradigm, exemplified by Facebook
(founded in 2004) and Twitter (founded in 2006), which
have rapidly displaced smaller predecessors such as
MySpace, Friendster and Bebo, much of the literature on
friendship in the online sphere predates this development.
As such, many discussions of friendship online, including
the virtual interactions considered by Cocking and Mat-
thews in ‘‘Unreal Friends’’ (2000), focus on preceding
generations of interaction, namely ‘‘e-mail, chat rooms,
instant messaging, newsgroups, and other means.’’
(McKenna et al. 2002) When used by strangers (that is, by
people who do not have pre-existing physical world rela-
tionships), the interactions available to people in this set-
ting are not conducive to the development of shared
activity. These methods of interaction are all primarily
means of communication. They are useful for people to
talk to one another, for the sharing of ideas and information
or the arranging of physical world activities. But within the
setting of a chat room or a newsgroup, it is only the con-
versation that is shared. Rather than doing anything new
with the other participants, you are engaged in recounting
things you have done, explaining your interests to others,
or potentially comparing your prior (non-shared) experi-
ences of particular kinds of activities with others. So for
example, a newsgroup may be formed for camping
enthusiasts in Central Otago, in New Zealand. Within this
newsgroup, people who have gone camping in the region,
or who wish to, or who want to determine whether they
wish to, can share experiences, offer advice and converse
with others who have these same interests. Importantly
though, any shared interactions that arise from this forum
will occur in the physical world. The newsgroup environ-
ment (and similarly, the email, chat room or instant mes-
saging environment) does not provide a realm for actually
engaging in shared activity.
This situation arises because, in these earlier online
interactions, the act of communication was itself central.
When one entered a chat room and began communicating
with the other participants, the novel factor was being able
to talk to distant persons who were previously unknown to
you, to develop relationships with people you have never
met and may never meet in the flesh. To develop ongoing
relationships with these people, in this environment, relied
upon a continuing interaction based solely on this com-
munication. To engage in shared activity required breach-
ing the barriers of the online world, and engaging in the
physical world. While this medium provided an expanded
range of means for people who already had the foundation
of a friendship to expand that friendship, the development
of the friendship was reliant on the physical world inter-
actions occurring additionally to the online. This means
that this form of communication is incapable of providing
the shared activity required for friendship.
The social media paradigm
The social media paradigm is at this stage exemplified by
Facebook, which is wildly successful and pervasive.5
5 The companies own statistics place them as having over 500
million active users, over half of whom access the site each day.
http://www.facebook.com/press/info.php?statistics (Last Accessed
December 6, 2010).
4 N. J. Munn
123
Facebook is a social media system whereby you form and
participate in networks of people, by choosing to ‘friend’
them. Having become someone’s ‘friend’ on Facebook,
you each have access to the information the other places on
the system, and are thereby able to track the activity of
those you have befriended. This activity may include
updates as to what the person is doing at a given time,
photographs and video of the person uploaded by them-
selves or others within your network, and information as to
products, companies or activities which the person is
tracking, planning, or participating in. Importantly, con-
nections on Facebook (and other similar services) follow a
social networking model whereby you befriend people you
have pre-existing connections with: Friends, family,
acquaintances and work colleagues. This is the ‘circle of
friends’ model of social networking, and it distinguishes
this type of interaction from the interactions considered in
‘‘Early forms of online interaction’’ above, which did not
rely on prior knowledge of other participants (Rosen 2007).
Facebook provides a means by which family, friends,
and mere acquaintances are able to maintain contact with
you, to learn as much as you are willing to tell them about
your current location, thoughts, and actions, and to find
other mutual acquaintances that they may have lost contact
with. Within this paradigm, there are two distinct levels at
which the concept of friendship operates. Some partici-
pants in the social media paradigm treat friendship as being
purely a matter of acquaintance, such that they will ‘friend’
people they know not because they want to talk to them,
but in recognition of a connection forged by shared place or
experience at some stage in the past. We are concerned
here with the stronger form of friendship operating within
the social media paradigm, which occurs when you use
social media to interact with people you consider to be
friends in a strong sense; involving mutual caring, inti-
macy, and shared activity.
In this strong sense, social media act as a means for the
continuation of friendships, not for their development.
They are used to contact old friends, to arrange activities
with them, not to make new friends. The importance of
Facebook is not to enable you to meet new people or make
new friends, but to keep you in touch with existing friends,
and, increasingly, to facilitate your participation in the
shared activities constitutive of that friendship. So, for
example, you may organise a party with a group of your
friends by inviting them to it via Facebook (rather than, as
in the past, by email or telephone). When you do this, it is
with the intention that you engage in a shared physical
world activity, which will (all going well) strengthen the
existing friendships you have with the people you invite.
After the fact, Facebook provides a repository for memo-
ries of the activity, through photographs and comments.
However, as with the earlier online interactions discussed
in ‘‘Early forms of online interaction’’ above, the actual
shared activities are engaged in outside of the interactions
in social media.
Immersive virtual worlds
The preceding two discussions have been relatively brief,
as the features of each are commonly known. Immersive
virtual worlds, however, are a more recent and less
pervasive phenomenon, with Everquest in 1999 being the
most successful early version, peaking at 450,000 sub-
scribers. (Sony 2004) While Everquest itself was by no
means the first such game, it was the first to capture
widespread public interest, particularly amongst the
broader society. Previous online worlds, such as the
MUDs (Multi-User Dungeons) which were popular
amongst those with the finance, technical acumen and
access to technology required to participate in them
during the formative years of the internet, lacked the
accessibility, population base, graphical interface and
widespread acceptance that characterise the immersive
virtual worlds I am concerned with in this paper.
Scholarly interest has focussed more recently on Second
Life, which is a free and largely user generated experi-
ence (See Cole and Griffiths 2007; Terra Nova 2011),
but the dominant force in MMORPGs is World of
Warcraft, which has recently passed 12 million active
subscribers. (Blizzard 2010) In the following discussion I
use World of Warcraft (henceforth WOW) as a para-
digmatic example of the virtual world impact on the idea
of friendship. Before doing so, however, I must
emphasise the importance of the ‘virtual’ component of
the immersive virtual worlds under consideration. The
activities engaged in in immersive virtual worlds could
equally be engaged in by groups of friends or acquain-
tances playing board games, or playing co-operative
games in a LAN cafe or similar environment in which
they have the opportunity for physical contact with the
other participants. Were friendships to develop because
of engagement in these sorts of activities, we would gain
no insight as to whether it was the activity engaged in or
the fact that it was done while in physical proximity that
enabled friendship development. By looking at activities
undertaken in immersive virtual worlds, we remove the
possibility of physical contact and thereby can establish
the possibility of friendship formation amongst people
who have never met in the physical world. The goal is
to show that it is the act of engaging in shared activity,
rather than the medium in which that activity is engaged
in, which is the crucial determinant of friendship
development.
The reality of friendship within immersive virtual worlds 5
123
The workings of immersive virtual worlds
In WOW style virtual worlds, participants have a choice as
to whether to pursue independent or collaborative
advancement. Participants who choose to play indepen-
dently are able to progress without relying on building
relationships with other players. They can level (progres-
sively gain abilities through activity and the performance
of tasks), instance (battle against particular sets of monsters
in scripted environments as part of a group of 5/10/25
players), and PvP (engage in player vs. player combat,
which pits teams of player characters against each other in
capture the flag and other styles of game). All of these
activities can be done in randomly generated groups, the
other participants chosen without reference to prior
knowledge or interaction. Alternatively, players who
choose to level collaboratively have access to a number of
tools to enable collaborative levelling. These can be divi-
ded into internal and external means. Internal means are
processes provided by the virtual world itself through
which friendships can be developed and maintained.
External processes are ones developed by the participants
themselves to facilitate the development and maintenance
of friendships within the virtual world.
Internal processes available in WOW include the ability
to befriend particular characters in the game, an ability that
has recently been extended to also allow you to befriend
the player, rather than the character. By doing this, you will
be informed whenever that character is online, enabling
you to maintain contact with them. By befriending the
player rather than a particular character, you are able to see
when the player is online in any capacity, whether on any
of a variety of characters, on any number of servers, or
even in any of the games which share Blizzard entertain-
ment’s tracking system. This friending process provides the
simplest level of interaction between participants in the
virtual world.
A further and more nuanced means of interaction is pro-
vided by guilds. Guilds are groups of participants who share
dedicated communication channels, pool their in world
resources, and are able more easily to track each others
online presence. Joining a guild is done both for the social
components of guild membership, through the availability of
like minded participants with whom to share the world
experience, and for the game-play benefits, such as the easier
availability of positions in raid groups, organised PvP com-
bat, and other components of the virtual world. Guilds
thereby facilitate the sharing and co-ordination of activities
over an extended period of time, that Sherman identifies as
requirements for the development of friendships (1987).
External processes through which interaction in these
virtual worlds is reinforced are often corollaries to guild
processes within the game. For example, guilds may
provide web-pages and forums in which their members can
converse outside the game. These are likely to be joined by
Facebook groups (currently, Facebook is sufficiently
dominant in the social media sphere as to be the only
system worth considering) and ventrilo servers. Ventrilo is
a voice over internet programme allowing verbal commu-
nication between large groups of people. WOW provides
for up to 40 person groups for some encounters (currently,
40 person groups only exist in some of the player versus
player battlegrounds. Against computer scripts, the largest
current content encounters are for 25 people), and Ventrilo
enables all 40 to listen to instructions and converse with
each other.
In combination, these internal and external factors allow
for very strong social bonds to be formed between partic-
ipants in immersive virtual worlds who choose to develop
relationships within them. Richard Rouse argues that the
kind of social bonds formed in these game worlds could
become the primary motivation for continued participation
(2000). This is very important for this debate, as it makes
the game world a vehicle for the relationship, rather than
itself being the reason for participation. Nick Yee has run
several studies on the perceptions participants have of
relationships in immersive worlds, which suggest that those
involved in the worlds consider them to be capable of
developing friendships, and to actually do so (2006a, b, c).6
With the general structure of the gaming situation estab-
lished, I turn now to the particular development of shared
activity in WOW.7
Shared activity in WOW
In contrast to the kinds of activity discussed in ‘‘Early
forms of online interaction’’ and ‘‘The social media para-
digm’’ above, the primary focus of social interaction in
WOW and similar immersive virtual worlds is not com-
munication, but shared experience. Shared experience, or
activity, is Aristotle’s core requirement for friendship for-
mation, as discussed in (‘‘The characteristics of friend-
ship’’) above. Participants in these virtual worlds find
others who they wish to share experiences with, within the
confines of the virtual world itself. Rather than facilitating
a physical world interaction, these virtual worlds provide a
realm in which shared activity takes place. This difference
is illustrated through consideration of the ways in which
6 See the project archives at http://www.nickyee.com/index-
daedalus.html for a description of his project and numerous articles
published reporting on the work.7 This section is a brief summary of the values of immersive virtual
worlds. For a more detailed discussion of MMORPG norms, see
Verhagen and Johansson (2009) For detailed examination of WOW in
particular, see Nardi and Harris (2006).
6 N. J. Munn
123
the activities detailed in ‘‘Early forms of online interac-
tion’’ and ‘‘The social media paradigm’’ interact with
immersive virtual worlds. The relationship in which an
environment such as WOW stands to Facebook, or to an
instant messaging system, is much more similar to that in
which physical world activities stand to these things, than it
is to the things themselves. That is, people use Facebook to
organise shared activity within the immersive virtual
world, or to reminisce about past shared activities within
that immersive virtual world. They plan shared activities
via email, or on discussion boards, and execute those
activities within the virtual world, in the way that others
plan to catch up with friends in a bar in town via email, and
then do the catching up in the physical world.
So, interactions within the virtual world system are
founded on shared activity. This is a relatively new devel-
opment in online communication, and has significant
implications for the way in which friendships can develop in
this medium. Because activity rather than communication
itself is central to interaction within an immersive virtual
world, the virtual world paradigm more closely mirrors
many of our real world engagements with our friends. While
it is often nice to catch up with a friend you have not seen in
some time and talk about what you have done without each
other, the bonds of friendship are initially formed by the
experiences you share. Virtual worlds provide shared
experiences that can fulfil the role of forming that initial
bond between people. You form a group of adventurers, and
you work together against often overwhelming odds toward
specific objectives. By doing this, those in the group dem-
onstrate their character, their roles and desires, and it
becomes apparent to the other members of the group whe-
ther or not the prerequisites for friendship are present. That
is, the virtual world activity is an opportunity for the
potential friends to be found ‘lovable and trustworthy’, as
Aristotle requires. (NE 8.3) Similarly, over an extended
period of participation in this shared virtual activity, the
group will grow closer as friends, and improve themselves in
terms of in game ability, and in general skills such as co-
ordination, co-operation and patience, thereby satisfying
Aristotle’s criteria for the importance of shared activity. (NE
9.12) As friendship develops within the group, you will
come to be interested in what they do when they are not with
you in the virtual world. When you meet for your weekly or
daily jaunt into the virtual world together, one of the ways in
which you pass the time is to ask how your friends have
been, as you would ask a climbing friend what they have
been doing since the last time you saw them at the wall, or a
bicycling friend what they have been doing since your last
shared ride. This suggests that interaction in these immersive
virtual worlds is distinct in kind from the interactions dis-
cussed in ‘‘Early forms of online interaction’’ and ‘‘The
social media paradigm’’. Before discussing whether there is
sufficient evidence to draw the conclusion that friendships
can develop in this kind of environment, I turn to a brief
discussion of the characteristics of shared activity in the
physical world.
The physical world
The physical world is the medium through which friend-
ships have traditionally been established. It is by compar-
ison to the patterns of friendship formation in this realm
that we can make judgements as to the ability of other
realms to generate friendships. We have seen from the
above three sections that immersive virtual worlds are
relevantly similar to the physical world, while social media
and previous generation online interactions are importantly
distinct from the physical world in terms of the kind of
interaction involved in them. In this section I argue that, as
per ‘‘Immersive virtual worlds’’ above, it is shared activity
in the physical world which provides the basis for friend-
ship formation.
We gain friends in many places. Our oldest friends are
likely to be the children of our parent’s friends, or perhaps
our classmates in the early years of school. Newer friend-
ships are developed and maintained at university, in
sporting teams, and through other activities, such as
debating and war-gaming. A key feature of these friend-
ships, aside from the physical realm in which they develop,
is that they are discerning. In the physical world, you are
not friends with everyone you went to school with, nor with
everyone you work with. You become friends with those
whom you connect with in some deeper way. Perhaps you
both enjoy a particular class, both despise a particular
manager. You choose a candidate for friendship and see if
a friendship will develop by doing things with that person.
When you mutually enjoy each others company, and find
activities you wish to engage in together, a friendship
forms. Subject to the analysis engaged in in Sect. ‘‘Shared
activity’’ above, in which the requirements for relevantly
shared activity were outlined, there is no special kind of
activity which makes this friendship formation possible,
because the kind of activity is not central to the process of
friendship formation. Rather, engaging in shared activity is
itself the key feature that enables friendship formation.
Over long periods of time, and through repeated and pro-
longed shared activity, the bonds of intimacy and mutual
caring also develop, such that the friendships we have
formed become close ones.
The capacity to develop friendships
It is uncontroversial that we are capable of developing
friendships in the physical world. We do, after all, do so. In
The reality of friendship within immersive virtual worlds 7
123
this section I argue that there is nothing to distinguish
between the capacity of the physical world and that of
immersive virtual worlds in developing the kinds of bonds
that lead to friendships.
The characteristic of the physical world which provides
the ability to develop close friendships is that the physical
world provides a forum in which potential close friends can
interact and learn about each other’s proclivities. They can,
as Sherman claims, share goods and interests in a jointly
pursued life (1987). This is reliant on activity, rather than
simple communication, and it is for this reason that I
classify immersive virtual worlds and the physical world as
relevantly similar, and distinct from social media and other
online communication. Consider our subjects Jesse and
Kelly from Sect. ‘‘The characteristics of friendship’’. They
met and became friends on a bike ride. They could have
met online, on a forum about bicycling, but if they had
done so, the process of developing acquaintance into
friendship would have been significantly more difficult, if
not impossible to achieve without at some stage tran-
scending the online and entering the physical realm. By
becoming friends, and engaging in their shared desire to
bicycle, they assist each other in achieving lives of the kind
they desire. Within immersive virtual worlds, online
friends can similarly assist one another. For instance, there
are significant costs associated with changing characters
within immersive virtual worlds, measured in the time
expended on the initial character, and the rewards that
character has obtained, all of which are bound to that
character. A player who wants to change characters then is
helped by the willingness of their friends to assist in lev-
elling and developing the new character, such that the
player can more quickly rejoin their friends in shared
activity. Because all concerned share an interest not only in
playing the game, but in doing so together, this mutual
co-operation facilitates the joint pursuit of this goal.
Amongst the chief difficulties for the development of
friendships online is the difficulty of ascertaining the truth
of the claims participants make, both about their desires
and about their character. Absent indicators such as body
language, tone and inflection in speech, determining whe-
ther someone is telling the truth online is more difficult
than in the physical world. Caspi and Gorsky recently
found that one-third of respondents admitted to engaging in
some level of deception online, while simultaneously
believing that the practice of deception was significantly
more widespread than this (2006). Shared activity, whether
virtual or physical, mitigates this difficulty, as it is sub-
stantially more difficult to maintain an assumed identity
under pressure or during exertion than it is when you have
the luxury of time that social media provides. Similarly, in
both immersive virtual worlds and the physical world,
voice communication is widespread, while it is not in the
other two situations considered. This minimises the ability
of participants in immersive virtual worlds to present the
‘‘carefully constructed self’’ that Cocking and Matthews
identify as preventing friendship formation online (2000),
while increasing the degree to which participants in
immersive virtual worlds engage in ‘‘non-voluntary self
disclosure’’, which they argue is valuable as a component
of the relational self developed in friendship (Cocking and
Mathews 2000). Unlike the chat room participant, or the
carefully sculpted public image someone presents on
Facebook, characteristics of participants in immersive
virtual worlds become apparent through actions. Leaders
take charge, the impetuous or impatient push ahead, the
careful and considered ask for advice, or discuss options to
overcome obstacles.
It may be argued against this claim that the kind of
activity undertaken in immersive virtual worlds is distinct
from that undertaken in the physical world, and this dis-
tinction is sufficient to render the connection between the
two unstable. For example, one who pursued this line
might think that there is something innate in performing
physical exercise that triggers friendship formation in a
way simulating that exercise does not. It is, however, dif-
ficult to see how this position could be maintained. Insofar
as the important aspect of shared activity is the interaction
with other real people, this is present both in the physical
world and in immersive virtual worlds. Through interaction
over time, you come to know the other people you engage
in these activities with, and the behavioural cues that
something is wrong or different are available in each
domain. It is, in other words, more difficult to hide your
personality when interacting in an immersive virtual world,
than it has been to do so in prior forms of online interac-
tion. Suppose, contrary to my earlier description of Jesse
and Kelly as having met while bicycling, that they met
instead inside the World of Warcraft. As new players in the
game, both create characters, and arrive in the first area in
which they can quest. They quickly discover the benefits of
co-operation, and together with a third player, Lee, they
quickly advance. Lee doesn’t want to talk, preferring
instead to try and advance as rapidly as possible, while
Jesse and Kelly talk as they play, and genuinely enjoy each
others company. Next week, Lee has moved on, while
Jesse and Kelly are still at around the same level. They
choose to level together again, and each takes a comple-
mentary role, such that their characters benefit from being
together. As they progress in the game, so to their friend-
ship develops. They come to know when the other is likely
to be online, and to communicate when this will change, in
order to allay concerns. While engaging in the shared
activity of gaming, they disclose their other interests and
goals, and discover that they share these also. They follow,
in other words, a path parallel to that followed by the
8 N. J. Munn
123
bicycling Jesse and Kelly discussed in Sect. ‘‘The charac-
teristics of friendship’’ above. I claim that this is just as
plausible a means for developing friendship as was the
bicycling example, and I hope that the considerations thus
far examined support this contention.
Some may also attempt to distinguish between the kinds
of activities that physical friendships lead to, and the kinds
that immersive virtual world friendships lead to. For
example, it has been suggested to me by an older academic
that ‘one would not invite an online friend to one’s wed-
ding’. This claim was intended, I take it, to point to a
meaningful distinction in the nature of the friendships
developed online by comparison to those developed in the
physical world, such that even if it were the case that
something properly called friendship could develop in an
immersive virtual world, the friendships therein developed
would be of a different kind to the friendships developed in
the physical world. Second-tier friends, who you keep out
of sight of your ‘real’ friends. For at least some people my
age, and I suspect for more as you examine younger
audiences, this kind of division simply does not exist. I
have, in fact, been invited to a wedding by people I had
only known through an immersive virtual world, and the
data collected by Cole and Griffiths suggests that meeting
online friends in real life is very common, with 55% of
females and 38% of males reporting having met online
friends in real life (2007). I suggest, therefore, that it is a
consequence of the relative novelty of the medium, not of
the nature, that results in the current desire to privilege
physical over virtual friendships. As people age in the
presence of virtual worlds, relationships formed in them
will become more common, less surprising, and more
accepted.8
Conclusion
In this article I have argued that the foundational compo-
nent of friendship is shared experience, and that the other
two components commonly held to be required for
friendship, namely mutual caring and intimacy, predomi-
nantly arise through shared experience, rather than inde-
pendently of it. With this conclusion in mind, I have
examined four different kinds of social interaction, three
online kinds and one more familiar kind, the interaction we
are used to from the physical world. I have argued that both
the currently dominant social media paradigm of online
interaction, as exemplified by Facebook, and the previous
generations of online interaction through email, discussion
boards, instant messaging, chat rooms and newsgroups, fail
to provide a forum in which shared activity can take place,
and as such, they are incapable of independently providing
a realm in which friendships can develop. I concede,
however, that in conjunction with either interaction in the
physical world or interaction in immersive virtual worlds,
they are capable of helping to maintain and strengthen pre-
existing friendships.
I argued that the third of the online means of interaction,
participation in immersive virtual worlds, or MMORPGs,
and in World of Warcraft in particular, is capable of pro-
viding the kind of shared activity that is required for the
development of friendships. Further, that this is possible
without the involvement of physical world interaction, such
that friendships can develop amongst people who have
never met in the physical world, solely through their shared
online experiences. To make this argument I relied firstly
on the nature of the shared activity common to the
immersive virtual world and to the physical world, and
secondly on the relationship in which these two realms
stand to the other two areas of online interaction discussed,
social media and older forms of online interaction. This
possibility has potentially important ramifications for the
kinds of obligations we have to others in the world. It is
commonly held that friendships are capable of generating
special obligations, and the ability to form friendships
online would drastically increase the number and location
of those obligations that are held by regular participants in
immersive virtual worlds.
Finally, I argued that an inability to provide for the
development of new friendships on the part of social media
and the previous generations of online interaction must not
be conflated with an inability to protect, encourage and
strengthen friendships. I argued that all four of the para-
digms discussed in this work are able to be used to
strengthen pre-existing bonds of friendship.
References
Aristotle. (1998). Nicomachean ethics (W. D. Ross, Trans.). Oxford:
Oxford University Press.
Blizzard Entertainment: Press Releases. (2010). http://us.blizzard.com/
en-us/company/press/pressreleases.html?101007. Last Accessed
December 6, 2010).
Bratman, M. E. (1992). Shared cooperative activity. The Philosoph-ical Review, 101(2), 327–341.
8 This discussion has centred on friendship formation. The mainte-
nance of friendship through online means is less controversial, and
accepted even by those such as Cocking and Matthews who deny that
friendships can be formed online (2000). I believe but will not
discuss, that in all the realms considered here, friendships can be and
are maintained. Social media and other online communication allow
distant (pre-existing) friends to maintain contact, and provide each of
these friends with cues as to their friends wellbeing, that while
insufficient to serve to establish friendships, are more than sufficient
to maintain them. See Cocking and Mathews (2000) for support for
this position.
The reality of friendship within immersive virtual worlds 9
123
Briggle, A. (2008). Real friends. Ethics and Information Technology,10, 71–79.
Caspi, A., & Gorsky, P. (2006). Online deception. CyberPsychology& Behavior, 9(1), 54–59.
Cocking, D., & Kennett, J. (1998). Friendship and the self. Ethics,108, 502–527.
Cocking, D., & Mathews, S. (2000). Unreal friends. Ethics andInformation Technology, 2, 223–231.
Cole, H., & Griffiths, M. (2007). Social Interactions in massively
multiplayer online roleplaying games. Cyberpsychology andBehavior, 10(4), 575–583.
Cooper, J. (1977a). Aristotle on the forms of friendship. The Reviewof Metaphysics, 30(4), 619–648.
Cooper, J. (1977b). Friendship and the good in aristotle. ThePhilosophical Review, 86(3), 290–315.
Facebook.com. (2010). Statistics. http://www.facebook.com/press/
info.php?statistics. Last Accessed December 6, 2010).
Gilbert, M. (2000). Sociality and responsibility: New essays in pluralsubject theory. Lanham, MD: Rowman & Littlefield.
Helm, B. W. (2008). Plural agents. Nous, 42(1), 17–49.
Helm, B. W. (2009). Friendship. In E. N. Zalta (Ed.), The stanfordencyclopedia of philosophy (Fall 2009 Edition), http://plato.
stanford.edu/archives/fall2009/entries/friendship/.
Helm, B. W. (2010). Love, friendship and the self. New York: Oxford
University Press.
Leib, E. J. (2007). Friendship & the law. UCLA Law Review, 54,
631–707.
Mason, A. (1997). Special obligations to compatriots. Ethics, 107(3),
427–447.
McKenna, K. Y. A., Green, A. S., & Gleason, M. E. J. (2002).
Relationship formation on the internet: What’s the big attrac-
tion? Journal of Social Issues, 58(1), 9–31.
Nardi, B., Harris, J. (2006). Strangers and friends: Collaborative play
in world of warcraft. In Proceedings of CSCW ’06: Proceedingsof the 2006 20th anniversary conference on Computer supportedcooperative work, pp. 149–158.
Rorty, A. (1993). The historicity of psychological attitudes: Love
is not love which alters not when it alteration finds. In
N. K. Badhwar (Ed.), Friendship: A philosophical reader(pp. 73–88). New York: Cornell University Press.
Rosen, C. (2007). Virtual friendship and the new narcissism. The NewAtlantis, 17, 15–31.
Rouse, R. (2000). Game design: Theory & practice. Plano, Texas:
Wordware Publishing, Inc.
Scheffler, S. (1997). Relationships and responsibilities. Philosophy &Public Affairs, 26(3), 189–209.
Sherman, N. (1987). Aristotle on friendship and the shared life.
Philosophy and Phenomenological Research, 47(4), 589–613.
Sony Champions of Norrath Press Release. (2004). http://
championsofnorrath.station.sony.com/headset.jsp. Last accessed
December 6, 2010).
Telfer, E. (1971). Friendship. Proceedings of the Aristotelian Society,71, 223–241.
Thomas, L. (1987). Friendship. Synthese, 72, 217–236.
Tuomela, R. (1993). What is cooperation? Erkentniss, 38, 87–101.
Velleman, J. D. (1997). How to share an intention. Philosophy &Phenomenological Research, 57(1), 29–50.
Verhagen, H., Johansson, M. (2009). Demystifying guilds: MMORPG-
playing and norms. Breaking new ground: Innovations in games,
play, practice and theory. http://digra.org:8080/Plone/dl/db/
09291.50401.pdf. Last Accessed December 8, 2010).
White, R. J. (2001). Love’s philosophy. Lanham, MD: Rowman &
Littlefield.
Terra Nova. (2011). http://terranova.blogs.com/.
Yee, N. http://www.nickyee.com/index-daedalus.html.
Yee, N. (2006a). Motivations of play in online games. CyberPsy-chology and Behavior, 9, 772–775.
Yee, N. (2006b). The labor of fun: How video games blur the
boundaries of work and play. Games and Culture, 1, 68–71.
Yee, N. (2006c). The demographics, motivations and derived
experiences of users of massively-multiuser online graphical
environments. PRESENCE: Teleoperators and Virtual Environ-ments, 15, 309–329.
10 N. J. Munn
123