ArticlePDF Available

Abstract and Figures

Concepts of space are fundamental to our understanding of human action and interaction. The common sense concept of uniform, metric, physical space is inadequate for design. It fails to capture features of social norms and practices that can be critical to the success of a technology. The concept of ‘place’ addresses these limitations by taking account of the different ways a space may be understood and used. This paper argues for the importance of a third concept: communication space. Motivated by Heidegger’s discussion of ‘being-with’ this concept addresses differences in interpersonal ‘closeness’ or mutual-involvement that are a constitutive feature of human interaction. We apply the concepts of space, place and communication space to the analysis of a corpus of interactions from an online community. A novel measure of sequential integration of conversational turns is proposed as an index of mutal-involvement. We conclude that a concept of communication space is needed to account for key features of the organisation of interactions in this community.
Content may be subject to copyright.
Communication Spaces
Patrick G. T. Healey, Graham White, Arash Eshghi, Ahmad J. Reeves & Ann Light
Interaction, Media and Communication Research Group, Department of Computer Science, Queen
Mary, University of London, London E1 4NS, UK (E-mail: ph@dcs.qmul.ac.uk)
Abstract. Concepts of space are fundamental to our understanding of human action and
interaction. The common sense concept of uniform, metric, physical space is inadequate for
design. It fails to capture features of social norms and practices that can be critical to the success of
a technology. The concept of placeaddresses these limitations by taking account of the different
ways a space may be understood and used. This paper argues for the importance of a third concept:
communication space. Motivated by Heideggers discussion of being-withthis concept addresses
differences in interpersonal closenessor mutual-involvement that are a constitutive feature of
human interaction. We apply the concepts of space, place and communication space to the analysis
of a corpus of interactions from an online community, Walford, which has a rich communicative
ecology. A novel measure of sequential integration of conversational turns is proposed as an index
of mutal-involvement. We demonstrate systematic differences in mutual-involvement that cannot be
accounted for in terms of space or place and conclude that a concept of communication space is
needed to address the organisation of human encounters in this community.
Key words: human interaction, online communities, communication, phenomenology
1. Introduction
This paper attempts to extend a trajectory set by Harrison and Dourish (1996)
who argued that the design of collaborative environments would be more
effectively supported if it focused more on the concept of placeand less on the
concept of space. The basic rationale behind Harrison and Dourishs argument
is that human activities are primarily organized in terms of the social practices
associated with a location not the physical properties of the space. For example,
an arena can equally well be used for a football match, a concert or a religious
service. Each event is subject to the same basic constraints imposed by the
physical space but each is distinguished by different senses of place. These
places are constituted by the social norms in each relevant sub-community
football fans, concert-goers, religious congregations. These norms determine
when, for example, singing and applause are (in)appropriate and the different
meanings assigned to the physical location e.g. as home endor stageor altar.
As a result, many basic design issues e.g., safety, lighting, layout can only be
addressed through an understanding of the place to be created.
Computer Supported Cooperative Work (2008) 17:169193 © Springer 2007
DOI 10.1007/s10606-007-9061-4
Our aim is to extend this line of argument by developing a concept of
communication space. The basic rationale is the same. We propose that there are
important differences in the quality of human interaction in degrees of inter-
personal, as opposed to physical, closeness that are important for the
organization of human activities and, consequently, for design. These differences
in inter-personal closeness are independent of both space and place. Just as the
same space can be associated with a variety of places, so the same place, we will
argue, can be associated with a variety of communication spaces.
The conceptual foundations for this proposal are drawn from Heideggers
(1927 [tr. 1962]) discussion of spatiality in Being and Time. The rst step in our
argument is to highlight the parallels between Harrison and Dourishs
distinction between spaceand placeand Heideggers distinction between
Cartesian space and regionsof equipment (c.f. Dourish 2001). Heidegger
develops this initial distinction further by highlighting an additional basic feature
of human-being
1
the solicitudewe have for one another. This, we will propose,
motivates a third concept of space that is constituted neither by simple physical
extension nor by the considerations of placebut by qualitative differences in our
closeness to others; being-withas opposed to being there(cf. Hollan and
Stornetta 1992).
This is an issue that, we believe, has not been given sufcient attention in the
design literature. The emergence of social computingand the success of
services such as MySpace, multiplayer games such as World of Warcraft, and
online communities such as Mixi demonstrate its timeliness. As Kang succinctly
puts it the killer applicationof the internet turns out to be other human beings
(2000 p. 1150, in Bargh and McKenna 2004).
Our concern here is not with the fact that people are connected nor with the
specic cues they have to one anothers activities per se, but with the inter-
personal character of their connections to others. This is distinct from the
concepts of shared presence and awareness as they are normally used in CSCW.
Our sense of closeness to others can vary independently of the kinds of
information we have available about, say, each others expression, persona,
location or activity in an environment (see e.g. Fitzpatrick 1998). It is also
distinct from a mutual-intelligibilityconcept of inter-subjectivity as knowledge
of shared meaning (e.g., Hollan and Stornetta 1992). Mutual-intelligibility does
not guarantee closeness and closeness does not guarantee mutual-intelligiblity.
The character of communication spaces, of being-with others, should be
understood, we will argue, as an independent level of analysis. The importance
of Heideggers phenomenology for present purposes is that it keeps the central
1
For expository reasons we use human beingrather than Heideggers term Dasein. This risks
confusing Heideggers position but is intended to make the discussion more accessible.
170 Patrick G.T. Healey et al.
importance of this aspect of human interaction in focus and treats it as a
fundamental, irreducible feature of human-being.
2
The empirical focus of our analysis is the communicative ecology of a text
based online community called Walford. As Harrison and Dourish (1996) argue,
text based MUDenvironments are particularly interesting in this context since
they are constituted almost entirely as places not spaces (p.74). Indeed, part of the
interest of Walford is that although it was originally constructed around a strong
spatial metaphor, the residents have progressively re-congured the environment
to overcome the room based spatiality of the original MUD architecture.
Our empirical analysis begins with a detailed discussion of how the
communicative infrastructure in Walford evolved. We go on to present an
analysis of interactions in Walford and introduce a novel measure of the
sequential integration of turns based on ellipsis. We use this and the distribution
of speech act types to demonstrate systematic differences in mutual-involvement
that are independent of both virtual space and place. We argue that these
differences correspond to differences in communication spaces that are
distinguished by the closenessof interaction.
We begin by highlighting the continuities, and some discontinuities, between
Heideggers discussion of spatiality and the spaceplacedistinction. We then go
on to consider the distinctive characteristics of being-withthat motivate our
concern with communication spaces.
2. Heideggers discussion of space and spatiality
Concepts of location provide some of the most basic structures for framing our
understanding of human action and interaction. The standard common sense
concept of physical space, usually attributed to Descartes, has two aspects.
Firstly, it has a structure which is quantitative, uniform, and which can be
specied in terms of a coordinate system. Secondly, it is something in which
human activities take place, but which is constituted independently of those
activities. As Harrison and Dourish (1996) noted, this Cartesian concept of space
is a cornerstone in the design of many collaborative systems.
Heidegger attacks both of these views: he maintains that, although space can
be given a quantitative structure, its basic structure is entirely qualitative and he
also grounds his conception of space in an account of human practice. This is a
reversal of conventional ontological priorities. It is the structure of everyday
human activities and concerns that provides the ontological foundations on which
concepts of physical objects located in space are built.
2
It is worth noting that this denes a key divide between Heideggerian and Husserlian
phenomenology and, consequently, distinguishes Heideggersaccountofbeing-withfrom
Schutzs concept of reciprocity of perspectives(Schutz 1954). Schutzs concept is intended as a
practical idealisedsolution to the problem of how inter-subjectivity is possible. This view has in
turn inuenced CSCW through the ethnomethodologcial tradition.
171Communication Spaces
For Heidegger, a primary focus on Cartesian concepts of spaceand distance
distorts our understanding of the way space and distance are actually
encountered. Our everyday sense of nearor faris more directly related to
our current concerns than by metric distance. For example, although the glasses
on your nose are physically close they are experienced as more remote [...] than
the picture on the opposite wallthat you are admiring (1962, p.141). Or, to use
another of Heideggers examples, although the pavement may be physically close
to my feet it is experienced as more remote than a friend I can see approaching
(1962, p.142). Because we are always already immersed in some course of
practical activity viewing pictures, meeting friends we do not normally
encounter space in purely Cartesian terms. To use the example above, an arena is
always encountered from within the horizon of our concerns as football fans or
concert-goers or builders or architects or events managers. There is no null
contextfrom which we encounter the abstract physical space of the arena.
Rather, that is what is left when we strip away the practical activities and
concerns that permeate our encounters.
It is because of this missing phenomenological context that Heidegger, and for
similar reasons Harrison and Dourish, see the common sense conception of
metric physical space as too limited to support an adequate understanding of the
organisation of human activity (c.f. Dourish 2001).
3
Instead, we need to focus
directly on how human activities are organized into courses of practical activity
and how these determine the spatiality, the nearness and farness, of what is
around us.
Things are not primarily encountered as isolated objects in particular physical
locations but as equipment
4
involved in some practical activity. Equipment is
organized in terms of regionsof involvements that are dened by our practical
concerns. A screwdriver has its placein a region of a workshop along with
other tools. These regions are organised with respect to activities such as making
a chair or opening a tin of paint. The screwdriversnearnessvaries as a function
of its availability and appropriateness for a particular ongoing course of action
not its position in physical space. A screwdriver that is readily available as it
becomes needed is experienced as nearbut one that is missing from its place or
out of reach or malfunctioning is experienced as far’–even if it is, from a
Cartesian perspective, physically nearby.
3
It should be noted at the outset that the match is not exact. For example, Harisson and Dourish
highlight relational orientation and reciprocity as a feature of their concept of space. For Heidegger
these are aspects of placeand the organization of equipment into regions(see e.g., Being and
Time §§2223).
4
Heideggers term equipment here does not refer just to tools engineered for some purpose. It is
intended to highlight the way in which all objects, including sticks and rocks, are primarily
encountered in terms of their (in) appropriatness as equipmentfor particular activities.
172 Patrick G.T. Healey et al.
For Heidegger it is the regions associated with practical activities that dene
the basic forms of spatiality; of places and directionality. Although it seems
natural to think of involvements and activities primarily as personal projects,
regions are not private cognitive maps of parts of an independently existing
world. For Heidegger they are inherently shared, public spaces. In part this is
because activities are typically done with, for, or against someone (1962, p.153).
Being a football fan or a concert-goer or part of a religious congregation depends
on being embedded in the right way in the activities and involvements of others.
More fundamentally, what it is to be a football fan or concert-goer or member of a
congregation is constituted by the shared understanding of how one should be
embedded in these activities and involvements not individual intention or
stipulation. To use a less obviously public example, the relative handinessof
a screwdriver for screwing or can-opening is not just a matter of its handiness for
one person alone. If it is handyfor one person it is handy for others doing the
same task (c.f. Mulhall 2005).
This is where Heideggers concept of region and Harrison and Dourishs
concept of place make contact. Places are dened by the patterns of behaviour
associated with some particular cultural understanding or set of practices. These
are public practices the communally-held(1996, p. 70) norms about what is
expected or appropriate. One point of contrast here is that whereas Harrison
and Dourish describe placeas secondary to or in space, as a space with
something added(1996, p.69), for Heidegger Cartesian space is a privative,
abstracted version of regions of concern; in effect, as placewith something
subtracted. For Heidegger, the Cartesian concept of space is thus secondary to
and supervenient on human practical activities and concerns. Despite this
contrast, the underlying point for design is the same. We must attend to the
social organization in terms of placesand regionsnot just the structure of
physical space in order to understand factors inuencing the (in)effectiveness of
technological interventions.
2.1. Being-withand solicitude: the structure of communication spaces
Heideggers discussion of equipment thus leads him to argue that the spatiality of
regions their sense of directionalityand their topology of places is a shared,
public spatiality. The individual sense of space that someone might have and
Cartesian physical space are both derivative from this more fundamental spatiality
of equipment and regions (see Dreyfus 1991 for further discussion). Moreover, we
encounter others through regions not as an adjunct to the equipment and
activities with which we are concerned but as integral to them; the world is
always the one that I share with Others(original capitals, 1962, p.155).
This leads, in Being and Time, to a consideration of being-withas a
fundamental existential characteristic of human-being. The structure of the
discussion parallels that for regions. Whereas the spatiality the nearness and
173Communication Spaces
farness of equipment is dened by concern, the spatiality
5
of being-with
others is dened by solicitude; the ways in which we can matterto one
another (1962, p.158). Just as equipment may be transparently available or
encountered in different modes of availability, so Heidegger distinguishes
different modes of solicitude. Solicitude is guided by considerateness and
forebearance(1962, p.159) and also has privative or decientmodes such as
indifference; The Other can be missing only in and for a Being-with(1962,
p.157, original italics).
This is an ontological claim about the nature of human-being not about the
particular contingent facts of some occasion. It doesnt matter to these
considerations if, as it happens, there are no other bodies near me in the same
physical space. No more than the fact that there are other bodies present around
me entails that I cannot feel alone. The same point applies to places. Even if we
had an exhaustive understanding of the social norms that constitute a football
game, or a concert or a religious meeting this would not exhaust the potential
variations in peoples closeness to (or alienation from) the others in those places.
It is worth noting that being-with is distinct from empathy or individual
cognitive reection on our relationships with others or even knowing oneself.
For Heidegger these are all secondary to and dependent on being-with others.
Being-with is an autonomous, irreducible relationship of Being(1962, p.162).
Solicitude is the existential ontological bridge to others and only on the basis of
Being-with does empathybecome possible(1962, p.162).
This discussion motivates our claim that an adequate understanding of human
interaction depends on addressing being-with as an independent level of analysis.
Heideggers argument provides a basic vocabulary for discussing these
phenomena and helps us not to loose sight of them as a basic feature of human
being. It also shows that analyses of placesand regionsare inevitably
incomplete in this respect because they do not address basic variations in the
qualitative character of human interaction; the differences in the nearness and
farness of others that constitute what we term communication spaces.
We can illustrate this by reference to Lights(2007) analysis of the experience
of mobile phone calls. When people are fully engaged with each other during a
call they describe this experience as being in a third spacedened primarily by
their sense of closeness. Space and place become conspicuous when this sense of
engagement breaks down e.g., when it becomes apparent that ones
assumptions about the other speaker are ungrounded (not holding their diary,
but driving), or perhaps, where background noises intrude (street noises or a
ushing toilet) that elements of the physical world or differences in place
5
Heidegger does not discuss solicitude in terms of spatiality. We adopt this usage here to make
the connection to the concept of communication spaceclear.
174 Patrick G.T. Healey et al.
become conspicuous. Technologies which alter the possibilities for connecting
people also alter the possibilities for maintaining different modes of being-with.
To provide a more systematic assessment of the empirical usefulness of this
concept we turn to the analysis of interactions in a text-based environment called
Walford. This environment is of particular interest because the residents have
progressively adapted their infrastructure to create a communicative ecology that
is signicantly richer than both commercial text-based group messaging systems
and avatar-based graphical environments (see, e.g., Brown and Bell 2004; Smith
et al 2000; Ducheneaut et al 2006). One consequence of this has been the removal
of all simple spatial constraints on interaction and the development of practices
such as remote emoting. We use detailed log analysis to assess the extent to
which concepts of space, place and communication space are needed to account
for the patterns of interaction observed in Walford.
3. Walford
Walford is a text-based online social community or talkerthat has been
established for more than a decade. It has approximately 1,500 regular users,
predominantly from North America and Europe. In January 2004 the daily
connection rate was 354 users with an average of 5 new users created per day.
The data for the analysis reported here comes from three sources. First, the
documentation of Walford including the user manuals, the version change logs
recording bug xes and changes to functionality and the usage statistics generated
by the program itself. Second, from contact with one of the developers of the
code, who set up additional logging facilities, and from our interactions with the
residents inside the environment. The third source of data, which is the main
focus for this paper, is a corpus of chat logs collected by specially developed
logging software over approximately one year (ignoring server crashes) in 2004
2005. For each person-to-person message sent the following data is captured: the
basic command type; unique numeric IDs for: the originator, the originators
location, the recipient(s) and their location. In order to protect the anonymity of
participants the names the environment, residents, characters, places, and
commands have been changed. Users agreed as a condition of use to the system
to the logs being used, in anonymised form, for the purposes of research and
publication (Figure 1).
Walford emerged as one of many variants on James AspenesTinyMUD
which was rst created in 1989.
6
One of the rst multi-user dungeon servers, it is
built around a strong spatial metaphor and four kinds of object: rooms, things,
6
Some documentation for TinyMUD can be found at http://www.mudmagic.com/codes/server-
snippet/2144.
175Communication Spaces
players and exits; all of which can have various properties associated with them.
Players can build rooms, furnish them with customized objects and even create
their own games and puzzles. This makes it possible for users to build up
relatively rich virtual environments and characters and subsequent versions have
progressively extended this capability. In Walford the spatial metaphor is
elaborated to include a complete town plan. There are shared public spaces such
as a high street, a pub, a townhall, a bank, a bus station. There is also a rubbish
dump and a network of roads.
Although Walfords basic architecture is built on a game it has developed into a
coherent social community in which the MUDs built-in game functions have
become subordinate to the social interaction. As will become clear below,
residentss main preoccupation is with their interactions and relationships with
each other. This point is illustrated by a sample of conversation topics from the
logs: outsourcing, mobile phones, births and deaths in users (real) families, their
childrens achievements, relationships (both inside and outside Walford),
chocolate, economics assignments, redundancy, taking sick days off work,
cross-stitch exhibitions, movies, bands and boredom. There is also a signicant
amount of discussion about the technical features of Walford itself, the
Number of Reci
p
ients
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Number of Messages
700
600
500
400
300
200
100
0
Distribution
local
mixed
remote
Figure 1.Number and location of message recipients for turns in one 24 hour period.
176 Patrick G.T. Healey et al.
competence or otherwise of its administrators and how its culture compares
with the social mores in other talkers.
One of the most interesting aspects of Walford is that its code base has evolved
through an iterative process of community development. The environment is
programmable by users and as they become more expert they can implement their
own soft codeon top of the hard code provided by the server. If new commands
or functions become sufciently popular they spread amongst the users and
eventually the developers incorporate them into the hard code. This has led to the
coevolution of a specialised command infrastructure and a unique communicative
ecology.
3.1. The communicative ecology of Walford
Table 1shows, the original set of TinyMUD communication commands: say,
whisper,gripeand pageplus a broadcast channel for system announcements.
These commands incorporate strong spatial constraints. The only means for a
player to communicate with another at a remote location is by using the page
command. This sends either a request for contact or a message but not both.
TinyMUD also provides a havenag (a property set on objects) that can be
used by a player to prevent other people from contacting them.
In Walford the communication infrastructure has evolved to become much
richer than TinyMUD. First, it incorporates a variety of synchronous and
asynchronous channels. There is a bulletin board system located in one of the
public spaces with topic threads, subtopics and facilities for anonymous and
remote posting. There is also a location independent chat channel system that
allows groups with common interests or concerns to sustain a more synchronous
form of topic-based remote interaction. This facility evolved out of the
Table 1. Communication commands in TinyMUD.
TinyMUD Command Behaviour
say <message> Says <message> to everyone in the room. Aliases: "and :. Can be
used for actions, e.g. if your name is Pat, and you type :waves,
everyone sees Pat waves
whisper <player>
<message>
Whispers the message to the named person, if they are in the same room
as you. Others in the room do not see the message.
gripe <message>. Sends <message> to the system maintainer
page <player>
[=<message>]
Tells a player that you are looking for them. They get a message telling
them your name and location. If you include the =<message>,it
will tell the player your name and your message, but not your location
@wall <message>. Only wizardsmay use this command. Shouts something to every player
connected
177Communication Spaces
administrators need to discuss issues related to system management. There is also
an internal email system, with an interface to external email systems, that
supports asynchronous interactions. It is clear from the logs that these channels
are also occasionally supplemented by the concurrent use of webcams,
commercial messaging applications, web pages and other connections external
to Walford.
The basic repertoire of chat commands has also been elaborated. The local
(same room) commands sayand whisperare extended to include think’‘ask
and emotewhich format messages as thoughts, questions or actions respective-
ly. In addition, two further commands provide forms of direct, i.e. addressee
specic, communication independent of location. One is an adaptation of the
original pagecommand. This allows private communication with both local and
remote addressees. Other people in the room only see a pagemessage if it is
directly addressed to them. There is also a tell<name> <message> command
that works in the same way as sayfor local addressees but can also be used to
send messages to remote addressees. Unaddressed participants at the local
location see the message but those at the remote location do not. Corresponding
to the page and tell commands are respond,resendand recallcommands that
make replying easier. A further extension of the emotecommand allows remote
emoting so that users can e.g. wave at each other from different rooms. Because
the emote command gives users signicant control over the formatting of text it
has subsequently been exploited to implement additional remote say,askand
thinkfunctionality and a range of more idiosyncratic forms of declarative
formatting (e.g., exclaim, tickle, hug).
Users have the ability to set up internal buddydistribution lists that can be
used as arguments of directed chat commands. There are also a number of ags
that support selective control of which communication commands people can use
to contact each other. Some of these ags are set by the users themselves and
some by system administrators (and occasionally used as a sanction). There are
also ags, like TinyMUDshavenag, that allow users to set global lters on
the kinds of message they receive. Finally, there are limited forms of natural
language processing that can be used to format turns.
The progressive, collaborative development of this rich communicative
infrastructure demonstrates the central importance of social interaction for the
residents. They have developed specic technical resources to support a variety
forms of interaction. The question that naturally arises is how these resources are
actually deployed by residents. We consider, in turn, the relevance of space, place
and communication space for understanding patterns of interaction in Walford.
3.2. Spacein Walford
One consequence of the changes to the commands made by the residents of
Walford is that there is no functional difference between communication between
178 Patrick G.T. Healey et al.
local and remote participants. In principle, everything that can be said or done
locally can also be said or done remotely. The residents have progressively
adapted their environment to the point where the original constraints on
interaction deriving from the underlying spatial metaphor have been removed.
Although certain resources, such as objects and games, are still tied to particular
locations, forms of communication are not.
In order to analyse actual patterns of usage, logs covering a ve month period
(ignoring downtime) in 2004 and containing just over three million messages
were selected. As Table 2shows, statistics on the distribution of turns in this
corpus show that users make extensive use of remote communication. More than
half of all messages are sent to recipients in different rooms.
As a result, interaction typically involves interleaving local and remote
exchanges. Excerpt 1, illustrates a typical exchange from the viewpoint of
Naomi. The rst four turns are the opening remarks in a local conversation
between Naomi and Sonia in Room A (note that turn 3 is a question although not
explicitly marked as one). At turns 5 and 6 Naomi initiates a conversation with
Garry who is in a different location. For this she uses directcommands addressed
only to Garry and so this exchange Garry responds in turn 8 would not be seen
by Sonia. At turn 7 a further conversational thread emerges. Frank, who is in a
different location to the others, announces to a larger group that includes Naomi,
Garry and Sonia that he is logging off. This is followed by a valediction from Frank
at turn 9. In turns 1011 Naomi and Sonia continue their conversation.
To assess how different this pattern is from face-to-face interaction we analysed
a random sample of 50 turn sequences from twenty-ve Walford users and an
equivalent random sample from the demographic section of spoken conversation
in the British National Corpus (BNC). This consists of spoken conversation
collected by people with portable tape recorders who record their own and their
friendsspeech over a period of up to a week. Conversations are recorded as
unobtrusively as possible, so that the material gathered approximates natural,
spontaneous speech.
7
Twenty-ve les consisting of the transcripts of a single
individuals conversations were selected and 50 turns transcribed.
This comparison shows that, on average, in a 50 turn sequence someone in
Walford will conduct three concurrent conversations whereas someone in face-
to-face interaction will conduct only one. This is of interest partly because the
text-based interface in Walford does not provide a natural way of grouping turns
into separate conversations. Rather, participants have a single window and have
to rely on formatting cues, scrolling and their own memory to track different
conversational strands.
7
All references to full names and addresses are removed from the BNC corpus and after each
conversation. All participants are advised they have been recorded and if they are unhappy about
the recording it can be erased.
179Communication Spaces
1. Local Room A.
Sonia: asks Hopws you?
2. Local Room A.
Naomi: tired
3. Local Room A.
Naomi: you
4. Local Room A.
Sonia: says Just got in from work
5. Direct Room A to B.
Naomi: heya hun
6. Direct Room A to B.
Naomi: logged in from home
7. Direct to Sonia, Naomi, Garry & four others:
Frank: says Ok, I m off again for now.
8. Direct Room B to A.
Garry: yep, just about to start my studying
9. Direct to Sonia, Naomi, Garry & four others:
Frank: chants "Yeah!, Yeah!, Yeah!!!!"
10.Direct Room A to A.
Naomi: aahh I see, kinda sick here and a bit pissed
11.Local Room A.
Sonia: asks "Drunk pissed?"
Naomi: angry pissed
12.Local Room A.
Excerpt 1: Sequence of 12 turns with three conversational strands. Turns are
in temporal order with indenting added to separate the conversational
strands.
A second contrast between Walford and face-to-face interaction that emerges
from comparison with the BNC relates to group size. Conversations in the BNC
are predominantly dyadic (54%) with occasional three or four party exchanges
and rarely anything larger. In Walford exchanges are still predominantly dyadic
but this is less common (32%) and proportions of group interaction are
correspondingly higher (see Figure 2).
These observations show that Walfords residents take advantage of the
exibility offered by the modied command set to circumvent the constraints that
the spatial metaphor places on interaction. The typical pattern of exchange
involves multiple concurrent conversations among partially-intersecting groups of
participants in a variety of locations. Face-to-face interaction is structured, in part,
by the spatial, temporal and practical problems involved in getting people
Table 2. Chat turn distribution in a sample of 3,018,191 messages.
Message distribution Frequency Percentage (%)
Local 578,633 19
Direct (same room) 704,027 23
Direct (different rooms) 1,735,531 58
180 Patrick G.T. Healey et al.
together in the same place. In Walford these constraints are removed and remote
communication is, ostensibly, just as popular as local communication.
One further observation helps to underline this point. In face-to-face interaction
a whisper is, all things being equal, a compromise between the desire to
communicate something privately and the potential visibility and therefore
detectability of the exchange to others (ignoring for now the use of staged
whispers as a device to provoke the interest of others). Although the whisper
command was one of the three basic commands for addressing others in
TinyMUD it is used for less than 0.04% of turns in Walford. If people wish to
communicate privately they can do this without any risk of detection by others
and thereby avoid one of the compromises of co-present exchanges.
3.3. Place in Walford
Although the spatial metaphor incorporated into its original architecture has
become obsolete as an organizing factor for communication, Walford is also
structured by a placemetaphor. The environment is divided into a number of
1e-06
1e-05
1e-04
0.001
0.01
0.1
1
1 10 100 1000
Log Proportion of Total
Lo
g
Messa
g
e Len
g
th
Direct Remote
Direct Local (sameroom)
Local
Figure 2.Changes in message size and frequency in local, direct (same room) and direct
remote interactions (n=3,025,979).
181Communication Spaces
public and private spaces individual rooms, town hall, rubbish dump which
have a strong semantics of place. Key amongst these is Walfords pub the
Queen Vic. This is one of the most frequently used locations in the environment
and there is evidence that the sense of placecreated here has an effect on the
kinds of interaction that take place.
Although on average residents make equal use of local and remote
communication, it is clear from Figure 2that this varies as a function of group
size. Remote exchanges are skewed towards dyads and there are a correspond-
ingly smaller proportion of remote group exchanges, particularly for group sizes
of two to seven recipients. They account for a higher proportion of local
interactions (69%) than remote (36%). This indicates a preference for conducting
small group interactions in the same virtual space. In fact, all of the local small
group interactions in the 24 hour sample take place in the Queen Vic.
Further evidence of an effect of placeon interaction can be seen in Figure 3.
This shows that for messages of between 10 and 50 words length there is a small
but highly consistent length advantage for local exchanges (command names and
their arguments are not included in the calculations of message length). This is
23310751849N =
Reci
p
ient Locations
MixedRemoteLocal
95% CI Frequency of Cross-Turn Ellipsis
.4
.3
.2
.1
0.0
Figure 3.Cross-turn ellipsis to different recipient distributions.
182 Patrick G.T. Healey et al.
not a consequence of constraints on the type of commands used since the same
commands can, in principle, be used both locally and remotely. It is also notable
that direct same roominteractions are intermediate between the local and
remote interactions indicating that this is also not a consequence of control over
message distribution or addressee specicity.
These observations indicate that the concept of place is useful for explaining
patterns of communication in this environment. This is reected, in particular by
the apparently distinctive character of small group interactions in the Queen Vic
(see Excerpt 6 below). However, we now go on to argue that neither spacenor
placeprovide a fully adequate account of interaction in Walford.
3.4. Communication spaces in Walford
The distinction between local and remote exchanges is problematic for a place
based account. Although the characteristics of local exchanges can potentially be
characterized by reference to the type of place where they occur, remote exchanges
occur in what amounts to a nullor non-place. Treating this as a special case of
place seems strained since it lacks well dened social norms of a kind comparable
to those that dene, for example, a pub conversation. Perhaps more importantly, the
preponderance of dyadic remote interactions (Figure 2) suggests that there are also
distinctions between types of remote interaction that a generic concept of nullor
non-place is unable to address. We return to this point below.
A second problem arises because of the possibility of mixedgroup interactions
in which some recipients are local to the speaker and some remote. All things being
equal we would expect these to have the same character as remote interactions
since they depend on exactly the same set of commands. However, there is
evidence that the residents are sensitive to the remote/mixed distinction.
As Figure 2illustrates mixed interactions are much less frequent than might be
expected by chance. In the 24 hour sample there are on average 30 times as many
possible remote addressees as possible local addressees. The most common
distribution pattern for multi-person messages should therefore be to remote
participants only, the least common to local participants only with Mixed
distributions falling between. In fact, 59% of group messages have exclusively
local distributions, 34% exclusively remote and only 7% mixed. It appears that
residents both systematically favour local group interactions and systematically
avoid mixed (local and remote) interactions.
Turn length and recipient distribution provide only gross measures of interaction
patterns. In order to develop our analysis of communication spaces we need measures
that index mutual-engagement or closenessmore directly. One intuitive measure is
the distribution of speech act types (e.g. talkvs. thinkvs. emote)indifferent
contexts. In addition, a novel, direct measure of the sequential integration of turns can
be developed by looking at the frequency of cross-turn ellipsis.
183Communication Spaces
3.4.1. Measuring sequential integration
Cross-turn ellipsis occurs where the meaning of one turn depends in a particularly
direct way on a constituent from another turn. For example, in Excerpt 1,
Naomis question youto Sonia in turn three can only be fully interpreted in the
context of the preceding hows you?The meaning of Naomis turn thus depends
in part on an elliptical constituent from another turn by another person (Sonias
hows) for its sense. Another simple example is provided in Excerpt 2 where the
meaning of Bsitis provided by Ashoney and lemon.
A : says "Take some honey and lemon too."
B : yeh I need it
Excerpt 2: Example of Cross-turn Anaphoric it. Local Dyadic Interaction
This phenomenon is useful here because it provides an intuitively valid
measure of the strength of the ties between different conversational turns. People
have a choice about whether to use of elliptical or fully articulated versions of a
turn as they do not alter its basic meaning. However, this choice does reect
assumptions presumably tacit about how easily other participants will be able
to retrieve the relevant context. The assumption is that the more we make direct
use of parts of each others turns the more closely integrated our contributions
are.
A : says "Grandfather died on Monday"
B : sorry to hear that
Excerpt 3: Cross-turn anaphoric that, Mixed Group Interaction.
There are a variety of different forms cross-turn ellipsis can take. Excerpt 3
illustrates a cross-turn anaphoric use of that(other examples are he”“she
him”“his”“her”“they”“their). Excerpt 4 provides another example of verb
phrase ellipsis in turn 3 I dont[come in here] and also an example of self-
repair where turn 3 is a re-working of turn 2 by the same speaker. Although repair
is not strictly ellipsis we include it here because it also signals a strong
dependency relationship across turns. Excerpt 5, which is a later continuation of
the sequence introduced in excerpt 1 provides further examples of self-repair,
short answers, complement ellipsis (turn 27 lucky you[to be home]) and
sluices (turn 36 why[lucky me/why do you ask]).
A : says "Take some honey and lemon too."
B : yeh I need it
184 Patrick G.T. Healey et al.
1. A: did not know you came in here
2. B: no I didnt
3. B: I don t
4. B: first time
Excerpt 4: Verb Phrase Ellipsis (2,3) and Self-Repair (3,4). Remote Interaction
23.Local Room A.
Naomi: you home from work noe
24.Local Room A. (Self repair noe )
Naomi: thinks now also
25.Local Room A. (Affirmative Polar Answer)
Sonia says yeah
26. Remote Room B direct to Sonia
Garry: I just read my next assignment to see what I
need to know
27.Local Room A. (Complement)
Naomi: hehehe lucky you
28. Remote Room B direct emote to Room A & C
Garry: hugs Sonia and Ruby
29. Mixed Room C direct emote to Sonia, Naomi, Garry,
Phil + 2 others
Ruby: hugs Garry
30. Mixed Room A direct emote to Sonia, Garry, Ruby,
Phil.
Naomi: hugs Garry
31. Remote Room D direct to Sonia, Naomi, Ruby, Phil +
2 others (not Garry)
Pat: thinks ƒ thankyou Peggy!
32. Mixed Room A direct emote to Pat, Sonia,
Garry,Ruby & Phil.
Naomi: licks Pat (Yeah!!)
33. Mixed Room C direct emote to Sonia, Naomi,
Garry, Phil + 3 others
Ruby: hugs Pat (Yeah!!)
34. Remote Room D direct to Sonia, Naomi, Ruby,
Phil + 2 others (not Garry)
Pat: says "Just on to check bingo
35. Remote Room D emote to Sonia, Naomi,
Ruby, Phil + 2 others (not Garry)
Pat: spanks and licks Naomi, Ruby, Sonia,
Phil, Guss and Dawn
36.Local Room A. (Sluice)
Sonia: asks "why?"
37.Local Room A. (Short Answer)
Naomi: just bored so asking.
Excerpt 5: (continuation of excerpt 1) Local Mixed and Remote Exchanges
with Examples of Cross-turn Ellipsis
All the turns from the 24 hour sample shown in Figure 2were coded for cross-
turn ellipsis following a simplied version of Fernandez et al (2004) taxonomy.
This includes, clarication ellipses, anaphora, short answers, prepositional
modiers, verb and noun phrase ellipsis, complement ellipsis, yes/no (polar)
answers and sluices. In addition we also coded for instances of cross-turn self
185Communication Spaces
repairs. For convenience we consider the analysis of group interactions rst
followed by dyadic interactions.
Group interactions:
As Figure 4indicates, the local, remote and mixed groups of participants use
different amounts of ellipsis. Turns to groups of local participants have a higher
proportion (26%) of cross-turn ellipsis than turns to groups of remote participants
(21%) and they have a higher proportion of cross-turn ellipsis than turns to
groups with mixtures of local and remote participants (12%). These differences
are statistically reliable (t-test for equality of means not assuming equal variances.
Local vs. remote: t
(2369)
=2.8, p(two-tailed)=0.00; remote vs. mixed: t
(412)
=3.8, p
(two-tailed)=0.00. Local interactions have a tighter sequential integration than
Remote interactions which, in turn, have a tighter sequential integration than
Mixed interactions.
For the analysis of speech act/command types each turn was coded for its
declarative formatting. To simplify the presentation these categories are collapsed
into four classes: ask,talks(page/say/tell) emotes(smiles, laughs, hugs, kisses,
waves, and specially formatted turn types such as yell)andthinksand asks.
S
p
eech Act T
yp
e
thinktalkemoteask
Percentage of Turns
80
60
40
20
0
Recipients
local
remote
mixed
Figure 4.Relative proportions of speech act types to local, remote and mixed distributions of
recipients.
186 Patrick G.T. Healey et al.
Mutlinomial regression analysis shows that the overall distribution of speech
act types in Figure 4varies with recipient group (global effect of recipient
location: Chi
2
(6)
=269.8, p=0.00, n=2,880). Given the relatively low use of
thinkand askspeech acts types, for focused contrasts only the emoteand
talktypes in the local, remote and mixed recipient groups are compared (local
vs. remote: Chi
2
(1)
=177.3, p=0.00, n=1,661; remote vs. mixed: Chi
2
(1)
=0.4, p=
0.51, n=679). Overall, mixed and remote groups make greater use of emotes
and proportionally less use of talkturns than local groups.
This pattern of interaction is illustrated by the exchanges in Excerpts 1and 5.
The local dyadic exchange between Sonia and Naomi involves no emotesand
displays a high level of sequential coherence including self-repairs and elliptical
expressions. In contrast to this the two strands of remote group interaction consist
almost exclusively of emotes(hugs, licks and spanks) or turns that do not
project for a response e.g. no direct questions. It also includes no self-repair or
ellipsis. Excerpt six provides a short contrastive illustration of the relatively
tightly integrated turn sequences characteristic of local group interaction.
Kevin: i'm bored
Billy: me as well
Ian: dances for your amusement
Kevin: thanks
[6 intervening turns of remote dyadic interaction]
Kevin: sooo
Billy: help me at choosing a new name
Denise: exclaims ""Now way!werdo!""
Excerpt 6: Local Group Interaction in the Queen Vic
Dyadic Interactions:
As noted above, dyads are an exception to the basic local remote contrast.
Although local group interaction shows a higher sequential integration than
remote group interaction, as Figure 5shows, the remote and local dyads are not
reliably different (t
(109)
=0.94, p[two-tailed]=0.36, equal variances not assumed).
Two further focused comparisons show that local dyads have higher sequential
integration than local groups (t
(544)
=2.60, p[two-tailed]=0.01, equal variances
not assumed) and, similarly, remote dyads have higher sequential integration than
remote groups (t
(1079)
=8.49, p[two-tailed]=0.00, equal variances not assumed).
As Figure 6shows, the distribution of speech act types for local and remote
dyads is also more similar than it is for local and remote groups. Dyads rely
predominantly on talkcommands and much less on emotesthan groups. A
focused, multinomial regression analysis comparing the relative frequency of
emoteswith talksin local and remote dyads shows no reliable difference
(Chi
2
(1)
=0.5, p=0.48, n=925).
187Communication Spaces
4. Discussion
The residents of Walford have developed a complex infrastructure for interaction
and a correspondingly diverse communicative ecology. This highlights the central
importance of human interaction to the participants in this environment. Life in
Walford appears to be more richly social, i.e. focused on the relationships
between residents, than that observed in other, graphically sophisticated, multi-
person environments (e.g., Brown and Bell 2004; Smith et al 2000; Ducheneaut
et al 2006). Although it is tempting to view text-based communication as one of
the most attenuated forms of human interaction, in important respects interaction
in Walford goes beyond what is possible in other environments and modalities
(cf. Kendall 2002).
One of the fundamental issues which this work raises is the following: in what
sense is the communicative ecologyof Walford still spatial? Although Walford
and its predecessors were constructed around a strong underlying spatial model,
intended as a metaphor for physically situated face-to-face interaction, this has
been progressively eroded by its residents through adaptations of practice and
technology. The extended repertoire of communication commands, each with
remote versions, highlights how, given the opportunity, people escape the spatial
647379N =
Reci
p
ient Locations
Remote DyadsLocal Dyads
95% CI frequency of Cross-turn Ellipsis
.40
.30
.20
.10
0.00
Figure 5.Cross-turn ellipsis in dyadic interactions.
188 Patrick G.T. Healey et al.
constraints of face-to-face interaction. Residents routinely maintain multiple
concurrent conversations, among large, partially overlapping groups of partici-
pants, in a variety of (virtual) locations. By contrast face-to-face interaction is
predominantly dyadic and rarely involves more than three or four participants. It
is clear that (virtual) space is not a signicant organizing factor in interactions in
Walford. In Hollan and Stornettas(1992) terms, residents communicate in ways
that go beyond just being there.
It is of course, possible to have qualitative, non-uniform concepts of space. For
example, we use qualitative notions of a place when we reason about ordinary
conversations: there is an everyday notion of Fwhere a conversation takes place,
which is clearly meaningful even though it need not be precisely or metrically
dened. Harrison and Dourishs concept of placeand Heideggers concept of
region provide such qualitative accounts of spatiality. However, there are aspects
of interaction in Walford that are difcult to account for by reference to placeor
regionalone. Although local group interactions are closely associated with a
single place the pub dyadic interactions have a common character regardless
of whether they occur in the same virtual place. In addition, the residents clearly
distinguish between mixed and remote interactions even though they both lack a
S
p
eech Act T
yp
e
thinktalkemoteask
Percentage of Turns
80
60
40
20
0
Recipients
local
remote
Figure 6.Use of Speech Acttypes by dyads in local or remote interactions.
189Communication Spaces
well-dened place and are mediated in exactly the same way. These distinctions
between dyad and group and mixed and remote do not depend on differences in
specic social norms of the kind that distinguish e.g. football matches from
concerts or religious meetings.
An alternative way to make this point is by considering how we normally
reason about the relations between people and places. If Ann is in the same place
as Bob, and Bob is in the same place as Claire, then Ann is in the same place as
Claire. It seems difcult to t any such notion of place onto the pattern of
interactions in Walford, however qualitative and non-Euclidean our understand-
ing of the virtual space or place might be.
Our argument is that we can preserve these kinds of natural inferences and get
a clearer understanding of how communicative resources media, environment,
artifacts are used in practice by using Heideggers concept of being-with to
guide our analysis. This, we propose, translates into a key organising principle in
human interaction: the development and maintenance of interpersonal closeness
or of different kinds of being-with. Communication spaces are thus operationally
distinguished by sets of participants and their relative level of mutual
involvement. The empirical utility of this approach, we claim, is that it uncovers
a practical organising factor in social life in Walford, one that is made manifest in
the differences between dyadic, local group and remote group interactions.
The dyadic communication space is characterized, we propose, by high levels
of mutual-involvement. This corresponds to relatively high levels of sequential
integration of turns and greater use of talkcommands rather than emotes.
Treating emotesas indicative of lower levels of mutual-engagement may seem
paradoxical however the Excerpts suggest that in Walford emotesfrequently
function as outloudsor response cries utterances that are designed to keep
others aware of whats going on but without directly engaging with a specic
addressee (Levinson 1988; Goffman 1981). Excerpts 1and 5illustrate this use of
emotes for rounds of greetings and to announce entrances or exits. In contrast to
this dyadic interactions such as those between Sonia and Naomi, in Excerpt 1, seem
to involve a greater commitment to the development of the unfolding exchange.
The local group communication space is characterized, we propose, by an
intermediate level of mutual-involvement. Local (and remote) dyads show higher
levels of sequential integration than local group communication. However, as
Excerpt 6 illustrates, the local groups are still engaged with the conversation but
the progression of topics seems less controlled than in the dyadic interactions.
The remote group communication space is characterized by low levels of
mutual-involvement. This form of interaction shows consistently low levels of
sequential integration of turns and a relatively high proportion of emotes. The
extracts indicate that this form of interaction seems to be primarily used for
maintaining group awareness of whats going on rather than direct engagement.
Interactions in these communication spaces share a relatively atparticipant
structure. People communicating face-to-face have to contend with the potential
190 Patrick G.T. Healey et al.
involvement of unintended recipients e.g. overhearers, bystanders and eaves-
droppers and this has consequences for the way they design their contributions
(Goffman 1981; Schober and Clark 1989). In Walford, overhearing, bystanding
and eavesdropping are only possible if people choose to use commands that allow
it. As the frequency of use of the whispercommand indicates, residents take
advantage of the opportunity to limit these possibilities.
Interactions in these communication spaces also promote parity of access. In
each case participants have mutually equivalent access to the conversational oor
and to each other. We hypothesise that mixed interactions are avoided, in part,
because they combine subsets of people with different kinds of access to each
other. One possible reason is that this would multiply the opportunities for, by-
play, collusion and innuendo (Goffman 1981).
A salient feature of Walford is the way people maintain parallel conversations
in different communication spaces. For individuals this can create complex
patterns of participation across the different spaces. For example, in Excerpt 5
Sonia is a ratied side participant to Naomis hug of Garry (turn 30). However,
she is also an overhearer to it in the parallel dyadic conversation she is having
with Naomi. Sonia and Naomis participant status in both of these strands is
mutually known to each other. However, Garry is also in a dyadic interaction with
Sonia that Naomi has no direct access to. Sonia therefore stands in at least three
different relations to Naomis hug, some mutually known to the other participants
and some not. A key practical problem for residents is managing these parallel
tracks. In contrast face-to-face interaction, there is no mutually agreed pattern of
domination/subordination between (sub)conversations in part because they
have disjunct sets of participants and are therefore not integrated within a single
mutually known participant framework.
5. Conclusion
We have proposed that the concept of communication space provides a useful
approach to thinking about the basic organization of human interaction. We have
identied four empirical criteria for distinguishing between communication
spaces: group of participants, sequential integration of turns, distribution of
speech act types and parity of access. These criteria are independent of space and
place although correlate with them. We have further argued that what underpins
these differences in communication space are not reducible to differences in space
and place but rather are due to differences in the forms of being-withthat people
create with each other. The technology of virtual spaces is encountered, we
propose, not just as a version of space in the physical world, nor just in terms of
places, but also in terms of the type of human encounters that it enables or
impedes; the ways in which people can bring each other close, keep each other at
a distance or be therefor each other. Peoples need for different levels of mutual
involvement on different occasions is a key factor driving the differentiation of
191Communication Spaces
instruments and practices into distinct spaces of interactional possibilities. We
have argued that the concept of communication space is needed to bring these
phenomena into focus and provide a better foundation for reasoning about and
designing for them.
Acknowledgements
The authors are grateful to Ivan Leudar, Greg Mills and Chrystie Myketiak and
the anonymous referees for their comments on earlier drafts. We are also grateful
to the residents of Walford for giving us access to their community.
References
Bargh, J.A., McKenna, K.Y.A. (2004): The Internet and Social Life. Annual Review of Psychology,
vol. 55, pp. 573590.
Brown, B., Bell, M. (2004): Social interaction in there’’ CHI 04 extended abstracts on Human
factors in computing systems. Vienna, Austria, pp. 14651468.
Dreyfus, H.L. (1991): Being-in-the-World: A Commentary on Heideggers Being and Time,
Division 1. Cambridge MA: MIT Press.
Dourish, P. (2001): Where The Action Is: The Foundations of Embodied Interaction. MIT Press
October 2001.
Ducheneaut, N., Yee, N., Nickell, E., and Moore, R.J. (2006): "Alone Together? Exploring the
Social Dynamics of Massively Multiplayer Games." In conference proceedings on human factors
in computing systems (CHI 2006) April 2227, Montreal, Canada. pp. 407416.
Fernandez, R. Ginzburg, J. and Lappin, S. (2004): `Classifying Ellipsis in Dialogue: A Machine
Learning Approach. In: Proceedings of the 20th International Conference on Computational
Linguistics (CoLing), pp. 240246.
Fitzpatrick, G. (1998): The Locales Framework: Understanding and Designing for Cooperative
Work. PhD Thesis, Department of Computer Science and Electrical Engineering, The University
of Queensland.
Goffman, E. (1981): Forms of Talk. Philadelphia: University of Pennsylvania Press. Chap 3
Harrison, S. and Dourish, P. (1996): Re-place-ing space: the roles of place and space in
collaborative systems. Proceedings of the 1996 ACM conference on Computer supported
cooperative work. Boston, MA, USA, pp. 6776
Heidegger, M. (1927): Being and Time English Translation Macquarrie, J. and Robinson, E. , 1962,
Blackwell, Oxford.
Hollan, J. and Stornetta, S. (1992): Beyond being there. In P. Bauersfeld, J. Bennett, and G. Lynch
(eds): Proceedings of the CHI92: SIGCHI Conference on Human Factors in Computing
Systems. ACM Press, New York, NY, pp. 119125.
Kendall, L. (2002): Hanging Out in the Virtual Pub: Masculinities and Relationships Online.
University of California Press.
Levinson S. (1988): Putting Linguistics on a Proper Footing. Exploring the Interaction Order,
Erving Goffman pp. 170.
Light, A. (2007) Transports of Delight?: What the experience of receiving (mobile) phone calls can
tell us about designin special issue on Enchantment, Experience and Interaction Design.
Personal and Ubiquitous Computing (in press).
192 Patrick G.T. Healey et al.
Mulhall, S. (2005): Routledge Philosophy Guidebook To Heidegger and Being and Time. London,
Routledge.
Schober, M.F. and Clark H.H. (1989): Understanding by Addressees and Overhearers. Cognitive
Psychology, 21, pp. 211232.
Schutz, A. (1954) :Concept and Theory Formation in the Social Sciences. The Journal of
Philosophy, 51(9) pp. 257273.
Smith, M.A., Farnham, S.D and Drucker, S.M. (2000): The social life of small graphical chat
spacesProceedings of the SIGCHI conference on Human factors in computing systems. The
Hague, The Netherlands, pp. 462469.
193Communication Spaces
... The term communication space was suggested by Healey et al. (2008) as a broad means of thinking about human interaction in online environments. Using the analogy of a physical location such as a stadium, the authors pointed out that use for different purposes (sports events, concerts, religious gatherings) results in widely different social norms (e.g., affecting appropriate behaviours regarding forms of communication: cheering, singing, being silent). ...
... The ability and desire for keeping distance or being close and supporting one other are different online and across individuals and occasions. These aspects of interpersonal closeness and mutual involvement, referred to as being-with by the philosopher Martin Heidegger, are an important aspect of human interaction and sit at the core of considerations around communication spaces (Healey et al., 2008). ...
... It builds on Saplacan's (2020) study to support investigation of the different tools and their combinations in the contexts of higher education and contemporary technologies. The overlaps between the concepts of common information spaces and communications spaces allow us to link to the work of Healey et al. (2008), investigating the effects on being-with, on bringing students and staff together. This led to the following research questions: ...
Full-text available
Article
This research examined the impact of supplementing a learning management system, Moodle, with communication tools, Discord and Teams, to support communication in blended and distance undergraduate courses in computer science, information technology, mathematics and statistics at a New Zealand university with well-established use of Moodle. Nineteen students participated in semi-structured interviews. Findings show that adding Discord or Teams increases information and knowledge exchange and helps students to connect with peers and teachers in their courses. Teams was beneficial particularly in settings with formal group work. Discord, which enables both students and staff to set up additional communication channels that are not restricted to course enrolments, was instrumental in connecting students across year levels with peers and alumni. This enabled discussions on course selection, career options and disciplinary topics beyond the course curriculum. Importantly, these beyond-course communication spaces nurtured belonging to wider discipline and study communities. The research establishes the importance of the increased levels of communication by highlighting the effects on student learning and connections to others. Looking beyond the specific tools, the level of formality and the degree of student co-ownership are identified as key factors in supporting the within- and beyond-course communication spaces. Implications for practice or policy: Learning management systems provide valuable course support but do not meet all learning and teaching communication needs, partly due to their formal and university-controlled nature. Educators and students benefit from using strong chat tools with improved information and knowledge exchange. Both educators and students experience enhanced belonging when using collaboration and chat tools. Students value Discord’s informality and student-led characteristics. Educators who use Teams to support formal group work must carefully manage integration with Moodle.
... − The site should enable communication to flow easily from one subpage to another, with design focusing on the communication space rather than the communication place (Healey, White, Eshghi, Reeves, & Light, 2007) Changes to the site were then implemented as follows (see figure 5 on p. 33 for the resulting community page design): ...
Thesis
http://urn.fi/urn:isbn:9789526216430 This thesis studies online community as a discursive phenomenon and as an experience. The ethnographic approach employed in this study allows the open exploration of meanings and experiences associated with community by site members, designers and outside commentators in three online environments. Extensive participant observation is supplemented by interviews, surveys and analysis of the interaction surrounding the topic of community. Nexus analysis provides an understanding of social action as the intermingling of historical bodies, interaction orders and discourses embedded in the scene of action. The thesis argues that the concept of community functions as a boundary object, taking different meanings in each context it is employed in. Community can be used to describe strong community experiences or lighter varieties of togetherness online; it can be a pragmatic term simply referring to the user base of a site; or it can incorporate many understandings related to the shared identities and shared practices in the social scene being referred to. The work thus provides a theoretical contribution to ongoing academic discussions related to defining online community, as well as a great deal of empirical knowledge on how experiences of togetherness are created online. Such knowledge may be used to inform future technology development and administrative practices that are sensitive to the many elements affecting social interaction in online spaces.
... Building on previous research, we consider transactional distance a multidimensional construct that includes both social and cognitive aspects, where interpersonal closeness, shared understanding and perceived learning are integral elements. In accordance with Moore's original definition, we emphasise the aspect of '(mis)understanding', while at the same time including the aspect of 'closeness' as a dimension of transactional distance that is consistent with the concept of 'psychological or communication space' (Fiedler, 1952;Healey, White, Eshghi, Reeves, & Light, 2008). Perceived closeness concerns the social aspect of transactional distance, while shared understanding and perceived learning concern the cognitive aspect. ...
Article
This study was an initial attempt to operationalise Moore’s transactional distance theory by developing and validating an instrument measuring the related constructs: dialogue, structure, learner autonomy and transactional distance. Data were collected from 227 online students and analysed through an exploratory factor analysis. Results suggest that the instrument, in general, shows promise as a valid and reliable measure of the constructs related to transactional distance theory. Potential refinement of the instrument and future research directions are included at the end of the article.
Article
Globally, 4.4% of the world's population suffer from depressive disorder, and 3.6% from anxiety disorder. Previous work found considerable negotiation between providers and patients about the nature of mental health problems and frequent patient resistance to treatment. However, how doctor-patient shared understanding of the problem is reflected in treatment recommendations and whether this is consequential for patient acceptance of treatment is poorly understood. This study explored shared understanding of the problem and patient acceptance of treatment using conversation analysis. In 52 U.K. video recorded primary care consultations (collected July 2014–April 2015), 33 treatment recommendations for medication or talking therapy were identified. Shared understanding was explored focusing on: whether the patient presented the mental health problem as their primary initial concern and how they characterised the concern; whether the mental health concern was raised by the patient; and how the doctor aligned with the patient's earlier characterisation of the problem in the treatment recommendation itself. These phenomena were coded for each treatment recommendation and the impact on treatment acceptance was explored. Patients accepted the recommendation immediately in 38% cases and actively resisted in 62% cases. However, two communication behaviors were associated with patient acceptance: recommending treatment for the patient's initial focal concern and doctors' turn design in the recommendation itself, i.e., using the patient's earlier words from the initial problem presentation to describe and characterise the problem. Given the global burden of mental health problems and frequent patient resistance to treatment, understanding how professionals can engage more closely with the patient's perspective is important. When doctors use the patient's precise words from the initial problem presentation in the treatment recommendation, this displays an understanding of the patient's perspective and personalisation of treatment based on the underlying biomedical or social causes, which then impacts on patient acceptance of treatment.
Chapter
This chapter establishes a wider context for language research about gender and sexuality that uses online data and to discuss the specific methodological issues in this study. The features of multi-user domains (MUDs) are discussed as well as the specific features of Walford, the MUD examined in this book, including the creative control participants have over their text-based environment, the conversational settings and commands used for synchronous talk, and discursive resources participants strategically use as privacy features. Consideration is given to the ethical issues that underlie researching both online communities and language and sexuality studies, including their adoption here.
Article
Purpose The purpose of this paper conducts a co-design framework, thanks to different kinds of interviews with expert pilots in different space contexts for each stage of the technical development of the system. Design/methodology/approach To speak about the question of “spaces” and especially spaces in the design process, this paper is focussed on the transdisciplinary design research for commercial airline pilots. This design research combined a “Human-Centred Design” approach (Boy, 2015) with an occupational psychology, including a “Clinic of Activity” perspective (Kloëtzer et al. , 2016). It integrates experts within the research process to design a new technological tool for indicating weather conditions in flight (Boulnois, 2018). Findings This study explains how the crossed self-confrontation interview under special space conditions allows dialogue to take place between the designers and the pilots for imagining the future workplace by using an emergent workspace. Practical implications The physical, psychological and social “space” conditions were worked on throughout the design process – including during the work analysis and design testing phases – and were revisited. Originality/value In this way, the occupational psychologist organised the framework for dialogue about the present situation and future workspaces, to anticipate changes in work organisation, enhance inter-professional collaboration and, finally, increase health and safety benefits for pilots, and potentially other workers, at the same time.
Article
Background: Participatory action research is a credible, culturally appropriate methodology that can be used to effect collaborative change within vulnerable populations. Aim/Objective: This participatory action research study was undertaken in a Western Australian metropolitan setting to develop and evaluate the suitability, feasibility and effectiveness of an Aboriginal peer-led home visiting program. A secondary aim, addressed in this paper, was to explore and describe research methodology used for the study and provide recommendations for its implementation in other similar situations. Methods: Participatory action research using action learning sets was employed to develop the parent support program and data addressing the secondary, methodological aim were collected through focus groups using semi-structured and unstructured interview schedules. Findings were addressed throughout the action research process to enhance the research process. Results: The themes that emerged from the data and addressed the methodological aim were the need for safe communication processes; supportive engagement processes and supportive organisational processes. Conclusions: Aboriginal Peer Support Workers and community support agencies identified three important elements central to their capacity to engage and work within the participatory action research methodology. This research has provided innovative data, highlighting processes and recommendations for child health nurses to engage with the Peer Support Workers, parents and community agencies to explore culturally acceptable elements for an empowering methodology for peer-led home visiting support. There is potential for this nursing research to credibly inform policy development for Aboriginal child and family health service delivery, in addition to other vulnerable population groups. Child health nurses/researchers can use these new understandings to work in partnership with Aboriginal communities and families to develop empowering and culturally acceptable strategies for developing Aboriginal parent support for the early years. Impact Statement Child health nurses and Aboriginal communities can collaborate through participatory action research to develop peer-led support for the early years.
Full-text available
Article
"Communicative space and its modeling" Communication between people can take various forms depending on a lot of circumstances – participants’ individual characteristics, their social roles, subject of conversation, etc. The paper introduces a work in progress on modelling one aspect of natural human communication – communicative space. Communicative space is a mental space where a communication participant places himself/herself with respect to other ones and where (s)he is ‘moving’ during a communication event. Communicative space can be characterized by different features, e.g., (social) closeness of a communication participant with the partner, collaboration, politeness, etc. These features of communication can be conveyed by language use as well as by different nonverbal means (body movement, facial expressions, etc.). The values +1, 0, and –1 are used for the coordinates in communicative space. Examples of human-human dialogues – both everyday and institutional – demonstrate how participants pass different points in communicative space during a conversation. The further aim is to include such a model of communicative space in an experimental system for modelling conversational agents in order to make interaction with the system more human-like.
Article
We study human-human dialogues in a natural language where the communicative goal of the initiator of dialogue is to bring the partner to a decision to do a certain action. If the partner does not accept the goal then dispute will start. Arguments for and against of doing the action will be presented by the participants and finally, one of them wins and another loses the dispute. We present a formal model of dispute which includes a model of argument. We discuss involvement of the notion of communicative strategy in the model. A communicative strategy is considered as an algorithm used by a participant for achieving his or her communicative goal. A communicative strategy determines also how a participant is moving in 'communicative space' during interaction. Communicative space is characterized by a number of coordinates (e.g. social distance between participants, intensity of communication, etc.). A limited version of the model of dispute is implemented on the computer.
Article
The title of my paper refers intentionally to that of a Symposium held in December, 1952, at the annual meeting of the American Philosophical Association.2 Ernest Nagel and Carl G. Hempel contributed highly stimulating comments on the problem involved, formulated in the careful and lucid way so characteristic of these scholars. Their topic is a controversy which for more than half a century has split not only logicians and methodologists but also social scientists into two schools of thought. One of these holds that the methods of the natural sciences which have brought about such magnificent results are the only scientific ones and that they alone, therefore, have to be applied in their entirety to the study of human affairs. Failure to do so, it has been maintained, prevented the social sciences from developing systems of explanatory theory comparable in precision to those offered by the natural sciences and makes debatable the empirical work of theories developed in restricted domains such as economics.
Book
This 2007 book considers how agencies are currently figured at the human-machine interface, and how they might be imaginatively and materially reconfigured. Contrary to the apparent enlivening of objects promised by the sciences of the artificial, the author proposes that the rhetorics and practices of those sciences work to obscure the performative nature of both persons and things. The question then shifts from debates over the status of human-like machines, to that of how humans and machines are enacted as similar or different in practice, and with what theoretical, practical and political consequences. Drawing on scholarship across the social sciences, humanities and computing, the author argues for research aimed at tracing the differences within specific sociomaterial arrangements without resorting to essentialist divides. This requires expanding our unit of analysis, while recognizing the inevitable cuts or boundaries through which technological systems are constituted.
Book
This book explores the brave new world of social relations as they have evolved on the Internet. It examines how men and women negotiate their gender roles on an online forum the book calls BlueSky. The result is an analysis of the emerging social phenomenon of Internet-mediated communication and a study of the social and cultural effects of a medium that allows participants to assume identities of their own choosing. Despite the common assumption that the personas these men and women craft for themselves bear little resemblance to reality, the book discovers that the habitués of BlueSky stick ... More This book explores the brave new world of social relations as they have evolved on the Internet. It examines how men and women negotiate their gender roles on an online forum the book calls BlueSky. The result is an analysis of the emerging social phenomenon of Internet-mediated communication and a study of the social and cultural effects of a medium that allows participants to assume identities of their own choosing. Despite the common assumption that the personas these men and women craft for themselves bear little resemblance to reality, the book discovers that the habitués of BlueSky stick surprisingly close to the facts of their actual lives and personalities. This book explores the brave new world of social relations as they have evolved on the Internet. It examines how men and women negotiate their gender roles on an online forum the book calls BlueSky. The result is an analysis of the emerging social phenomenon of Internet-mediated communication and a study of the social and cultural effects of a medium that allows participants to assume identities of their own choosing. Despite the common assumption that the personas these men and women craft for themselves bear little resemblance to reality, the book discovers that the habitués of BlueSky stick ... More