Interaction

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 
136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 
153 154 155 156 

This section outlines and reviews the issues involved when a central actor

manages interactional asymmetries in order to be able to engage in fluid mobile

work. Firstly, we must recognise that distance matters (Olson & Olson, 2000),

and that a host of issues related to the struggle for maintaining fluid mobile work

emerge from the mobilisation of interaction and from the fluidisation of work

activities across temporal, spatial and contextual boundaries. Many of the

issues discussed below are also interesting to study and discuss in settings

where the interaction is “at arms length,” but here the main objective of such a

study could be the opposite — demonstrating the richness of means by which

collocated interactors manage fluid interaction (Goffman, 1982; Heath & Luff,

2000; Olson & Olson, 2000). Studies show that collocated office workers

spend between 25% and 70% of their time on face-to-face conversations.

Collocated synchronous interactions can be characterised in terms of (Olson

& Olson, 2000): rapid feedback, multiple channels, known identity of contributors,

shared local context, impromptu interactions on arrival and departure,

easy establishment of joint references to objects, free individual control of

attention and participation, implicit peripheral cues, and the spatiality of

reference. As outlined previously, empirical studies have characterised most

interpersonal interaction as dyadic (involving two people), brief, opportunistic,

synchronous, focused on shared objects, ongoing rather than one-shot, and

containing multiple threads (Whittaker et al., 1997). Conversations are typically

brief, synchronous and opportunistic interactions with multiple concurrent

threads, thus leading to issues of context regeneration and conversation thread

tracking (Whittaker et al., 1997).

The first issue to consider relates to the relevance of the interaction. Although

much of the contemporary discourse on the application of ICTs seems to focus

on the increased ability to interact anytime and anywhere, and that this apparent

ability is translated into a subsequent necessity of interacting anytime and

anywhere, studies of actual work activities demonstrate that by far all work can

indeed be conducted anytime and anywhere (Wiberg & Ljungberg, 2001).

Distance matters not only in terms of geographical distance but also in terms of

temporal and contextual asymmetries between interactors. The initiator may be

in one frame of mind, focused on for example getting a meeting schedule

finished, whereas the person he or she contacts on a mobile phone in order to

clear some details may be in the middle of an important meeting, or concentratFluid

ing on writing a memo. In the case of the telephone engineers studied by Wiberg

(2001), their work depends critically on travelling to where the fault has been

reported. They must therefore negotiate access with the house owner. Wiberg

(2001) found examples of lack of mutual awareness of what the status of the

cases were at a given time, such as in the example where the engineer had

travelled quite far only to find out that rebooting the telephone switch had solved

the problem remotely, rendering the long trip unnecessary. We here have a

potential deepening of the interaction asymmetries, when at the one hand

mobile work is conducted in certain spatial, temporal and contextual circumstances

and on the other hand pervasive mobile communication technologies

offers an apparent stability of dissolving temporal, spatial and contextual

barriers — the mobile worker can be reached no matter where and when they

are and irrespective of what they are engaged in (Agre, 2001).

Generally, interaction modalities can be characterised in terms of their degree

of obtrusiveness and ephemerality (Schmidt & Simone, 1996; Ljungberg &

Sorensen, 2000). Interaction can be perceived as more or less obtrusive in

terms of how the interaction forces upon the participants the need for them to

devote their attention towards the interaction. Ephemeral interaction unfolds in

time and space without leaving behind external traces, whereas persistent

interaction is characterised precisely by traces being sedimented from the

interaction. Both modes of interaction offer continua and not distinct categories.

Interaction can be perceived as more or less obtrusive to the fluid

accomplishment of mobile work dependent on the actor’s subjective perception.

Mackay (1988) showed that different actors experienced different levels

of stress coping with email overload. Similarly, the degree of persistency can

depend on the actual technology used and the specific way in which it is

appropriated and combined with other technologies. Whilst a face-to-face

conversation clearly can be characterised as ephemeral since it only verbal, and

an email clearly is persistent in that it leaves a trace behind that can be inspected

at a much later stage, then instant messaging such as ICQ and MSN message

service clearly can be viewed as both asynchronous persistent interaction as

well as synchronous ephemeral interaction. Although the streams of interaction

may be stored at the server for later retrieval as digital traces of human activity

(Sørensen et al., 2000), the situated use can equally be viewed as real-time

conversation where the trace only serves a highly temporary primary purpose.

When investigating the struggle to obtain fluid mobile work, interesting insights

can be gained from investigating configurations of the two interaction modalities.

Using a traditional stationary phone with no caller ID displays results in a

fairly obtrusive and ephemeral interaction. Modern mobile phones, on the other

hand, offers increasingly advanced ways of setting the interaction requests as

unobtrusive, so only the acceptance of an incoming call will lead to obtrusive

ephemeral interaction. It is not unrealistic to assume that mobile ICTs will

support more sophisticated awareness, filtering and notification mechanisms

(Ljungberg & Sorensen, 2000), for example by allowing the semi-automatic

filtering of incoming requests for interaction and subsequent conversion of a

request for a conversation to a reply by SMS explaining the interactional

preferences of the receiver of the request. Generally, the current mobile ICTs

primarily support interaction modalities that are not ephemeral and unobtrusive

— the aspects of human interaction, which have been demonstrated, handled

in very sophisticated ways when interactors are collocated and working at

“arms length”. We will below discuss the related issue of mutual awareness

further.

Taking a closer look at the relationships between ephemeral and persistent

modes of interaction, work studies have illustrated how actors actively will

make ephemeral interaction persistent; for example by audio or video taping

conversations, by photographing whiteboards, or simply by taking notes during

meetings or during telephone conversations (Ljungberg & Sorensen, 2000).

Since it is associated with considerable transaction costs — for example to

transform audio notes to written text — some research has looked into enabling

easy retrieval of passages in audio material such as voice mails (Whittaker et

al., 2000). In a sense, these activities partly can serve the purpose of

crystallising work activities, for example by making discourses and discussions

publicly known for others to inspect and comment. Written documents, audio

or video notes, email trails, voice mails, SMS messages etc, therefore provide

a common and shared, although fragmented, awareness of dyadic interrelationships

for later inspection and reference. Such awareness is at least to some

extent quite crude since they can not be assumed to be synchronous, nor can

they be assumed to reveal significant information of the other part’s willingness

or readiness to be interrupted.

Increasingly technologies offer ways of managing fluid mobile interaction by

offering the receiver of the request for interaction to postpone or stack the

interaction (Wiberg, 2002; Wiberg & Whittaker, 2004). These technologies

are either asynchronous, such as the basic email that can be accessed from a

mobile phone or PDA, or can be gateways between synchronous and asynchronous

technologies, for example allowing a telephone conversation to be

postponed by having the number stored in a “missed calls” register or by

allowing the caller to leave a message to be picked up at a later point. These

mechanisms of postponing synchronous interaction can be combined with

various filtering and awareness mechanisms (Ljungberg, 1999; Ljungberg &

Sorensen, 2000), such as a discretely vibrating phone, which from the user’s

trouser pockets alerts of an incoming call, or the more advanced awareness

filtering of assigning different groups of callers to different ring tones. The ability

to translate a synchronous attempt to interact into an asynchronous trace can

obviously lead to severe consequences for everyone involved. Establishing

fluid mobile work in the short-term by disconnecting and stacking all interaction

requests clearly may result in others’ struggling to maintain fluid mobile work.

Around 60% of workplace phone calls have been reported to fail to reach the

recipients (Nardi & Whittaker, 2000). In the case of networked mobile

workers, this can of course be assumed to change, but against an increasing

intensification of interaction, it may still be a significant problem. Wiberg (2001)

reports of telephone repair engineers with three mobile phones for each

engineer. Furthermore, not all problems may go away by being transformed

from synchronous to asynchronous requests for interaction. Postponing implies

spending time at a later stage retrieving the information and subsequently

attempting to contact people who have attempted to contact. Easy access to

the available interaction threads is therefore important (Nardi & Whittaker,

2000). Recent empirical studies of availability management have also revealed

that having technological support built into the mobile phone for being able to

briefly and instantly micro-negotiate with the sender of an interaction request

upon an alternative and more appropriate time to initiate the interaction is more

efficient for the receiver than taking calls immediately which might be interruptive

to the task at hand (Wiberg, 2002; Wiberg & Whittaker, 2004).

Given the increase in available means for communicating with others and the

general increased technical sophistication of mobile ICTs, maintaining fluid

mobile work is also increasingly a matter of managing multiple ongoing

conversations over multiple technologies. Here, mutual awareness comes to the

fore as a crucial issue. In cases of single synchronous technology scenarios,

such as the traditional office with one stationary telephone, others will immediately

be aware of essential interactional aspects of the owner’s behaviour —

if the phone is busy, the person is on the phone. However, with an increased

fragmentation and mobilisation of interaction across multiple ICTs, lack of

mutual awareness can lead to increase in the interaction asymmetry. Establishment

of mutual awareness of location has been promoted as an important

element of mobile interaction (Mäenpää, 2001), but also conflicting accounts

of the awareness of activities as the primary element has been promoted

(Weilenmann, 2001). However, in both cases, a generalised notion of location

awareness is being negotiated since Weilenmann argues that the awareness of

activity infers awareness of location. It can, therefore, be argued that the use

of mobile phones, for example, socially constructs a location-based service,

both in terms of allowing constant update mutual awareness of locations, as well

as in bringing the interaction to the location. Much research has discussed

applying specific awareness technologies, AwareWare (Nilsson et al., 2000).

The most common of these technologies is a stationary interaction technology

in its own right, namely Instant Messaging, where platforms such as ICQ allow

participants explicitly to declare their interactional status. Other systems, such

as the one reported by Nardi et al. (2000) supported implicit awareness by

monitoring user keystroke rates and therefore enabling others to gain an

impression of whether or not the person to be contacted is situated by their desk

or, alternatively, perhaps is too busy to be contacted at all. Tang et al. (2001)

demonstrate a multi-platform mobile awareness system with implicit location

logging. Dix et al. (2000) suggest a generic systems architecture for mobile

awareness technologies that integrates the technologies technical “awareness”

of internal state with the inclusion of the external context to support mutual

awareness between users.

The extent to which ICTs model the properties and behaviour of the interactors

can greatly affect both the degree of sophistication to which the ICT can

support the establishment of mutual awareness, as well as the extent to which

the people involved will find the technology a breach of their privacy and more

serve as a surveillance technology. Here, we would assume as a first hypothesis

the same interrelationship as argued by Schultze & Vandenbosch (1998)

concerning the perceptions of information overload when using Lotus Notes.

Here it was demonstrated that, although there were initial reports of information

overload as a result of implementing Lotus Notes, a subsequent study showed

that the actors had adapted and therefore no longer experienced information

overload. Similarly, Nardi et al. (2000) discuss how the introduction of the

Instant Messaging system led to initial discussion of how people felt observed

since everyone else could inspect their key-typing rate constantly. However,

this concern was subsequently forgotten and a positive stance had prevailed —

people saw the immediate benefits of gaining awareness of the typing speed of

the person they would wish to interact with in order to know if they were busy

or not. This mechanism is an example of an awareness mechanism, where

information pertaining to the interaction is provided to others in order to

facilitate their decision as to whether or not they still wish to interrupt.

In terms of managing fluid mobile interaction in situations where a number of

participants are engaging in ongoing interaction, traditional CSCW issues

pertaining to session management and floor control are brought to the fore.

Clearly most organisational conversations have an ongoing character as

opposed to the traditional view of sessions discretely organised in time

(Whittaker et al., 1997). Participants will need to engage in prioritisation of

competing conversation threads, and the generally poor support for ephemeral

and unobtrusive modes of interaction, which are essential for obtaining fluid

session management (Whittaker et al., 1997; Olson & Olson, 2000; Wiberg,

2001), implies significant challenges.

We have based the analysis of fluid mobile interaction on the assumption of

proactive actors exercising their judgement in the situation as to who they wish

to interact with, for how long and if they choose to turn on the answering

machine and email instead of engaging in a telephone conversation. Here, we

therefore critically emphasise the role of the individual’s subjective preferences

as opposed to organisational or task rationality in terms of purpose, duty, or

need. We acknowledge that there of course exist systems of power, domination,

division of labour and so on.

Discussion

This chapter has, through theoretical inquiry, aimed at outlining the main

challenges for understanding fluid mobile interaction from the perspective of the

individual actor constantly faced with detailed decisions as whether or not to

engage in interaction with others. Our discussion in this chapter has been based

on the explicit choice of attempting to integrate theories from social topology,

sensitising how we generally can characterise changes to work practices, with

theories emerging from detailed operational studies of how new technologies

can be integrated within organisational practices. Traditionally the former

strand of research retains great distance to discussions of actual technologies

and work practices, whereas the latter almost exclusively focus on how

concrete technological innovations integrate with the performance of specific

tasks. We find that both perspectives ought to be called upon when attempting

to address some of the most important issues organisations are faced with at the

moment. One of the reasons for this relates to a particular discourse we have

not related to in this chapter, but which naturally would be a key subject for

further consideration, namely the issue of the relationship between modes of

organising and the mobilisation of interaction.

Rapidly adopted technologies such as email, mobile phones and instant

messaging can be characterised by their interpretative flexibility in terms of the

individual user locally defining the patterns of use. They basically provide a

backbone mobilising conversations, and relative little else. Thereby, they

support the individual actor in engaging in encounters with others whilst

retaining emergent processing of information, thus enabling the individual user

effectively to cope with tasks of relatively low complexity but relatively high

degree of complexity (Mathiassen & Sørensen, 2002). The local application of

locally conditioned microprocedures (Lanzara & Patriotta, 2001) is used as a

means of coping with conflicting interactional requirements. The individual

actors specific configuration and application of a single ICT, as well as the

specific patterns of use and portfolio choice of multiple technologies, can be

viewed as locally conditioned improvisation to resolve the local messiness

(Ciborra, 2002).

Some of the unintended consequences of the flexibility and immediacy of these

technologies are then their flexibility to be applied, for example, for tasks of high

complexity and low degree of uncertainty. Here, simple mobilised conversation

representing emerging information processing could, for example, be replaced

by a systematisation of the information processing, thus reducing the risk of

participants experiencing interaction overload (Mathiassen & Sørensen, 2002).

Imagine the slightly silly proposition of having to replace Amazon.com’s

automatic order handling system with a call centre. However, in order to

comprehensively analyse the mobilisation of interaction and the struggle to

obtain fluid mobile work, we cannot exclusively look at the overall social or

managerial issues of rational ways of organising work. The individual actor’s

intentions, desires, moods, local dispositions, etc., will greatly affect the microprocedures

they employ for managing their own availability and interactivity

(Ciborra, 2002). Here, one of the key issues pertaining to the individual’s

experience of fluid work will be the daily unfolding of both interaction and

outeraction, where the latter characterises communicative processes employed

solely to establish and discuss connections with each other, as opposed to

actually communicating (Nardi & Whittaker, 2000). When, for example,

mobile voicemails only containing the message “ring me at...” are served in

a ping-pong fashion between people, all they do is outeraction. If there are

constantly discrepancies between individual actors’ desire to be interactive or

interpassive, the increased mobilisation of interaction can lead to dramatic

increases in this “coordination of coordination work” (Ljungberg & Sorensen,

2000). The recursive relationship between interaction and outeraction does,

however, not stop after the first recursion, and a question for further empirical

and theoretical investigation is the changes to the ways in which individuals and

organisations manage interaction in general. When all members of an organisation

have mobile phones, then there needs to be both general as well as highly

specific discussions of how these can be used to facilitate fluid mobile work.

However, the organisation may also need to discuss how to coordinate and

discuss the use of new interaction technologies.

Neither can we exclusively study these phenomena in a bottom-up fashion. The

mobilised interaction always takes place within a social and perhaps even an

organisational context. This raises socially conditioned issues of power,

influence, domination, culture, privacy, surveillance, etc. It may, for example,

be a very good idea for mobile workers to know what close colleagues are

doing and where they are located, such as discussed from a technical and

operational perspective by Tang et al. (2001). However, in some contexts it will

be considered surveillance of work if the information is shared with others in the

organisation (Ciborra, 1996). Similarly, if the project team members are

distributed and highly mobilised, it is probably not acceptable from an

organisational point of view if a key member of the project decide that

maintaining fluid mobile work involves sustained periods of disconnectedness.

There are, in our view, significant methodological consequences of the

mobilisation of interaction. When studying CSCW systems, it is a significant

challenge to sufficiently cover and understand the roles, opinions and detailed

actions of distributed interdependent actors (Grudin, 1994). Even more

distribution and mobilisation of interaction will potentially imply even more

methodological challenges for fieldwork design. The mobilisation of interaction

also brings novel approaches to the fore, such as conducting virtual ethnographies,

studying mediated interaction patterns (Hine, 2000).

Summarising, it is clear that a host of research results from varied fields can

inform the discussion of fluid mobile work, and that this research at the same

time must be appropriated to situations where the interaction as well as the

actors are highly mobilised. The state-of-the-art technologies we have seen so

far show us interesting glimpses of the future, but it is evident that the real issues

in contemporary working life are radically changing and that the current

technologies being used do not sufficiently address the main issues.