8th November 1993



CHAPTER 8: Collaborative interactions *around* and *through* computers



What is typically suggested by the phrase "collaborative learning"?

Probably the image of individuals gathered together at some materials

that they are trying to understand. Naturally, from time to time such

individuals may drift into solitary reflection. However, to qualify

as a genuinely collaborative interaction, we normally expect there to

be an underlying common focus for attention. In the last Chapter, I

illustrated an approach to analysing such occasions. The analysis

highlighted varieties of social coordination that might be achieved at

these focal points. It explored the natural concern of collaborators

to construct common understandings and shared systems of reference.

The analysis also evaluated problem solving environments in terms of

how far they promote or undermine such efforts. Computers may

sometimes be effective environments for learning in this sense;

sometimes they are not. However, the localised and interactive

properties of this technology do suggest that it has a special

potential for resourcing the social construction of shared knowledge.

Yet, our spontaneous image of collaborative learning as involving a

gathering *at* problem solving materials may be too narrow. As

Landow has put it: 'I suspect that most peoples' conception of

collaborative work takes the form of two or more scientists,

songwriters or the like continually conferring as they pursue a

project in the same place at the same time' (Landow, 1990, p. 407).

This conception is too limiting. In the present chapter, I wish to

discuss two broader configurations for the organisation of

collaborative interactions. They both entail a reconsideration of

what defines a shared problem solving environment. In particular, I

shall discuss circumstances in which collaborations may be dislocated

in time - the participants do not need to be co-present. And I shall

discuss circumstances in which the participants have less

comprehensively overlapping concerns - where the problems that they

are each addressing are more loosely coupled. These represent,

respectively, central features of interacting "through" computers and

interacting "around" them.




Of all the educational arrangements whereby new technology might

support collaborative activity, this is the one that has attracted

least research or commentary. Regretfully, I shall not report new

empirical material here that helps fill this gap. Instead, I shall

identify the questions that need to be addressed and argue that they

are interesting and worth research investment.

Exactly what belongs under this heading can be approached by

reflecting very generally on the material nature of learning

environments. For example, Walkerdine (1984) has drawn attention

to the typical environment of a primary school classroom. Once one

has stepped back from this familiar setting, it becomes possible to

see a variety of constraints and supports that are literally built

into it. Chairs and desks are oriented into characteristic patterns;

walls are decorated with particular material; areas of the room are

furnished in ways that afford defined possibilities for acting. In

short, the very fabric of the place is designed to manage the business

of learning according to distinctive ideas about good practice. It is

typical of socio-cultural theorising to dwell upon the manner in which

environments serve to mediate cognitive activity. As reviewed in

Chapter 2, this theoretical tradition understands cognition in terms

of a human subject located in relation to mediational means. In

respect of schooling, this approach will consider how local

mediational means can resource pupils with particular interpretative

practices. Accordingly, one way socio-cultural theory should direct

research on new technology is towards investigating options for

integrating computers into this "fabric" of the educational


The present Chapter addresses such a concern. This section of the

Chapter does so in relation to those configurations of computers

*around* which social interaction may be organised. In clarifying the

idea of such configurations, the above example of the classroom fabric

is helpful. It raises a concern for the material setting of

collaborative encounters, and it encourages us to think about them

very broadly. The example leads us beyond considering only the

familiar case of collaborating by intimately working together *at*

some location or artefact. For the design of classrooms vividly

illustrates the principle that material environments will constrain

and facilitate a whole range of social interactions that can occur

within them. So, the structure in some particular environment may

influence all sorts of collaborative engagements that we may be party

to. A sensitivity to such ecological considerations could guide

decisions about the optimal deployment of new technology. For it

should suggest a constant need to consider how any particular

technology (say, computers) is best incorporated into some established

material environment (say, a classroom), such that it becomes an

effective component within the very fabric of working practice.

As I commented at the outset, the issue of how to locate computers

within the material environment of schools has not been closely

studied. In practice, the choices may seem rather restricted. In

primary schools there is still too little equipment. So, to ensure

equable access and to encourage its use in all aspects of a

curriculum, the preference has been to distribute computers evenly

across classes. Where some mobility of equipment is possible, this

policy may allow creating short term pockets of extra access: such

that a class might enjoy a more intensive period of computer activity.

There has been very little attraction to the idea of concentrating

computers in circumscribed work areas.

Secondary schools, on the other hand, have been more likely to

configure computers into clusters that occupy distinct areas -

"computer rooms" perhaps. Sometimes these are networked and in some

schools that network may run through more of the premises (I shall

discuss these arrangements further in later Sections of this Chapter).

It is the particular case of grouping together computers to support

parallel patterns of working that interests me here. This

configuration creates the possibility of what I term "collaborative

interactions *around* computers".

To a casual observer, these areas can lead classrooms to look more

like working environments from the world outside of school. In fact,

it is more in relation to social practices in the workplace that these

arrangements for new technology have attracted some research

attention. Such interest has been partly inspired by new perspectives

within the tradition of "human-computer interaction" (HCI) research.

These involve challenging the prevailing concern of HCI researchers to

study the solitary user - as if the character of human-computer

interaction could be understood independently of the broader cultural

setting within which interactions get situated. These new

perspectives are well captured in a seminal collection of papers

edited by Norman and Draper (1986). This volume includes an essay by

Bannon (1986) in which the issue of a computer system's communicative

effectiveness is seen to involve consideration of the niche it

occupies in a context of existing working practices. So, for example,

the familiar problem of users failing to read computer application

manuals is understood in terms of tensions with established and

preferred modes of acquiring knowledge of this general kind. These, it

is argued, are often socially organised: knowledge gets sought and

exchanged with other people - on the fly. In many working

environments there is a rich but informal pattern of casual

communication that supports this. Studies of how new technology is

implemented at work tend to stress a need to respect such structures

(Huber, 1990; Olson and Lucas, 1982).

Many of us will recognise the force of this description from our

own experience of working environments. It is less recognisable as a

description of working patterns within classrooms. Sometimes, this

may be because classrooms impose some prohibition on such

communication - identifying it as "cheating" perhaps. However,

sometimes a lack of such communication among students may reflect

constraints inherent in the arrangements, materials and goals of

working. There may be grounds for thinking that some configuration of

computer systems can release a richer exchange in classroom settings.

A project reported by Kafai and Harel (1991) raises this possibility.

The work of these researchers does concern the support of

collaboration among pupils. However, they distinguish the scope

of their own interest in processes of collaboration from that entailed

in a more 'conventional' use of the term. The conventional

understanding, they argue, will involve two or more individuals

working towards a *single* product. In the situations documented by

Kafai and Harel, there is a layer of joint activity that is

superimposed upon individual work or conventional collaborations.

What this amounts to is the existence of an umbrella goal that a whole

class will be sharing; the goal is realised by individuals or small

groups producing their own distinctive products in relation to that

shared goal. In fact, the particular example Kafai and Harel describe

is an "Instructional Software Design Project" (ISDP): thus, the

umbrella goal becomes 'to use LogoWriter [an authoring tool] to design

a piece of software that teaches about fractions - but each of them

also expresses his or her own ideas and produces his or her own

project' (op. cit., p.87). The style of working then encouraged is

one that permits what they term 'Optional Collaboration' (choosing to

work alone or with others) and 'Flexible Collaboration' (deciding with

whom to work, when and for what purpose).

It is argued that the way in which computers are integrated into the

classroom environment serves to make these working practices realistic

and successful. In the ISDP project, the computers were organised

into circles with a great deal of freedom of movement and obvious

opportunities to take in what peers were doing. This arrangement

seemed to support what the researchers termed 'collaboration through

the air'. This corresponds to something like the patterns of exchange

typical of the corridors and coffee rooms of many traditional

workplaces. Kafai and Harel describe a number of case studies that

identify children gaining from the loosely-knit communication that the

setting readily affords.

Valuable though these observations are, I believe that more needs to

be done to characterise the social dynamic that is involved.

The nature of this alternative collaborative structure is defined

above in terms of the existence of an umbrella goal. But this is

hardly a distinctive feature: it may also characterise other routine

circumstances arranged in conventional classrooms. Children are often

working on such shared, overarching goals: they are, as a class,

writing project descriptions, making maps, and so on. Moreover, the

open-plan character of many British schools does encourage some degree

of lateral communication within a classroom. So, what must be

clarified is how far the possibilities for creative, through-the-air

collaboration are extended because of the particular properties of

working around computers. I think it is likely that computers are a

powerful context in this sense. Moreover, the source of their

strength may lie in a referential anchoring capability: a capability

of the very kind discussed in the last chapter in relation to more

conventional collaborations.

This analysis assumes that the constraints on collaborating through

the air arise from the "air" normally being too thinly resourced:

there are not usually enough available anchor points at which action

and attention can be coordinated. Loose communication can often

flourish in workplaces because the collaborators there are frequently

drawing upon a richly articulated body of mutual knowledge. This is

an inter-mental achievement; it will have developed over a long

history of interconnected communications. School work is less

likely to be easily grounded in this way. The immediate concerns of

school work are rarely situated within such intimate and evolving

understandings. Instead, schools create problems that are much more

localised and much less easily related to some prevailing and shared

set of institutional purposes. However, where school problems are

explored in the medium of new technology, then collaborators may be

better provided with referential anchors for the development of their

shared interests. For they will be pursuing their various goals with

overlapping sets of tools; these tools will be visible and manipulable

as a common point of reference; their mechanics will provide a

sensible vocabulary that a pupil confidently can use with any

collaborator who has had experience with the same resources.

In summary, I am not suggesting that configuring computers into

classroom clusters is necessarily the ideal way to arrange them.

There may be other institutional factors that dictate other

possibilities. In particular, there is a valid concern that

too rigid a separation of computers into their own rooms serves to

identify them with a curiously dislocated and special class of

activity (Chandler, 1992). Yet, however such organisational decisions

are made they should include some sensitivity to the issue of how

different working arrangements can afford different possibilities for

collaboration. These possibilities are not confined to the case of

small groups working *at* computers, and as already discussed here.

There is a level of community-based collaboration that can arise where

class activities are more loosely-coupled. Nevertheless, in both

kinds of situations the technology may be serving to support

collaboration by providing strong points of shared reference. A

similar claim can be developed for the networking of computers - a

strategy that is sometimes adopted as an alternative to the clustering

discussed here. I shall turn to this strategy next.






The key issue introduced in the last chapter has arisen again. That

issue concerns how we may resource the constructing of a shared object

of understanding. In many circumstances recognition of this "object"

will arise from a common concern to make progress with some

self-contained task (such as writing a story, or scoring high on a

number puzzle). The activity takes place, together, at the

site of the problem. However, in the configurations to be discussed

in this Section, the form of a shared understanding may sometimes be

less tightly related to some such circumscribed problem. What it is

that comes to be held in common - that becomes a source of shared

reference - is more a set of broader intellectual practices. So, I

shall be partly interested here in how new technology can mediate

forms of activity that create *communities* of shared understanding.

Understandings that are held in common need not be exclusively

relevant to the short term goals of working together on localised

problems. There are circumstances where mutual knowledge provides a

general underpinning relevant to a whole range of collaborative

encounters: this arises in situations where people are held together

into communities that share a common set of concerns - such as might

sometimes arise within institutionalised education.

In this Section, I intend to sketch the nature of network

configurations in general and, then, consider the particular

implications of networking for educational practice. I shall pursue

the educational theme by reporting an example of network-based

innovation with which I have a close involvement: an initiative

directed at an undergraduate population. This will enable me to make

a number of points about collaborative structures mediated *through*

computers. Some of these points I shall take up later to consider

their significance for practices in the earlier years of education

will be discussed.



The nature and application of distributed computing

For a long time, the popular representation of computer use was that

of a user occupied at a self-contained machine. This may still

capture well enough the domestic experience of this technology.

However, in institutional settings, it is increasingly likely that the

user will be occupied at a machine that is less isolated: it is likely

to have connections to other computers at other locations. Equipment

that is linked together in this fashion is said to be networked.

Local area networks (LANs) allow communication within the premises of

some workplace. Wide area networking (WAN) involves links with

geographically very remote computing systems. The original attraction

of such connectivity was to permit a large number of individual

computers (network stations) to gain shared access to a central

resource of data (a file server). Quite simply, this centralisation

offered a saving on the amount of computer hardware required for data

storage. It also allowed central administration of those

data facilities that members of an organisation might wish to share.

However, the design of networks has become more sophisticated in

recent years. The model of a central file server passively delivering

data to connected stations as they request it is now a fairly

primitive conception of networking. The modern concept of

"distributed computing" identifies a more dynamic environment.

"Servers" in such environments can themselves be very powerful. They

can take responsibility for some of a user's computational needs - as

well as delivering files that might be accessible to the computational

resources of that user's own network station. This has led one

company active in this area to adopt the commercial slogan: "The

network *is* the computer". Certainly, at any moment, exactly where

the computing is taking place can be well hidden from a typical user

of a distributed computing network.

These configurations of computers might have evolved originally to

achieve straightforward economies of resource. As such, they might

have been seen as merely extending and optimising the computing power

available to some community of solitary users. However, it has become

clear that these arrangements have significant implications for

patterns of coordination among those users themselves. A widely-cited

example is that of the ARPAnet - a pioneering US defence network,

originally conceived to distribute computing power among the research

community. It quickly became apparent that this network unexpectedly

was supporting a great deal of interpersonal coordination.

Researchers were finding their needs for collaborative contacts were

being increasingly mediated by a person-to-person connectivity

made possible by file transfer over the net. Newell and Sproull

(1982) provide an early summary of how research communities

appropriated this technology for such purposes.

Network configurations permit two key procedures: the transport of

files between users and the controlled access to data files held

centrally. Network-based collaboration is then made possible by tools

and structures that elaborate these core capabilities. Files become

objects that a community of users may view, share, and transform

through the use of these tools - much as they might manipulate

resources in the material (non-computed) world of collaborating. More

recently, networks have also permitted the direct transmission of voice

and video: thereby, allowing more vivid communication between remote

users, as well as their coordinated exchange and manipulation of data


For our present interests, it is not important (and probably not

possible) to establish a rigorous taxonomy of the supporting structures

that have been documented for communication within networked

environments. However, certain broad distinctions and themes are

useful to identify, particularly as they relate to developments within

educational settings. One class of widely-used resource stresses

structures for interpersonal exchange. These more generic tools

permit text-based communication on a one-to-one or one-to-many basis.

The best-known example is electronic mail (email) which allows a user

to compose text messages at a network station. A mailer program can

then be instructed to transfer them immediately to the file space of

some remote user(s) - who will activate their own mailer to read,

reply or forward the material. Increasingly, it is now possible to

incorporate sound and visual images into these messages. A variant on

this pattern is the electronic bulletin board, whereby material may be

posted in a central file space that can be widely read by users within

some networked community. This structure has, in turn, been

elaborated in the form of so-called "conferencing programs". These

allow orchestrated discussions by imposing topic structures on

text entry (and, sometimes, a moderating mechanism). Users may write

their own entries into a free space within this structure: the effect

is to create threads of contributions. Participation in these

computer conferences may come to resemble the experience of a seminar

or workshop interaction.

Other network resources are tailored to supporting interactions with

more circumscribed purposes. That is, they may impose greater

structure on a group interaction and/or offer more specialist tools

for manipulating material of joint interest. Such software is

sometimes termed "groupware" and useful characterisations of the genre

can be found elsewhere: particularly, within reviews by Ellis, Gibbs

and Rein (1991) and by Johansen (1988). Naturally, the emergence of

this software has encouraged associated research enterprises concerned

with its impact in organisational settings. Thus, there is now an

active community of researchers studying the general phenomenon of

computer-mediated communication (Lea, 1992) and the more particular

circumstances of "Computer-supported Cooperative Work" (Bowers and

Benford, 1991; Grieff ,1988). There is also an active interest in how

the culture of organisations can be effected by the penetration of

computer-mediated communications (Rice, 1992).

Many of these networking developments have indeed penetrated a variety

of workplaces (Collins, 1986; Malone and Rockart, 1991). Commercial

contexts, rather than educational ones, now provide the real stimulus

for creating new resources. Educationalists can take advantage of

these developments, although not all the structures developed for

commercial purposes are relevant to the needs of education. For

example, in the commercial sector there is much interest in Group

Decision Support Systems (Vogel and Nunamaker, 1990), some of which

may be fashioned for networked environments. But, framing and

converging on "decisions" is less central to the concerns of

educational practice, and tools that help students do this may of

limited application outside specialised problem solving exercises.

So, the network-based structures that have been of most interest in

teaching settings are these. (1) Resources that support interpersonal

exchange and debate, particularly email and conferencing.

Institutions that service non-residential students who might be widely

dispersed have made the most effective use of these tools to support

their "distance education" (Harasim, 1990; Mason and Kaye, 1989). (2)

Information servers. These are programs offering individual users

easy access to database collections of files relevant to some

curriculum or organisational structure. These files may be located

within some menu-style screen environment and then examined copied or

printed by students. (3) Co-authoring tools. These allow two or more

users to jointly edit or create documents in an asynchronous manner

(Landow, 1990).

Most educational initiatives exploiting these structures have been

concentrated in the University sector. I shall briefly review their

progress later in this Section. First, we might note certain findings

that have emerged from more workplace-oriented research: these may

provide some pointers regarding what can be expected in educational

contexts. A recurring proposition is that network-supported

communications is potentially subversive: it can challenge existing

procedures that maintain a social order within some organisation.

Perhaps for this reason, the implementation of groupware-based

practices are not always successful. Thus, in reviewing this mixed

progress, Grudin (1990) comments '...a medium which allows widely

separated people to aggregate their needs is, in fact, quite

frightening' (p. 181). In general, organisations may be slow to

respond to these structural possibilities. Yates (1989) locates

innovations of the kind being discussed here in a broader historical

context. He notes that few management technologies were ever adopted

when they were invented, but only when shifts in management theory

made the possible applications more apparent. He comments: 'Real

gains await innovative thinking about the underlying managerial

issues' (Yates, 1989, p. 275). This may be a common fate for all new

media (Winston, 1986). SO, by the same token, we may assume that

these new technologies of communication will not easily penetrate and

transform established practices within the organisations of education:

this is already implicit in the historical review of teaching and

technology authored by Cuban (1986).

A further recurring claim about communication in networked commercial

environments is that the exchanges supported become deregulated or

uninhibited (Hesse, Werner and Altman, 1988; Sproull and Keisler,

1986). It is widely believed that this arises from the lack of social

cues that can be conveyed in text-dominated and asynchronous

communication. However, some commentators have questioned whether

this medium is really so impoverished in terms of its social texture.

Lea, O'Shea, Fung and Spears (1992) argue that uninhibited exchanges

are quite rare in computer-mediated communication: they may simply be

more memorable and visible in this text-based, archival context.

Moreover, they accept that interpersonal social cues are necessarily

minimal in this medium; but suggest that users make active regulatory

use of social identities as given by other kinds of cues - text

and format cues that specify affiliations to various social

categories. It is uncertain, therefore, how this medium might

influence the style of communication in educational settings.

However, it is clear that educational settings have traditionally

respected hierarchical structures of authority; whether or not these

become somewhat flattened by computer-mediated communication will be

one outcome of interest.

The most optimistic message to be derived from studies of workplace

communication through computers is that it loosens up patterns of

exchange within an organisation (Malone and Rockart, 1991): this may

support more creative coordinations. However, simply putting

network-based structures in place is not enough to ensure that real

transformation of communication practices ensues. This has often been

observed for the case of email (Carasik and Grantham, 1988; Eveland

and Bikson, 1986). We may heed Bannon's warning (1986) that the

success of innovation within some setting will depend upon innovators

proceeding with a sensitivity to the informal system of rules and

procedures that already govern working practices. I will consider

this more closely in the educational context of Universities - the

teaching institutions where collaboration through computers has been

most actively considered.



Computer-mediated communication in undergraduate education

There is much to be worried about regarding the manner in which

universities are being encouraged to deploy new technology (eg.,

Hague, 1991); however, it is unfortunate that critics (eg., Robbins

and Webster, 1985) are led to wholesale rejections of new technology

in this sector before its potential for supporting more collaborative

opportunities has been fully explored. This is a direction for

computer-based innovation that has yet to be properly explored and

judged. When the application of computers to university teaching is

discussed, it is curious how little attention is given to distributed

computing; and curious how slow educational practitioners have been to

recognise the relevance of networking to the support of collaborative


For example, in a recent collection of papers summarising IT-based

teaching innovation in various UK universities (Gardner and McBride,

1990), there is no mention of networks as a teaching resource.

Similarly, Darby (1991) does not include communication infrastructures

in his review of future needs - when reflecting on the outcomes and

implications of a UK initiative to promote computers in university

teaching. Hale (1990) describes an example of a computer-rich

teaching department: yet this model does not incorporate communication

considerations. Part of the reticence on this matter may arise from

not being able to see networks as offering anything more radical than

a passive file serving mechanism. For example, Kay's recent review of

networks in higher education converges on the recurrent theme of more

efficient information delivery: '..pervasively networked computers

will soon become a universal library' (Kay, 1991, p. 106). Gardner

(1989) may have a similar vision of the "electronic campus" when he

concludes: 'There is at the moment no irrefutable series of arguments

which demonstrates that electronic campuses are necessarily better

places for staff or students.' (1989, p.348). Unfortunately, it is

hard to promote convincing arguments when there are so few model

systems to refer to.

As it happens, an educational interest in network infrastructures is

more active in North America. Thus, some campuses there have made

considerable investments in distributed computing for teaching:

notably, Carnegie-Mellon University (Hansen, 1988) and MIT (Stewart,

1989). Technical and academic strategy within the major US

initiatives has been summarised in a review by Issacs (1989). Issacs'

review indicates that networks often only service a fairly traditional

teaching strategy: they deliver computer-aided learning packages to

students at workstations. However, it is also apparent that these

computing environments, to some extent, have made possible new and

imaginative forms of coordination within their respective communities.

Some of these possibilities have been reviewed in more detail by

others. In particular, Kiesler and Sproull (1987) have published a

volume of essays describing the impact of networking at

Carnegie-Mellon. Their commentary suggests that the computing

infrastructure is transforming patterns of communication within the

campus community.

At first glance, this might seem likely and desirable: electronic mail

opens up otherwise sluggish lines of communication, course material

can be made widely accessible when and where it is needed, student

assignments can be transferred within this medium - and so on. Hiltz

(1990) has evaluated four courses intensively run in this way, with

most of the communication managed by electronic distance teaching

methods. These 'virtual classrooms' are favourably judged by students

who report a greater sense of participation and more access to the

course tutors. However, some caution is necessary: the student

constituency was self-selected, Hiltz indicates access to computing

resources has to be very good and considerable commitment is required

from staff and students to master the new tools. Final grade scores

were no better or worse than controls. Other reports of successful

classes suggest a similarly mixed picture. Philips and Santaro (1989)

describe the experience of running four parallel speech communication

classes that made extensive use of electronic mail and bulletin board

systems. The system was heavily used, the course ratings high, and

the outcomes good. Yet, engagement was localised: a third of the

class made extensive use of the computing resources while 20% barely

used them at all. The mainframe computer system at its core was

somewhat cumbersome and this may have made workloads high. Evidently,

the usability of systems will be relevant to the progress of these

initiatives. Although, in the present case, it is reasonable to

suppose that the distribution of course engagement was at least what

one would expect from more traditional teaching methods.

These initiatives are slow to spread. Indeed there are numerous

commentaries suggesting opposition to developments of this kind. Hiltz

and Meinke (1989) indicate that their virtual classroom provoked

'active resistance' from many faculty members. McCreary describes the

University of Guelph's long-standing commitment to using computer

conferencing and notes that its adoption has not been pervasive. In

particular, there is a lack of presence on the system of senior

members of the community. Komsky (1991) describes efforts in one

university to deploy email for more administrative purposes,

commenting: 'Despite a high degree of computer literacy and frequent

use of computing for other applications, these faculty have been

unwilling to alter their existing communication patterns to include

electronic mail.' (p.311). So, where faculty (or students) are not so

computer literate, we can presume the resistance will be greater

still. Moreover, that obstacle is not easy to anticipate. Stewart

(1989) indicates that resistance has been evident even at MIT where

'technical blood flowing through their veins' has not ensured that

members of this community would be active computer users.

Of course, all of these cautions are made from within traditional

university settings. Where the student constituency is more

geographically dispersed, then it is likely that computer-mediated

communication will be particularly successful. There is some

indication that such success is possible (Harasim, 1990; Mason and

Kaye, 1989). The UK Open University is one of the largest and most

experienced distance teaching institutions in higher education. It

has reported some success with computer conferencing (Mason, 1989).

However, the success was for students taking a technology course

and the conferencing resource has yet to spread further into the

University's teaching program. Moreover, there are indications that

such course-related communication may not always be experienced as a

liberating opportunity for equable student participation. Grint's

(1992) report of a small group of users indicates that the resource

was not necessarily radical for them in this way. Partly, he argues,

the difficulties arise because student participation depends upon a

community 'solidarity', and that is not easily created within the


Yet, in terms of our present theme of "collaboration", this is very

much the quality of experience that it is hoped interacting *through*

computers can furnish. Here, I am hoping that communicative

activities pursued through this technology might create for

learners new forms of mutual knowledge. Such achievements could then

underpin further collaboratively developed understandings: much as

interacting *at* computers and interacting *around* them might. My

brief review of developments within the present university community

suggests that computer infrastructures do not yet have a significant

role in teaching and learning practices. This may reflect a lack of

faith and funding from policy makers. It may reflect interfaces to

the technology that are still too cumbersome or students who, in any

case, are still uncomfortable using computers. It may reflect a

failure to achieve critical masses of participation within the


There is one further consideration relevant to making such initiatives

work; it concerns the institutional level at which network-supported

practice is organised. In North American Universities, network

structures may be generic resources oriented to the campus community

as a whole. It is then possible for these resources to be

appropriated into particular teaching units where they are adapted to

local needs. Thus, the *course* becomes the level for organising

interactions through this technology (cf. Barrett and Paradis, 1988;

Hiltz and Meinke, 1989; Kinkead, 1987; Landow, 1990; Philips and

Santoro, 1989). In Britain, the natural unit of organisation may be

the *department*. This is because most administration assumes that

individual students have strong affiliations to such an academic body

- sometimes, in the case of combined degrees, they may have more than

one such affiliation. Evidently, modularisation of courses (currently

underway) is likely to loosen these kinds of link. Perhaps this is

all the more reason to be contemplating new strategies for sustaining

them under such pressure. Many students will continue to pursue study

plans that focus on a circumscribed academic subject: yet the options

of modular structures may serve to undermine their sense of

association with a fixed cohort of peers pursuing a common set of

goals. While, computer-mediated communications may well be conceived

to meet the typically course-focused needs of these students, perhaps

it might also help recover some of their sense of involvement in a

larger academic community.

I will describe below a case study that is organised at this level.

It is unusual in that it represents a consensual effort to create

structures for interacting through computers at the level of a whole

teaching department. The observations are also interesting, as they

cover a period during which all other circumstances of the department

were stable. Moreover, this was a lengthy period (5 years) and thus

conclusions are not subject to the limitations of "snapshot" research:

observers of electronic communications in office settings have

cautioned against drawing conclusions from brief accounts relating

only to the early period of innovation (Rice, Grant, Schmitz and

Torobin, 1990). Space does not permit a full description of this

project and - as it is my own department - I will rely, to some

extent, on participant observations. However, the exercise should be

adequate to make some general points about the problems and

possibilities of collaborative structures realised in this network




University case study: context and implementation



The observations that follow refer to a medium size university

department teaching Psychology students entirely on its own premises.

These students come from both Arts and Science backgrounds. The group

of interest here are those combined cohorts of 40 or so students who,

in any of the years of a three-year course, are registered for an

Honours degree in Psychology. Each such group was recruited into a

networked communication structure during their second year of study;

other undergraduates taking various combined degrees involving a

component of Psychology could also make use of the network resources

but they experienced less organised encouragement to do so. The

exercise also involved all academic staff, a small technical staff

(around 7 people) and a postgraduate and research community (around 15

people at any one time). My remarks here cover a 5-year period

(1986-1991) where circumstances were notably stable. So, the core

curriculum and teaching strategy in the Department happened not to be

altered during this time. Student numbers remained level and the

computing infrastructure itself did not change in fundamental ways.

Moreover, academic staffing remained constant (9 lecturers) with only

one (temporary) appointment changing during this period.

It seemed to the staff in this Department that Durham provided an

attractive context to explore network-based resources at the

departmental level. Even in 1986, Durham enjoyed a well-developed

computing infrastructure. This partly arose from the University's

involvement in a regional consortium that had taken early steps to

promote campus-wide, multi-user services. These were based around the

innovative and powerful operating system, MTS. Thus, access to a

central service was very widely available: in libraries, in

departments, from telephone dial-up services, and in student

residences. The strongly residential character of Durham (70% of

undergraduates live in college-style accommodation) was significant in

several respects. It ensured that most undergraduates would have

residence-based access to network resources. However, it also defined

part of the reason these resources were thought educationally

valuable. A strongly residential student community leads students to

make their colleges the organising setting for social life.

Departmental staff tend to note that this can undermine the ease with

which shared academic interest catalyses student relations - and,

perhaps, gets explored within them. For example, 40 final year

Psychology students would be evenly dispersed across a dozen

residential settings (and some would live in private accommodation).

This offers most of them a rich social life, but usually not one that

exploits common academic commitments. In 1986, an unannounced poll of

the Psychology students assembled for final year registration,

revealed that, on looking round the room, the average student could

only identify by name half a dozen peers from this common academic


In 1986, this department needed to upgrade computing equipment that

supported teaching. The decision was made to invest in 30 new

connections into the University network. Terminal access would then

be available to all staff and there could be generous facilities in

public spaces for students. It should be stressed that these

innovations were achieved within the normal budgetary framework for

teaching support: I am describing relatively low-technology

innovations within the reach of many university departments - should

they so chose. However, as is widely appreciated in commercial

settings (eg., Strassman, 1985), the real initial investment is not

usually concentrated in hardware provision, but in handling the human

issues of introducing a set of new working practices. The present

example was not unusual in needing what McCreary (1990) terms a

'diffusion manager' - someone taking responsibility for drawing

members of the community into acting within this medium. This was a

role that I partly took upon myself and, therefore, can comment the

implementation process from a close association with it.

Interactive communication media are vulnerable to start-up problems:

where there is less than universal participation at the outset, this

usually imposes a high cost and lower benefits for those who do take

part. Markus (1987) has analysed this paradox and highlighted the

problems of creating an early critical mass. Success appears to

depend upon a small group of key users making a disproportionate

contribution - perhaps making themselves more openly available than

would otherwise be the case. In the present example, the teaching

staff could be said to have made this gesture in respect of the larger

community of students. Of course, these staff themselves have to

arrive at their motivation. The present situation may have been

unusual in that these particular individuals would be guided by

pedagogic commitment and, perhaps, by theoretical curiosity. This

might not so easily drive such an initiative in other academic

contexts. It might even suggest social science departments as likely

innovators in this area (Hiltz and Meinke's (1989) virtual classroom

was for sociology students and, at Durham, the one department so far

to have reproduced our own strategy is Sociology). Yet, the diffusion

manager in this case did find it necessary to create network

facilities that staff could not otherwise easily enjoy: in particular,

easy access to software for statistical analysis as well as central

printing facilities that improved on what was locally available.

Moreover, key administrators in the department made regular use of

electronic mail for management purposes.

Managing the involvement of students was in some respects easier. It

was seeded by building a requirement of usage into their coursework.

This was not shamelessly manipulative: these students had always been

required to learn the use of a statistical analysis program. All the

necessary resources for that exercise were located in the same

computer environment as the communication facilities under

consideration here. They were introduced in parallel early in the

students' second year. Also, a potentially prescriptive atmosphere

was, in a small way, defused by introducing a number of frivolous

resources (cf. Marvin, 1983). These included on-line jokes, a

database of student events, and randomly selected humorous items

delivered when signing off. Students also benefited by having

unrestricted access to a full screen text editor and printing

facilities which many of them were able to use for basic word

processing. Keyboard skills was a further need that had to be

met: typing-tutor programs were made available at workstations and a

professional instructor took a voluntary class in the department.

Even after 10 years of computers being in schools, it should not be

expected that contemporary undergraduates will find learning the

basic use of a new system easy or agreeable. A computer attitude

scale (Stevenson, 1986) was completed by all these students at the

start of each year. This instrument comprised 20 statements requiring

endorsements on a 5-point scale from "strongly agree" through

"undecided" to "strongly disagree". Figure 8.1 indicates that a

similar pattern of responding was reproduced in each generation. If

anything, the course-based experience of a networked environment is

correlated with a gradual softening of attitudes towards using

computers. Yet, it is noticeable that there remains a significant

unease over using this technology. The survey included supplementary

questions probing particular experiences that students felt were

associated with negative or positive feelings about using computers.

The most commonly cited category of negative experience concerned

feelings of being less competent than others or feeling resentment

over having had inadequate opportunities to gain expertise during

their education. Ironically, the second most commonly cited problem

related to bad experiences of being taught about technology in

previous contexts.

----------------- Insert Figure 8.1 about here -----------

There has remained a significant minority of students (around 15%) who

seem seriously intimidated by this technology and feel very reluctant

to use it regularly. Typically, they have mastered the small number

of basic actions that provide them with teaching-related material

through this medium (elaborated below). On the other hand, the

overall picture of usage suggests a buoyant situation. The use of

selected local commands (including signing on to the service) was

recorded in a system log file. To ensure anonymity, this data was

directed into a Computer Centre account and the identity of particular

users hidden - although their status as individual representatives of

a particular student or staff cohort was coded. Figure 8.2 summarises

the extent of system usage over this period - in terms of average

number of weekly log-ons by teaching staff, by second year and by

third year students. All teaching staff became and remained active

users of the network. So, it would be normal for them to log into

this system more than once a day. Student use is also active with

most students logging-on three or four times a week.

------------------ Insert Figure 8.2 about here ---------------

There were three kinds of network-based resource relevant to the

support of collaborative interactions. I shall define them briefly

before saying more about the fate of each in practice. (1) Electronic

mail furnished a tool for interpersonal communication. It also

supported communication to groups. Over 50 such group aliases were

defined. These mainly related to the various teaching units -

seminars and practicals, taught option courses, and so forth. Thus,

rapid person-to-person and person-to-group communication became

possible. (2) An information server was written to collate and

distribute documents relevant to teaching, administration and research

within the department. A menu-driven interface allowed users to

converge on files classified under such headings; they could then

view, copy or print this material. (3) Such common access to files

also allows more interactive structures. The concept of computer

conferencing mentioned earlier is a case in point. In the present

environment, a program ("Forum") was available that allowed the

creation of a conference (for example, it might concern a particular

taught course) and, then, the development of "topics" (conversational

strands) falling under that heading. Users could add their own

written contributions to a developing exchange. A simpler program

("Intray") was also written to allow self-defined groups of

individuals to share access to a single file. In some ways this

resembled an electronic mail program. On running the utility, a user

would see a list of files to which they currently had shared access.

Any one could be selected through menu-style reference and, then,

edited or printed. When signing on to the network, a message would

indicate whether any such shared file had recently been updated.

Observations below on the fate of these resources is based upon three

kinds of information. Firstly, some system-level programming allowed

usage statistics to be gathered. Secondly, student reaction was

polled at various times through interview and written accounting.

Thirdly, staff recounted their experiences informally to myself under

periodic questioning as a record of network activity was developed. I

shall also draw upon my own experiences as a participant who may have

been especially active in promoting network resources for

collaborative purposes - arising from a professional interest in their

value. The social-psychological distinctions I wish to make do not

map cleanly onto the system resources as summarised above. I shall

review the experience in terms of varieties of collaborative

arrangement - each of which might involve some mix of these tools.



Computer-mediated discourse

Under this heading I am considering how communications *through*

computers can support collaborations that stress interpersonal

exchanges. So, the resources of interest are those that help

reproduce the opportunities of that discourse typically enjoyed in

face-to-face discussion. Evidently, electronic mail is of special

interest. It provides an accessible means of communication that can be

referred to at times that suit individual users. Computer

conferencing provides a further framework for the support of

text-based discussion. However, in the present case, conferencing can

be dealt with quite quickly: it never proved a resource that could be

sustained. Part of the problem may have arisen from a user interface

that many people found cumbersome. However, even when thorough

preparation had been ensured and where it was formally built into the

work of project groups, it did not prove a popular or effective


The experience with electronic mail makes an interesting contrast.

There are two important differences to note. One is that using the

mailer program was never felt to be as difficult. The other is that

electronic mail *was* widely used and widely appreciated. In annual

surveys at the end of each academic year, students reliably cited mail

as the most engaging network resource. During the first two years of

this initiative, the mailer happened to deliver unusual levels of

feedback to the senders of messages. Thus, it could be determined, at

any given time, who had seen a particular message. When a message was

sent to a group, this feedback was listed for all its members. By

sampling such group-directed mail that happened to be sent by myself,

it became possible to estimate the typical delay between sending an

item and it being widely read within some target constituency. I

performed such calculations in terms of "working hours" - assuming a 9

am until 5 pm day - for a message to have been read by at least half

the receiving group. During the second year of the project this

message half-life was 2 hours for lecturing staff and 10 hours for

students. The point I wish to emphasise is that these statistics,

along with student feedback on the attractions of email, indicate that

the resource was accessible and effective (see also Figure 8.2). We

might therefore expect it to be widely used for tutorial exchanges.

Indeed, staff anxiously expected this. At the outset of this

initiative, lecturing staff were invited to write brief declarations

of how they expected the use of networking resources to enter into

department life (these kept sealed and only examined very recently).

A common concern was that electronic mail would open up a flood of

student contact that would be difficult to manage.

In understanding the network-based practices that have actually

evolved, there are other background observations that are relevant.

These concern the extent to which this community already supported

strong traditions of interpersonal collaboration - in the form of talk

between staff and students and talk among students themselves. These

traditions were sampled in two ways. Firstly, academic staff kept

daily diaries during three one-week periods just before the networking

initiative was launched. In these diaries, staff availability and

contact with students and colleagues was logged. Secondly, the extent

of student collaboration was assessed with an unannounced survey of

how one particular piece of coursework was managed.

Staff diaries summarised the opportunities that students would

normally expect for informal advisory contacts - and the extent to

which such opportunities were taken. The pattern of statistics was

very even across staff. When time allocated to formal teaching,

official meetings, and research-related absences are subtracted, an

average of 4.7 (range 3 - 6.8) hours per day remained as time during

which lecturers were potentially "available". Undergraduates claimed

a relatively small proportion of this time. 17% of it was absorbed in

meetings with other individuals; 12% of it was taken up by

undergraduates. This was enjoyed by a daily average of 2.9 students;

around half of which were students being supervised in specialist

project work by that member of staff. Moreover, for most staff this

number was made up of a significant core of individuals seen quite

regularly (usually in relation to their supervised projects). These

statistics suggest substantial opportunities for staff-student

tutorial contact, but relatively modest uptake. The same picture

emerges if we approach the question from a more student-centred


The student-focussed exercise considered the natural history of one

substantial piece of coursework required towards the middle of the

second year course. This work was associated with fortnightly

staff-led seminars, each involving about 6 students. The work set may

have varied between groups but was to be submitted at the same time.

At the submission meeting, students completed an anonymous

questionnaire concerning the scheduling and support of this piece of

work. The assignment needed to be done within a 12-14 day time frame,

but it would have been the only written work required in that period.

It was assessed but the mark did not count towards degrees. For

present purposes, I only wish to comment on the collaborative

relations that might support this work. My point is that a project of

this kind was typically a solitary achievement. 72% of the students

had no further conversation with their tutor in relation to the

assignment. And 62% had no discussion with their peers. Moreover, of

those that did, three quarters commented that these conversations were

"very brief" or "cursory". The relative lack of collaborative

engagement with peers is supported by a survey of study time

allocation carried out within the same class. Students reported only

1% of their study involved discussions with peers, and over 75% of the

class claimed no time at all invested in this form of learning.

Hounsell's (1987) study of undergraduate practices of essay writing

endorses these findings: he also discovers 'no substantive discussion

of essay writing amongst peers' (p.113)

I am dwelling on this background context of working practices because

it aids interpreting the subsequent impact of computer-based

communication media. In general terms, a description of the present

initiative might be helpful in suggesting likely outcomes elsewhere;

but only if the present local circumstances are fairly fully

articulated for purposes of comparison (not because those

circumstances are "typical"). In this spirit, there is one further

dimension of this situation to comment upon. That is, the traditions

of communication within this setting as they were interpreted by the

students themselves. To get some sense of this, 10 final year

students were randomly chosen and approached to keep a reflective (but

anonymous) record of their learning and interactions across this year.

They were encouraged to be alert to attitudes among their peers and

incorporate these impressions into an accounting of student

experiences as they perceived them. At the outset the terms of

reference were fully discussed and organised in relation to a number

of open-ended, orienting questions concerning: 'activities' (the

processes of study), 'contexts' (the locations and resources for

study), and 'people' (relations with staff and peers). The exercise

generated 10 accounts, each around 2000 words in length.

The features of these accounts that are most relevant to present

concerns are those about which there was most unanimity. Firstly, it

was widely claimed that the course was engaging and respected.

Secondly, staff were regarded as approachable and the social

atmosphere relaxed:

Student C: Generally, staff are pretty accessible: one merely knocks

on their door...many members of staff are only too pleased to help out

with questions.

Student E: Students feel fairly few pressures compared to other

students in different departments mainly due to the fact that the

staff are fairly accommodating and understanding.

Thirdly, although less widely identified, it was remarked that

discussion among students was limited. Where this was claimed to

occur it was often expressed in terms of students monitoring

their peers: to check that they were themselves "keeping up", or to

confirm that others were experiencing similar classes of difficulties

and pressures. Some students identified the need for some kind of

formal structure that could promote what they perceived as a


Student A: While Psychology is a fascinating subject, very few

students actually sit around amongst themselves and talk about, people need timetabled tutorials in which to talk about


Student C: The lack of communication between psychology students is

depressing. Somehow, discussion about the course "outside of hours"

as it were should be encouraged. Perhaps Psychology students should

only be accepted by certain colleges, thereby creating a hard core of

psychologists with increased opportunities for communication.

Student perception of the collaborative nature of their learning seems

in keeping with the formal sampling of this described earlier. Their

reference here to limited study-centred peer interaction is consistent

with those observations. However, their sense of an accessible and

"collaborative" staff might seem less consistent. The diaries and

student work summaries have suggested that, in reality, a relatively

small number of students have tutorial contact with staff outside of

formal teaching settings. Of course, there may be no real

inconsistency here. Students may detect a receptive teaching culture

and yet chose not to take advantage of it by initiating frequent

collaborative conversations.

There are good reasons to expect that electronic mail between

students and staff might flourish in such a setting. For one thing,

email has been widely described as a communication medium that causes

users to be less inhibited. This may facilitate exchange in

educational contexts where there are perceived status differences.

Keisler, Siegel and McGuire (1984) have noted its effectiveness in

this sense with undergraduates. Some commentators (Turner, 1988)

have even expressed concern at an excessive loosening of manners that

the medium can encourage among university users. However, our own

experience has not been one of electronic mail radically altering

patterns of communication. This conclusion arises from summary

accounts elicited from staff, sampling of individual student mail use,

and more detailed participant records of my own.

All teaching staff took part in a lengthy interview one year into the

present period. The topics for discussion were circulated in advance,

responses were noted during the interview, and summarising accounts of

these confirmed or negotiated later. There was universal agreement

that the network resources had been a great asset. Several staff

commented that they could not remember how they managed previously;

others remarked they had not at first appreciated the significance of

the initiative and had been surprised at how effective the facilities

were. Also, they all agreed that electronic mail was a significant

advantage in support of their teaching. However, the reasons cited

for this all concerned opportunities for contacting students -

delivering course material, inviting meetings, cancelling meetings,

calling in missing books and so forth. No staff admitted to

participating in anything like tutorial exchanges over this medium.

The most widely cited examples of student-to-staff academic queries

related to the supervision of individual practical projects. My own

detailed records of mail to and from students echo these reports

across the five year period discussed here (and beyond). In every

year, mail relating to these final year projects outnumber all other

incoming student mail. Even then, they are relatively scarce: in an

average year they amounted to nine messages from a typical yearly

group of five students. Almost all other incoming mail took the form

of two-turn exchanges. These arise within the context of some

particular academic contract (usually a recently required piece of

work): I can only trace six items that are context-free academic

inquiries - such as might arise as reactions to a particular lecture

or reading.

Thus, the summary picture from staff describes a pattern of

communication very similar to that already apparent in their diaries

of informal face-to-face student contact. There are relatively few

unsolicited queries and much of the communication grows out of more

sustained and personalised obligations associated with supervising

specialist practical work. Six final year students who kept logs of

their own mail confirm this pattern. The student with the largest

number of outgoing email items to staff reported 25: most of which

where brief administrative matters. All reported more outgoing

mail to other students and more incoming mail from staff. A small

amount of student-to-student mail (on average around 6 items in the

year) were course-related. However, it is clear that the availability

of this medium does not transform or greatly amplify existing patterns

of collaborative communication. This is true even in circumstances

where the need for greater staff and peer collaboration is acknowledge

and even after the following operating conditions have been satisfied:

(i) a reasonable period of network institutionalisation has occurred,

(ii) electronic mail has become familiar, regularly used and

appreciated, (iii) the freedom to approach staff and peers in this way

has been well advertised.

For these reasons, I am inclined to view unstructured use of

electronic mail as a rather limited resource for extending the

collaborative experience of learning - at least, within the

traditional culture of British department-based teaching. Arguably,

it is too cumbersome ever to suit tutorial-type dialogue and

conceptions of computer-mediated communications in education surely

need to look beyond this simple analogy. Only in distance teaching

contexts might this tutorial application make sense (Cf. Verdun and

Clark, 1990). Electronic mail remains useful to coordinate other

activities, including scheduling face-to-face meetings. And it

*could* be useful for student-driven, two-turn exchanges that might

not otherwise occur. That this tends not to happen reflects a need to

refashion other aspects of the culture of teaching and learning:

electronic mail does not seem, by itself, to create new conditions for

empowering such collaboration.




Network structures for joint activity


One enthusiast for computer-mediated communications in education

comments: 'Knowledge is not something that is "delivered" to students

in this process, but something that emerges from active dialogue among

those who seek to understand and apply concepts and techniques'

(Hiltz, 1990, p. 135). Many such innovators properly wish to

dissociate themselves from 'delivery' models of educational practice,

fostering instead opportunities for participation in educational

discourse. Yet, it may be optimistic to assume that new arenas of

dialogue will naturally open once the electronic infrastructure and

tools are in place. Certain persuasive studies within distance

teaching settings might encourage this belief. However, empowering

collaborative practices within more traditional cultures of teaching

and learning may require going beyond simply giving access to these

new conversational tools. Productive educational dialogue usually

depends upon a previous investment: a background of more carefully

cultivated social practices. So, in the end, the challenge for

developing collaborative interactions *through* computers may be one

of re-mediating established practices: using these tools to

transform existing forums of socially-organised teaching and learning.

Once such new structures are discovered, then they may motivate

activity within the dialogue-supporting devices we have been

discussing above. The purpose of this section is briefly to elaborate

this argument.

I will mention two exercises carried out within the context of my own

teaching. They each made use of network resources to structure a

traditional activity in a more collaborative fashion. Neither

depended centrally on the dialogue functions of electronic mail and,

thus, they take us beyond the examples of the preceding section. The

first example concerns support for the management of written

assignments. The second concerns student-led discussion groups. They

illustrate something of what is possible, but they also illustrate an

optimistic assumption that new structures can be easily bolted on to

existing learning practices.

To introduce the first, I shall return to the exercise described

above, in which students were confronted with an inquiry to

characterise the natural history of an assignment they had just

completed. I commented earlier on the solitary nature of this work.

The other striking outcome was poor time management. This might have

been anticipated, as student procrastination is a well documented

problem (eg., Silver and Sabini, 1981). The present exercise required

a breakdown of the work in terms of time spent on preparatory reading,

drafting and final writing; students described how these activities

were distributed across the two-week period. Two particular

observations are worth highlighting. (i) The proportion of time given

to preparing plans and drafting was relatively small (around 15%) and

(ii) most work on the assignment was concentrated in the latter part

of the available period. Thus, 65% of the work was done within the

last four days and 20% was done within the final 24 hours. Only a

fifth of the students had given the work any attention at all within

the first week of the allotted time. Arguably this is a problem of

procrastination rather than lack of commitment. On average, ten hours

was spent preparing these essays: this is a fairly generous investment

in absolute terms, the problem is that its distribution across the

available period is skewed.

I suspect the underlying pattern of management endorses Hounsell's

conclusions from an interview study of essay writing among history and

psychology students:

...there was no substantive peer discussion and communication from

tutors to students appeared to be largely formal, post hoc, product

oriented and limited in scope. Essay writing thus appeared to be a

central assessment activity but a peripheral pedagogical one.

(Hounsell, 1987, p.118)

The situation invites, among other possibilities, more imaginative

support from tutors, and support that encourages more effective use of

preparatory time. The writing of an assignment might become a

collaborative activity, if procedures could be instituted to draw

tutors into the writing *process* (rather than them only entering at

the end as commentators on the *product*). Evidently this does not

happen spontaneously: observations made above on tutor-student

contact indicate that students do not seek out such involvement.

Neither does technology greatly assist by offering an accessible and

legitimate new communication medium (electronic mail). The problem

seems to require establishing a recognised (and economical)

collaborative practice: some set of procedures whereby joint

involvement in the preparing and planning of work becomes possible.

This might require the construction of a more accessible object of

shared reference than might be naturally achieved within a brief

face-to-face (or computer-mediated) tutorial dialogue. Participation

in a networked computing environment might make this possible.

My own exploration of this possibility involved two teaching forums.

The first was the same assignment-requiring seminar referred to above

when discussing student procrastination. Students were asked to

produce a first plan for their written assignment during the following

week. This could then be developed iteratively within further

exchanges, if that seemed useful. The exercise either used electronic

mail or the Intray file-sharing program. Perhaps such a procedure

could be effected with reference to paper drafts, but its management

would be more difficult. Collaborating through the computer network

means that the task can be turned around very promptly and editing

facilities mean that comments can be effectively interleaved into

existing text. This affords something more like a dialogue in

relation to the topic under development. Moreover, when the time

comes to discuss the finished product (as might occur in a seminar

meeting), participants will have a richer understanding of the nature

of that product through shared access to its conception.

Such procedures work, up to a point: the medium supports the activity

and it does not seem to over-stretch a tutor's commitment. Yet, the

collaboration is hard to precipitate and hard to sustain. Of course,

other tutors might do so more skilfully and more successfully.

However, I suspect that there is more to the problem than this.

Additional effort is required to identify procedures that could be

effective: procedures that make sensible contact with existing work

practices and, yet, can then act to enrich or extend them. In this

case, formalising the development of an essay in successive text

drafts was accessible and useful for some students, but unfamiliar and

awkward for others. The discrepancy in reaction makes it more

difficult to sustain the initiative: for the culture of seminar groups

tends to encourage procedures that *all* members can appropriate.

However, the obstacles may be substantial. To consider this, I have

implemented the same procedure with different operating circumstances.

In particular, I have encouraged computer-mediated dialogue around

text plans for a final year assignment that was part of formal degree

assessment. Thus, this work was important; and because the completed

assignments were blind marked and the class was large (35), there

should be limited fear of tutor prejudices carrying over from planning

dialogues to influence ultimate assessments. Yet, whenever it has

been instituted, this procedure has never attracted more than one or

two collaborating engagements. Moreover, passing the responsibility

to a postgraduate assistant - who might be less threatening - makes no

significant difference.

One of these postgraduates raised this topic during interviews with

randomly selected students following completion of their degree.

From the transcripts, it became clear that the assignment was taken

seriously. It was also regarded positively (coursework assessment for

degrees being unusual in the Department). Procrastination was

frequently cited as a problem: to submit material for discussion close

to the deadline was seen as only exposing one's own inefficiency. A

further problem was simple lack of exposure to an interaction of this

kind. These students were one year further into their undergraduate

education than those described above: procedures for collaborative

interaction around the development of an assignment were still more

unexpected and unfamiliar. Thus, the networked environment does offer

a resource for this possibility and, for a small number of people, it

can be effective. But, for the majority, more preparatory effort has

to be invested in making such structures work. The *collaborative*

involvement of tutors and students in ways other than

institutionalised discussion forums (seminars) remains unfamiliar.

I will outline a second collaborative structure that is more

student-centred and which also suggests a similar conclusion:

networks furnish real possibilities for new structures but they may be

hard to launch as *widespread* practices. This case case concerns 35

or so students taking an optional course in their third year. The

class is semi-formal; partly, because it incorporates one hour each

week when small (self-selected) groups within the class take

responsibility for leading a discussion. Encouragement is given for

students to exploit these groupings as a basis for creating their own

structures of mutual support. This rarely happens. The interviews

with graduating students mentioned above confirm this. However, I am

aware of two such groupings that did sustain such activity, and that

did draw upon collaborative support through the computer

infrastructure. These groups enjoyed regular meetings at which a

nominated scribe summarised what they had discussed and understood.

These were mailed to the tutor (or placed in a shared Intray file) and

became the potential target of some further dialogue - either within

that file or at other times when the participants (and, perhaps, the

tutor) met.

What is notable is that the experience was very potent for those

participants. But, also, this venture was notable for being unusual:

few students easily adopt this manner of collaborating. I have

highlighted the examples above to stress two points. (1) The computer

network does suggest novel structures through which collaborations can

be supported: the networked technology furnishes a resource that helps

build up productive shared understandings. (2) The promise of such a

resource can only be assessed in relation to a background culture of

collaborative practices. This is merely to stress that, when deployed

within traditional undergraduate settings, the technology is

unlikely to function like a magic bullet.

Yet it *can* be subversive in these settings. So, other colleagues in

this Department recently have reported localised successes in using

the computer network to support small group collaborations within

their advanced (third year) teaching forums. What seems to help

precipitate this are ingredients not included in the relatively

unstructured initiatives described above: there may need to be a

degree of enforced group responsibility and also a well-specified

purpose or goal. Then, it is more likely that the shared computer

space will emerge to offer a valuable repository for group-based

understandings, and it is possible that these can become an authentic

part of common knowledge.




Participatory structures


Bruner (1991, p.3) quotes Nobel prizewinner Harriet Zuckerman as

suggesting that the chances of winning the prise increase immeasurably

if one has worked in the laboratory of somebody who has done so

themselves. With this observation, we are alerted to the significance

of cultural contexts for learning. This requires us to think from yet

another perspective about the technological resourcing of

collaborative structures. In discussing interactions *through*

computers, I have touched on possibilities that reproduce some of what

happens during interactions together *at* problems. Thus, electronic

mail might support the creation of shared understandings through

affording approximations to conversational dialogue. I have also

discussed possibilities involving more diffuse collaborative

structures: in particular, possibilities based upon using a common

computer space to maintain (asynchronously) certain documents that can

become shared objects of reference. Under the umbrella of

"participatory structures", I wish to introduce the idea that

interactions through this technology might be relevant to supporting

cultures for learning - in the sense that Bruner (and others) identify

them. In this case, more overarching shared frameworks of

understanding are generated; accessing these frameworks may help

create platforms from which new understandings get collaboratively


Theorising in the socio-cultural tradition recently has been

influenced by anthropologists who describe informal settings for

learning. Lave (1988), in particular, has studied such settings;

thereby conceptualising teaching and learning in terms of learners

encountering activity within 'communities of practice'. Such social

structures, it is argued, relate very broadly to educational agendas:

'Even in cases where a fixed doctrine is transmitted, the ability of a

community to reproduce itself through the training process derives not

from the doctrine, but from maintenance of certain modes of

coparticipation in which it is embedded' (Lave and Wenger, 1991,

p.16). Such a perspective has encouraged interest in apprenticeship

relationships. However, Lave cautions that the master-apprentice

terminology should not become a disguise for traditional teacher-pupil

organisations. Instead, she toys with an alternative vocabulary that

dwells more on the fabric of social relations and their its role in

transforming persons and practices. For example:

Newcomers develop a changing understanding of practice over time from

improvised opportunities to participate peripherally in ongoing

activities of the community. Knowledgeable skill is encompassed in

the process of assuming an identity as a practitioner, of becoming

a full participant, an oldtimer. (Lave, 1991, p.68)

I believe this vocabulary is helpful for thinking about the conditions

of undergraduate learning. The present influence of psychological

theory on University education is rather limited (Laurillard, 1987;

Saljo, 1987). It sometimes surfaces in a concern to cultivate in

students more effective "study skills". Yet, research suggests that

such skills can not be abstracted and taught independently of subject

content or meaningful contexts (Ramsden, Beswick and Bowden, 1987). A

more useful theoretical perspective may be one that considers the

nature of students' engagement with the practice of their discipline.

This issue concerns the participant structure of learning as it might

be created within, say, a university department. Our particular

interest here is whether technological resources can contribute to

enriching such structures.

What might this involve in the current example of a Psychology

department? It must involve some kind of exposure to the *doing* of

Psychology - an encounter with the subject in a context where it is in

progress. Fifty years ago, Lewin was stressing that learning is most

effective when pursued collaboratively within communities of practice.

He argued that watchmakers becoming carpenters involves more than

them learning how to use certain tools: it involves coming to swear

like carpenters, to walk, eat, and see the world from the carpenter's

point of view (Lewin and Grabbe, 1945). In an undergraduate context,

this might be achieved, in part, by a certain simulation of the

discipline's professional routines. It is tempting to demand of

students scaled-down versions of professional practices: such as

research reports, conference-type presentations, traditions of peer

commentary and so forth. Yet, the most potent learning experiences

may arise from opportunities to witness the art, values, tricks and

vision of practitioners as they actively engage in the subject.

Exposure to this culture affords newcomers a chance of 'legitimate

peripheral participation' (Lave and Wenger, 1991). It might allow

them to approach the status of oldtimers 'through a social process of

increasingly centripetal participation, which depends upon legitimate

access to ongoing community practice' (Lave, 1991, p.68).

I believe that considering the management of communication structures

within an educational setting will play some small part in

transforming it into an accessible 'community of practice'. There are

some pointers to be drawn from the undergraduate example that has been

discussed in the present section - although, admittedly, they are

modest. One network-based resource that relates to this purpose is

the information server: the tool that collates and distributes

documents relevant to the life of the department. Teaching material

is one category of material that dominates this database. At a

superficial level, it is valuable to students as a repository of their

course-related documents. In the final weeks of the period described

here, there was an average of 900 accesses to this database per week.

However, the availability of this resource has done more than merely

duplicate existing paper-based operations. In particular, it has

lead some staff to incorporate novel material that would not

previously have been in circulation: their own lecture notes for

example, or (anonymous, and with approval) examples of previous

student approaches to tackling some topic of mutual interest. An

incidental effect of this is to create a more vivid birds-eye view of

departmental activity. This is achieved partly because the material

is centralised in a format shared by all users and partly because it

is topical and detailed in a way that traditional handbook summaries

are not.

One feature of usage that became clear from system logs was that

users consulted entries that were not relevant to their own

course-related study needs (this observation includes staff - some

of whom were active in cross-checking the curricula of their peers).

It was hoped that this curiosity among student users could be deflected

towards aspects of the department's activity less centrally related to

the undergraduate syllabus. Thus, the database came to include

material relating to research activity, recent publications, the

administration of the department, new library and equipment purchases,

minutes from staff (and staff-student) committee meetings, and so


This exercise in opening up windows onto departmental life is cited,

as with several other examples in this case study, to illustrate a

principle: namely, that of linking computer-based structures to

collaborative purposes. Thus, in the present case there may be a real

possibility of drawing individuals towards a closer appreciation of

the life of the department, as defined by its core concern with the

practice of Psychology. However, in reality, these materials

peripheral to the official syllabus have attracted very little student

attention. Again, the existence of such access does not, of itself,

create a new level of engagement with the local culture of the




Undergraduate network: concluding comments

The departmental case study sketched above illustrates three senses in

which collaborations may be supported by interactions *through*

computers. They all relate to joint concerns that are relatively

dislocated in time and space: collaborators are not co-present in the

familiar sense. Firstly, computer networks may facilitate some of the

dialogue-based processes that are a traditional part of forming shared

understandings: electronic mail is the most accessible tool to help

achieve this. Secondly, access to shared file space may resource the

creation of useful objects of shared reference: I have discussed the

example of staff and students collaborating on the conception of a

traditional assignment, and also the example of a tutor's

computer-mediated involvement with the products of student-led

discussion groups. Thirdly, collaboration may be resourced more at

the level of very loosely-coupled shared interests and understandings.

This is a matter of electronic media creating for learners a greater

sense of awareness of overlapping purposes and a greater sense of

participation in a community of practice.

In relation to this final sense of collaborative support, it would be

naive to hope that investing in computer networking infrastructures

naturally met these needs. It is certainly possible that they could

undermine them instead. For example, electronic communication could

be deployed to obstruct the circumstances of joint understanding, it

could comprehensively displace rich interpersonal communication and

render relationships more remote. This has certainly not happened in

the quite long experience of the present case study. Yet, it is also

clear that widespread participation in the present initiative has not

had a productively transforming effect on collaborative structures


The important lesson of the example may relate to the *potential* of

computer-mediated communications when implicated in such

transformations. In the present case, I believe it has become clear

that much needs to be done in terms of confronting the limited culture

of collaboration that students bring to their undergraduate

experience. There is also much to be done at the level of

facilitating a richer intersubjectivity between the learners in this

situation and the tutors with whom an effective collaborative

relationship needs to be built. This is true even in a setting that

students may rate as relatively successful in this respect. In the

following section I make some brief observations regarding related

initiatives at much earlier stages of education. The cautionary

observations of a sociologist discussing electronic media in schools

may serve to link the above discussion and the one that follows:

...the professional optimists confuse community and communication.

What they forget or affect to forget is that communication is the

expression of a will, of a will to be or live together, which almost

invariably pre-exists it. In one sense, it is the community which

precedes communication and not the reverse, even if communication

may eventually reinforce the community. (Balle, 1991, p.107)








Networks may have been slow to penetrate teaching within undergraduate

contexts; they have made still less impact on practice in the earlier

stages of education. Yet, I believe the forms of collaborative

interactions made possible by interacting through this medium also

hold promise for younger learners. Again, my discussion will be

dominated by personal experience with one particular model system. I

shall draw on this to suggest ways in which computer networking may

underpin valuable collaborative structures in early education. The

discussion in this Section will be shorter, as many points generally

relevant to this form of activity have been introduced above.

Moreover, the case study to be discussed is still in a relatively

early stage of development.

In the first section below, I shall review the status of networking

structures within early education. Then I shall move on to describe

particular possibilities for practice as revealed in one primary

school setting. The discussion will focus on collaborations as they

might evolve in relation to children's early experience of writing.



Computer networking in early education

Few educational commentators have considered how the networking of

computers in schools can support innovative teaching and learning.

Handbook reviews rarely mention networking (eg., Salomon, 1989; Smith,

1989). A significant European conference on information

technology and schools gave no serious consideration to network

infrastructures (Eraut, 1991). Where they are considered, it is often

in terms of configuring an isolated room to create a more efficient

file serving mechanism (Henderson and Maddux, 1988). The schools

that do have networks are almost always in the secondary sector and

their networking is almost always clustered into a circumscribed space

- rather than integral to the school (Wellington, 1987).

Yet, in both primary and secondary settings, there has been an

attraction to the broad idea of collaborative interactions *through*

this medium. Curiously, it has inspired communication initiatives of

a different kind to those pursued in the universities. In schools,

interest has started from communication over wider areas - links made

outside of the school itself. In universities (as reviewed above),

the communication structures have been more at the local level. Thus,

in the school system there has been some lively exploration of

cross-site communication, but relatively little interest in local

structures of computer communication. While in universities, there

has been more interest in the local level and virtually no exploration

of cross-site links. Universities have been well enough equipped to

follow either route. While schools, in choosing to foster remote

communications, arguably have chosen the route with more cumbersome

technical demands. Their commitment may reflect traditional concerns

to challenge the often too-insulated nature of pupils' school


Communicating through computers to other institutions is made possible

if a school has a modem. This is a device allowing some computer to

be linked to a telephone line and, thereby, to exchange files with a

remote site similarly equipped. In practice, the usability of such a

link depends upon access to certain software tools that organise and

direct such communication. For this reason, early projects often

called upon university contacts to mediate the link - thus allowing

access to electronic mailers. These more effectively formatted and

targeted the files transmitted. More recently, commercial concerns

have acted to occupy this mediating role. In Britain, a news company

has launched the "Times Network for Schools": for an annual

subscription, this offers electronic mail as well as conferencing and

bulletin boards (Wellington, 1987). In North America, AT&T have

supported the development of a comparable resource aimed at schools

(Reil, 1990).

An early and influential exercise based around cross-site

communication was reported by Levin, Reil, Rowe and Boruta (1985).

Their rationale was to become widely endorsed by others. They argued

that this form of communication created a credible functional

environment to support student projects. It created new topics, new

audiences, and new purposes for such work. Indeed, some researchers

(Cohen and Reil 1989; Reil, 1985) were able to demonstrate that

participation in these exercises resulted in improved grade point

scores for writing. This seemed to result from an active engagement

in the editorial responsibilities that this form of communication


There have been similar case studies of successes in British schools

(eg. Keep, 1991; Wishart, 1988). Yet, the overall impression is that

these initiatives can be hard to promote and hard to sustain. Maddux

(1989) cautions the use of telecommunications before careful

considerations of cost and whether the same goals could not be

achieved more economically in other media. In reviewing a number of

such initiatives, Reil and Levin (1990) argue that users do quickly

discover the communication medium offers a qualitatively distinct kind

of interaction. But they also stress that a coherent need for the

exchange must be in place, and that here must be an active commitment

by some enthusiastic project manager. In reviewing a number of

computer-mediated school twinnings, Turnbull and Beavers (1989) also

refer to very careful planning and serious consideration of the real

compatibility of interacting communities. Keep (1991) conveys a

similar impression of the need for much energy in order to maintain

that critical level of exchange that sustains pupil interest. The

consensus seems to be that this medium can support valuable activity

if it is embedded in a larger framework of cross-site communication

and concern. The passing of text messages alone is too

decontextualised an experience to sustain serious pupil interest.

In the ten years since this idea of schools communicating through

computers was seeded, it is hard to conclude that an active and

effective culture of such communication has emerged. There are

still only pockets of success. It would be unfortunate if the general

principle of such collaborative school-based interactions were judged

on this variety of activity alone. From studying the reviews

mentioned above and from my own involvement in supporting a primary

school develop such links, I believe the goal of such cross-site

communication is worthwhile but it needs to be approached more

gradually. The problem is well crystallised by Hawkins, who comments:

In the educational world of the kindergarten through twelfth grade

there is, in most places, a relatively narrow band of activities that

these technologies make more efficient. As has been often noted, our

classrooms tend to be quite self-contained in terms of interactive

relationships and resources. (Hawkins, 1991, p.161)

Thus, it may be necessary to institute stronger traditions of *local*

computer-mediated collaboration before embarking on these more

ambitious communication exercises. It may be necessary to establish

practices of coordinating with the next classroom, before venturing

towards collaborations that span cultures and continents.

At least, this was the thinking behind a primary school networking

project established in cooperation with two colleagues, Geoff

Alred and Jack Gilliland. In the last three years, I have been

considering the kind of software structures that can support novel

forms of coordination at the level of a self-contained school site.

This, therefore, is a project focussed at communicating *through*

computers - within the forum of a single community. In the longer

term, one of its achievements may be to establish a culture of

communication that would be well-prepared for further-reaching




Open-plan computing

I have chosen the phrase in this heading to suggest a link between

traditions of social organisation in British primary schools and a

particular computer infrastructure to be described below. This

configuration for school computers is conceived to create an

environment in which innovative communication might flourish - rather

as might also be achieved within the physically open-planned layout of

many classrooms.

Our configuration of computers entailed installing a low-cost

(Acorn Computers) network throughout the premises of a school.

Various programs were then written to manage this network and to

sustain certain forms of pupil activity within it. The school

comprises 7 classes corresponding to the successive years of primary

education: from 4/5 until 10/11. All classes always have at least one

network station and, by negotiation, around 5 additional machines can

be clustered in areas close by any class to promote more dedicated

activity. As with the university case study described in the previous

section, this is a structure realised at a fairly low-technology

level: the aim is to explore what is possible within the resources of

a typical school, rather than to demonstrate some state-of-the-art

(but inaccessible) model system.

I will summarise the main features of the software environment to

convey the possibilities for collaborative interactions through this

technology. The network centralises a great many programs that are

characteristic of those used in any British primary school. Thus, the

widely-recognised advantage of networks for simply serving such files

to pupil-users is realised in this instance. However, there are four

particular programs that are crucial to developing the structure in

more educationally distinctive ways. I will describe each of these in

turn. (1) A menu-driven interface allows pupils to navigate their way

through the disc space of the central file server. Teachers may

configure menus to be relevant to activities current in their own

classes. Moreover, each pupil is allocated their own file space and

file space may be associated with short term collaborative projects.

(2) A conventional word processor suitable for primary pupils is

available to all users. (3) Pupils may nominate files to be

incorporated into a database structure that organises them into a

system of "folders". Folders are associated with particular pupils

and also with particular projects and/or classroom communities. Their

contents may be examined through suitable movement through the network

menu-structure. The route a child might make to enter a new item

in her folder is illustrated in Figure 8.3 that shows a schematic

sequence of screen displays. (4) An electronic mail utility allows

pupils to have their own mail box. Text messages may be dispatched

within the community and to addresses of outside sites (these are

collated and dispatched as an automatic night job).

--------------------- Insert Figure 8.3 about here ----------------

As in the case of university networks, when we are considering

structures for coordination, we might think first of electronic mail.

After all, this utility supports communication that is closest to what

spontaneously we associate with collaboration. It supports dialogue

between individuals. Although this school has still had only modest

experience with electronic mail, it is clear that this resource can be

engaging and effective. However, it has acted by gluing together

other more substantial coordinated activity - rather than being itself

the central feature of some collaboration. For example, it has proved

a very effective resource in support of a school-wide project for

producing a newspaper. Material could be distributed over the network

to reach groups working on different aspects of this common concern.

Moreover, this mailer (SJ Research's *Interspan*) allows seamless

links to outside sites. So, given a meaningful context, such as

producing a magazine or newspaper, cross-site electronic mail can be

effectively integrated with a well-grounded local project.

However, here I am concerned to highlight less obvious collaborative

resources. In particular, the folder structure that was described

above. At the moment, much of the material within this system

comprises pupil's written work of a fairly traditional kind. Yet, the

collation that is achieved - perhaps in conjunction with certain

orienting teaching practices - serves to create a stronger social

framework for these children's efforts. The most straightforward

transforming effect relates to the creation of a strong sense of

audience. Surveys indicate that much pupil writing in education is

crafted only for the teacher-examiner (Britton, Burgess, Martin,

McLeod and Rosen, 1975). Educationalists have recently promoted a new

paradigm of writing instruction that puts more stress on the

rhetorical nature of literacy (Davies, 1989; Dipardo and Freedman,

1988). The National Writing Project (1989) gives a range of examples

of how this can be organised. Yet, in practice, it remains hard to

foster a sense of the school as a 'community of writers' (Dunn, 1989).

Evidently, the network-based structures I am describing take us some

way towards an infrastructure that can support credible audiences in

this sense. For example, it may facilitate the kind of exercise

described by Somerville (1989) in which older children write stories

that become the focus of reading and discussion among younger children

in the school.

The existence of a structure to collate and make accessible pupil's

work invites other collaborative engagements. One that I believe is

poorly exploited within all sectors of education can be expressed in

the phrase "leaving tracks". Martin (1989) has discussed the general

place of furnishing for writers models of practice: these may become

resources for reflection and inspiration. This principle can be

usefully realised in the present computer structure. Insofar as each

school class reproduces the aims and efforts of its predecessors, it

may be a source of some insight to have access to the achievements of

previous generations. The network allows a class to leave these kinds

of traces behind it. I view this as a "collaborative" initiative, as

it involves a socially-organised concern to create and to coordinate

around objects of shared understanding.

The principle of leaving tracks applies to individual pupils also.

So, pupils in this environment build up a portfolio of work that

captures their own development as writers (or scientists, or

journalists etc) and this portfolio can become a resource for their

own reflections on this process (cf. Brown, 1989). Indeed, pupils can

be found referring to these folders spontaneously, apart from any

encouragement they might encounter to do so in the context of

teacher-led activities. There is a link in this to one feature of

traditional computer-based learning tools that has been widely

discussed in the literature: that is, the capacity of these resources

to make visible the *processes* of intellectual construction. Thus,

J.S. Brown comments: 'We are missing the real source of power for

computer-based tools: the computer can record and represent the

process underlying the created product' (Brown, 1985, p.182).

However, this idea is always cited in reference to underlying

processes of a private, cognitive kind. It has not been effectively

developed at a level that recognises the socially-grounded nature of

much intellectual construction. In a sense, what the organised,

archival character of this network affords is opportunities for such

process of reflection in the social domain.

It is clear that this example of interacting *through* computers

involves a resourcing of "collaboration" in the broadest sense of that

term. We may summarise the resulting opportunities as follows. These

structures can create circumstances where distinct groups of children

can coordinate their efforts across barriers of time and classroom

separation. The production of a newspaper is a vivid example of this.

The networking can also create structures that allow children's work

to reach further into the common knowledge of the school and, thereby,

play some part in motivating new achievements. There may also be a

deeper sense in which the structure can support such common knowledge:

this finds a parallel with the "participatory structures" discussed

for a university community above. It is possible that the

coordination and access that networks can create may serve to bring

pupils nearer to the socially-organised experiences of *being*

biologists, or historians, or dramatists, or whatever. This is a

point made by other researchers who have witnessed something of what

local level networking might achieve in early education (Newman, 1989,

1990; Ruopp, in press).



The examples discussed in this chapter should put into sharper relief

the sense in which the concept of "collaboration" may characterise

effective settings for learning. At the heart of this conception are

the activities of creating and exploiting structures of common

knowledge and shared reference. However, I am stressing that there

are various ways in which understandings can be held in common among

learners and teachers. Collaborating should not narrowly describe the

circumstances of small groups working together in time and place at a

focussed problem. In the institutional settings of learning,

important common knowledge exists as part of the diffuse background to

a variety of problem solving circumstances. The present chapter has

discussed ways in which computers may resource learners investing in

this broader framework of understandings.

So, in discussing interactions *around* this technology, we

encountered circumstances in which a transient learning community -

sharing a common working space - might collaborate 'through the air'.

Here, clusters of computers furnish focal points at which individuals

(or small groups) may concentrate their work; from these points they

may drift in and out of lateral communications with others similarly

engaged. Moreover, the computer tools they share provide concrete

referential anchors that may more effectively support such

collaborative talk.

In discussing interactions *through* this technology, the varieties of

collaborative structures encountered was still greater. In these

cases, the interactions supported may be displaced in both time and

space. I argued that to focus on tools for supporting versions of

(text-based) conversational exchange was to present too narrow a

characterisation of collaborative possibilities in this medium. An

accessible shared file space offers more than this: it offers a rich

environment for creating objects of shared reference. Particularly

when utilised in a framework of organised social practices, this

further illustrates how new technology can successfully resource

(rather than deny) collaborative possibilities.

It might be argued that many of these examples do not crucially depend

upon computers: the relevant collaborations could be supported in

other ways, with other technologies. I have two reactions to this

argument. The first is that whatever may be *possible*, in practice,

many of these collaborative structures simply are not being realised

in other ways. Thus, for example, there are various means whereby

primary pupils work *could* be collated and made into an accessible

and managed resource for their peers and successors but - by and large

- it is treated this way. These shortfalls often do await a

technology that simply makes a critical difference to what may

realistically be achieved. My second reaction to this argument is

somewhat strategic. Actively identifying new technology with

collaborative practices serves to deflect attention from certain less

agreeable models of computer-supported learning. If we do not wish to

see new technology transform the experience of learning into something

solitary and dislocated - then we must demonstrate that it has a

credible place in a more collaborative framework.