In this Chapter, I shall not attempt a comprehensive review of the

various themes that have arisen in what has gone before. I hope the

summary sections associated with each individual chapter achieve this

purpose well enough. Instead, I wish to extract a small number of

issues that I think recur throughout this book and make some closing

remarks in relation to them.

Of course, the book has been concerned with computers and education.

But that is a broad topic, and my own interests around it may appear

to be rather particular. Certainly, I have not attempted an overview

of all the various ways in which computers can now enter into

curricula. Their influence has become so far-reaching that such a

task would be too ambitious. In fact, some of the educational

software employed in my own observations may seem rather

modes illustrations of what computers can offer - these examples may

not reflect a state-of-the-art by which some enthusiasts might prefer

the technology was judged. However, my interest has not been to

describe these frontiers; my real starting point has been a concern

with the broad character of educational experience. Observations of

learners working at computers have furnished a helpful opportunity for

developing this interest. In some respects it has been useful that

the tasks observed have *not* been too sophisticated. This should not

imply that my interest in computer-based activities has been entirely

a contrivance. In the end, I believe that this technology does offer

a special potential for enhancing the experience of learning. I shall

return to stress this point towards the end of this Chapter.

Before going further, I should restate and reinforce the perspective

on educational practice that I have promoted throughout this book. In

Chapter 3, I suggested that many psychologists approach questions of

learning equipped with one of two pervasive metaphors: computation or

construction. These are broadly associated with, resepctively, the

academic traditions of cognitive science and of cognitive

developmental theory. My own preference has been to foreground a

third conception: collaboration. This I derived in Chapter 2 from the

socio-cultural theorising inspired by Vygotsky and his contemporaries.

For many developmental psychologists, this theorising has been

effective in drawing attention to the socially-grounded nature of

cognition. In part, that has stimulated attention to processes of

social interaction, as they relate to cognitive change. Yet, I feel

the empirical orientation towards this topic has been over-attentive

to ideas arising from the scaffolding metaphor - as it is often

used to describe the interpersonal dimension of instruction.

Socio-cultural research could gain from developing a greater concern

for the nature and management of intersubjectivity. I have argued

that a socially-grounded perspective on learning must acknowledge a

human capacity for projective understanding of mental states in

others. It is this that creates a deep human concern for the

formation of socially-shared cognition. Finally, I have argued that

such a concern is at the core of what is entailed in collaborating.



On the experience of collaborating

I have suggested that psychological research into collaborative

learning has made only modest progress. It has been dominated by a

preference for analysing collaborative exchanges by coding and

counting discrete social acts - usually utterances. There is some

value in such categorising systems but they fail to characterise the

sustained and discursive nature of many collaborative encounters.

Firstly, these encounters have a temporal dimension: they involve

partners in a protracted constructive process. Secondly, they involve

a commitment to coordinating action and attention in relation to some

focal point existing between collaborators. In fact, these two

features of collaborative interactions are related. Often, what it is

that is constructed in this protracted fashion is what can become an

"object" of shared knowledge; this then serves to focus or direct the

coordinated action.

The important point is that collaboration should be recognised as a

state of social engagement that, on any given occasion, is more or

less active and more or less effectively resourced. So, collaborators

may vary in their concern to create shared understandings; and their

circumstances of joint activity may vary in how readily they permit

such achievements to be brought off. The challenge is to discover how

discourse is mobilised in the service of creating joint reference; to

see how what is created gets used as a platform for further

exploration; and to see how the material conditions of problem solving

can be more or less friendly towards efforts after this mutuality.

In Chapter 7, I characterised children talking in ways that were

variously successful in achieving shared knowledge, and I described

computer-based problems that were variously successful in supporting

them attempt to do this. Within those accounts, I also suggested that

participating individuals might enjoy very different experiences of

this collaborative effort. This notion was prompted by recognising

that some children could claim a different degree of investment or

ownership in the shared knowledge that underpinned the joint activity.

However, the same reasoning raises a question about collaborations

that are successful in this respect: can they claim a particularly

*positive* affective tone. This notion is entertained in Argyle's

discussion of cooperation: 'One possibility is that the experience of

closely synchronised interaction is intrinsically rewarding, the

result of evolutionary pressures favouring cooperation' (Argyle, 1991,

p. 10). Unfortunately, Argyle only pursues this idea in relation to

bodily synchronisation - such as we might enjoy during dancing.

Yet, the pleasure of being 'sychronised' may extend beyond the case

of physical coordination. Our own casual experience of being

collaborators surely endorses the idea that being in posession of

shared knowledge (and being conscious of its mutual nature) can be a

positive experience. Perhaps this is all the more so, the greater our

awareness of an interpersonal investment lying behind its

accumulation: this may impart to the possession of shared knowledge a

certain agreeable intimacy. In fact, this affective dimension may

touch on something very basic in human nature. Trevarthan comments:

'But only humans have the kind of appetite a one-year old begins to

show for sharing the arbitrary use of tools, places, manners and

experiences' (Trevarthan, 1988, p. 55). This appetite we might chose

to read as a precocious expression of the motive for shared reference.

Furth and Kane's (1992) discussion of the social creation of mental

(narrative) objects seems to make a similary point for the preschool


These observations of very early manifestations of collaborative

interactions introduce the second of my closing themes.



On becoming collaborative learners

I am stressing the central place of socially-shared knowledge in

learning; and, thereby, the central place of collaborative

relationships within which such knowledge gets constructed. Yet, this

emphasis may seem in some tension with classroom realities. In

particular, ethnographies of classroom life tend to suggest that

peer-based collaboration can be very difficult for young children to

sustain (cf. the review in Chapter 6). Moreover, researchers who have

discussed with young children their experience of academic

collaboration tend to find that many pupils do not regard those

experiences very positively (Cullingford, 1991; Galton, 1990). This

might raise in our minds a question around which there is a long

history of speculation and reasearch (reviewed by Pepitone, 1980):

namely, are we basically cooperative or collaborative creatures? Our

attitude to this issue may colour our enthusiasm for organising

collaborative circumstances for learning.

Certainly, various anthropologists have argued that we have a

fundamentally cooperative nature (Hall, 1976; Trivers, 1983).

Moreover, the research of developmental psychologists might suggest

that there is indeed something basic in ourselves that is relevant to

this issue (a "basic-ness" that may or may not demand reference to

evolutionary arguments). I have in mind the basic capacity and

motivation to exercise an intersubjectivity. In fact, Trevarthan

regards interpersonal motives as 'the primary organisers of mental

growth' (1987, p.178). He describes a developmental course whereby

infants first manifest a primary intersubjectivity - they achieve a

harmony of affect and emotion with others. Then they move towards a

secondary intersubjectivity - they manifest concern to establish with

others shared reference towards external objects and circumstances.

Moreover, casual observation does suggest that the achievement of this

shared reference is a visible source of considerable pleasure for


Other researchers have traced this concern into the preschool years.

Toddlers may be particularly interested in objects that are seen to be

within the attentional span of a play partner (Eckerman, Whatley and

McGhee, 1979; Eckerman and Stein, 1982; Hay, 1979). Budwig, Strage

and Bamberg (1986) report a (rare) longitudinal study of these

processes as they transform from parent-dominated encounters to

encounters of shared reference with age-mates. They make the point

that parents make a distinctive contribution to these encounters.

They naturally work hard to create and sustain shared objects of

reference and this leads to a peculiarly rich platform from which new

learning and exploration can be supported.

On the other hand, Dunn's (1988) research on preschool sibling

relations suggest a sometimes less harmonious atmosphere within

playful family arenas. Hay, Caplan, Castle and Stimson (1991) have

observed the fracturing of collaborative attitudes among young peers:

they argue that maturity does bring a concern with personalised

interests, particularly in conditions of jointly acting in relation to

scarce resources. Of course, it is not clear that we should regard

such developments as symptomatic of some basic human striving for

individual gratification - in conflict with collaborative motives.

Again anthropological research is valuable in drawing attention to how

changing conditions of living re-direct basic human motives. Graves

and Graves (1986) review a rich variety of such changing conditions

relevant to the developing sense of self and community. To express it

starkly in their terms: new economic and social orders make it less

necssary to 'store food in the neighbours belly' (Graves and Graves,


It is important to see the technologies and rituals of modern

schooling as part of this cultural influence on development. The

infants' concern for shared reference in relation to material objects

may continue to feature in collaborative situations; they may reliably

recur through the remainder of life. But this secondary

intersubjectivity must also become elaborated in new ways with the

growth of a more symbolically-dominated intelligence. Children will

explore new forms of mutual knowledge that are more representational

in nature. In particular, shared reference may become increasingly

located in narrative structures. For during the preschool years,

children will naturally explore a variety mutual of involvements

with narrative (cf. Furth and Kane, 1992). Schooling may further

mobilise such playful collaborations; but it also puts a new kind of

demand on collaborators. They must come to generate shared objects of

understanding that are abstract, but which can not readily take on

a narrative format: they must, for example, collaborate about

mathematics and language itself. These are the classroom

circumstances where joint work is observed to flounder (Bennett,

1991). As the terms of the task become more complex in this way, so

the social interaction may be observed to disintegrate into parallel

and solitary forms of work (Perlmutter, Behrend, Kuo and Muller,




On resourcing collaborative encounters

The above remarks imply one way of looking at the problems and

possibilities of school-based collaborations. Such a perspective

seeks to understand how schooled tasks create special demands upon

collaborators. Thus, ait would be argued, consideration needs to be

given to how these demands can be resourced. The formation of joint

understandings is a natural and visible achievement of children

functioning in out-of-school contexts: the challenge is to discover

how those achievements can be mobilised for realising purposes defined

in classrooms.

This is not a traditional psychological orientation to the problem of

making collaborative learning happen. More typically, psychologists

might approach this problem in terms of a skill model: so it might be

suggested that school children could benefit from some form of

training in being collaborators (cf. Johnson, Johnson and Roy, 1984).

Alternatively, it might be suggested that collaborative learning is

assimilated to a developmental sequence - arguing it can only occur

following the stage-like emergence of various cognitive structures

(Tomasello et al, in press). I prefer to start from the insight that

children are collaborators from a very young age. Certainly, before

they get into schools they are already highly competent at forming

shared reference - and highly interested in doing so. What is needed

is a greater sensitivity to how continuities can be forged between the

successes of collaborating in domestic and playful contexts and the

demands for doing so in schools. Of course, preschool children's

capabilities for coordinating activity with others is a particular

kind of achievement: it will be dominated by shared reference that is

located in material or narrative structures. So, perhaps the trick

will be to appropriate and extend these achievements into schooled


Computers offer considerable promise in this respect. They can

furnish flexible representations that may become the objects of joint

reference for learners. This capability reflects their interactivity

and their sophistication as a general symbol manipulating technology.

Theorists who adopt a Constructionist (Harel and Papert, 1991)

perspective on computers in education dwell on these properties. Such

theorists are keen to highlight the capacity of computers to make

learning experiences concrete. Yet the orientation of the influential

constructionist tradition has been very much towards the needs of

self-contained learners. So, in a paper discussing what is understood

by making knowledge "concrete" through computers, Wilenski remarks:

'The constructionist paradigm by encouraging the externalization of

knowledge, promotes seeing it as a distinct other with which we can

come into meaningful relationship' (Wilenski, 1991, p.202). In

considering the place of computers within collaboration, I have

stressed their capability for creating such externalised resources.

But the 'meaningful relationships' thereby afforded need not be simply

those between the individual learner and some knowledge domain. They

may also be relationships held in common with others and creative

collaborations may be especially enhanced by that possibility.

In Chapter 7, I described tinkering with a computer program

(one that supported early number work) in order that it might furnish

such a referential anchor for pupils using it together. This

tinkering was successful, in that pupils' activity became increasingly

coordinated around this point of shared reference: they collaborated

more effectively. Developing technology to be supportive of the

collaborative experience of learning is partly about developing

such ways of resourccing joint activity *at* the site of some problem.

But in other chapters, I have stressed that the collaborative

experience extends beyond this famililar paradigm. So it will be

important to understand the assimilation of computers into these

broader engagements. For one thing, this means being alert to the

danger that computer-based activity may fail to get drawn into the

community of discourse that characterises organised learning. The

collaborative interactions of teachers and pupils need to be organised

*in relation to* computers - just as they are to all class-based

learning experience (Chapter 5). In addition, there are other senses

in which the collaborative experience of learning is organised. In

Chapter 8, I considered how communities of learners may collaborate in

a more loosely-coupled way *around* a resource - collaborating

'through the air'. It will be helpful to understand more of how the

technology can be integrated into a learning environment to take best

advantage of opportunities for social exchange. Finally, I have

discussed (Chapter 8) collaborations that may be experienced *through*

this technology.



On the prospects for computers within collaborations

In this book, I have promoted a particular theoretical perspective on

how we might best organise teaching and learning. This has involved

foregrounding the collaborative experience of being a pupil. It has

involved attending to how organised learning invariably is about the

construction and deployment of shared understandings. As participants

in this endeavour, we may proceed more or less effectively, and we may

be resourced to do so more or less imaginatively. My interest in

educational technology is very much for its supporting role in this

sense. I have tried to review something of the exisiting social

structures of educational practice and to suggest some of the ways

that computers may enter into them to support what people are

attempting together. This has involved arguments about the *general*

way in which technology is deployed: not more focussed considerations

of the value of particular computer-based applications - tutoring

systems, simulations, tools, or whatever. My attitude towards any

such particular application is very much determined by how effectively

it is being incorporated into the participatory strucutre of learning

(including consideration of how readily its design might afford such


These concerns are worth rehearsing. There is currently much

enthusiasm for educational software exploiting the new multimedia

presentation capabilities of microcomputers. Certainly, if this book

had attempted to review computers in education in terms of their

technical sophistication and creative ingenuity, then it would have

been necessary to have discussed these applications at some length.

But in terms of the agenda that I have set here, it is not clear that

these new developments are especially promising: it is not clear that

they are going to resource a socially-grounded experience of education

as I have been discussing this. Indeed, accounts of new practice

based upon, for instance, hypermedia tools suggest that these are very

much tools conceived for isolated learners. Landow's (1990) work is

an excetion to this. A recent article by Lehrer (1993) is also

exceptional in evaluating a hypermedia resource in terms of its

relation to the background culture of educational practice into which

it is placed. Perhaps it should be hoped that this signals a new

research trend among developers of this challenging technology.

Another area of fast technical development is networking tools and

connectivity. Yet, as reviewed in Chapter 8, it is not obvious that

the potential of networking is being realised in educational practice.

Universities may be the best settings in which to show what is

possible here. However, the signs that it will offer such a model are

not encouraging. For example, the UK government's support for

developing computer-based teaching resources in higher education has

been quite generous. Yet, none of the many individual projects

financed so far have focused on communication structures, and very few

of them pay any explicit attention to the support of collaborative


I believe the prospects are not particularly encouraging for an IT

development strategy that respects the social themes I have been

considering. If anything, the climate is drifting towards

"delivery-system" models of computers in education. This allows

politically influential commentators to see new electronic media as a

technology for 'capturing the worlds' best brain power' such that it

may 'dramatically increase the rate of circulation of intellectual

capital' (Hague, 1991) This commodity approach to education is

increasingly visible. One recent advertisment for a British

university invites potential students to put it 'at the top of their

shopping list': this institution's concerns being illustrated by a

supermarket basket filled with packages labelled "ecology",

"genetics", "modern art" and so on. Some of the arguments developed

here in Chapter 3 may help us see how computers could be effective in

supporting such delivery of packaged knowledge.

All new technology in education is inherently vulnerable to

deployment in this way. Thus, it is important to remain vigilant to

trends as they are developing. My own advocacy of networks (Chapter

8) should probably have been more sensitive to the potential of this

infrastructure to support the delivery model of education. For

example, there is now a network-based academic journal called the

'On-line Journal of Distance Education'. It has many interesting

articles on Distance teaching. However, its masthead carries a

sobering prediction:

In the industrial age, we go to school, in the information age, school

can come to us. This is the message implicit in the media and

movement of distance education.

This message may be part of the vision shared by many innovators and

many politicians. There may be good grounds for contemplating more

innovative structures of organised education than those that are

currently exemplified in western schooling. But I hope that whatever

alternatives evolve they will respect the need for learners to

participate within rich communities of understanding: to partake of

the collaborative experience of learning.