14th September 1993




The deployment of computers in education is a venture not greatly

influenced by theories of cognition. One reason for this is the

simple fact that psychologists have not made computer-based learning a

topic of special empirical interest. Perhaps that is not surprising,

given the confusingly multi-faceted nature of the technology. There

is nothing uniform about what these instruments "do" for learners,

even though their uniform appearance may suggest that there is.

Practising numerical estimations, using a word processor to compose

stories, programming the movements of a floor robot, exploring an

ecological simulation, the graphic design of a poster: this a very

mixed bag of practical uses for a classroom computer, although it is

typical of what we can witness happening - even within primary


Some of these learning activities seem commonplace enough (say,

estimating); while others appear to be familiar but, on reflection,

turn out to be mediated by computers in quite distinctive ways (eg.,

word processing a story). Yet others, while related to the

established curriculum (say, maths), seem to involve radically novel

approaches to its content (eg., controlling robot movements). It

might be thought unlikely that this mixed bag could be easily embraced

by singular psychological theories. If there is to be any generative

relationship between theory and practice it might seem likely only at

the piecemeal level of inspiring particular software in particular

curriculum areas. Some such examples have already been mentioned:

mathematics learning software has been informed by cognitive

psychology (eg., Resnick and Johnson, 1988; Sleeman, 1987);

computer-based reading and writing aids have been informed by ideas

from research on metacognition and its development (Salomon, 1988b;

Woodruff, Bereiter and Scardamalia, 1982).

These localized influences of psychological theory are certainly

welcome. However, can Psychology furnish any more overarching

theoretical perspective to help us think creatively about the

development of this technology in education? I believe that it can.

Indeed, to an extent, psychological thinking may already be exerting

a concealed influence. I will argue later that certain significant

directions taken by current computer-based educational practice can be

readily legitimized by psychological "worldviews" (Agre, 1993) -

although such influence is not always explicitly identified. This

creates one good reason for discussing certain broad traditions of

psychological theorising here. It will help to articulate some of the

theories of learning and cognition that form a background against

which some recent design and practice has been managed in this area.

Another reason for reviewing psychological theory is more

forward-looking: to find a framework for addressing some of those

problems associated with computer-based practice that were identified

in Chapter 1.

Later (in the next chapter), I shall make a thorough comparison of

three psychological perspectives on learning and cognition and

consider what they each imply for the effective use of new technology.

Two of these perspectives (computational theories of cognition, and

constructivism) are well-developed and well-described in other

sources. The third (socio-cultural theory) is of more recent

influence and still subject to mis-representation. For this reason

(and because it is the approach that I favour), I shall use the

present chapter to describe it more fully. I believe it is the

perspective that best addresses some of the problems of implementing

computer-based learning that I have already identified: particularly

those relating to the social context of educational activity.

In this respect, one issue that theory should help clarify is the

basis for believing that education should preserve a strong

interpersonal dimension. It is significant that practitioners are

worried about new technology on this basis. So, a formal account of

learning as a socially-grounded achievement would inform any challenge

to technological visions of the isolated pupil. Secondly, it will be

valuable to have a theoretical platform for dealing with these

concerns in a concrete, practical manner. In short, an integrating

theoretical perspective could be a powerful resource to help guide

computer-based educational ventures.

Both of these tasks can be addressed by the particular theoretical

perspective outlined below. This perspective is the cornerstone for

the remaining discussion in this book. It establishes learning as a

fundamentally social experience. It encourages the assessment

of new educational resources in terms of their potential for enriching

the interpersonal contexts of learning. This view also suggests a

framework for thinking about real options whereby this social

incorporation might occur - for example, in respect of a resource such

as computers. The theoretical perspective in question is one

associated with "socio-cultural" thinking in Psychology. More

recently, the term "cultural psychology" has been used. It is a

position pitched at a fairly grand level: indeed, it is about the very

nature of cognition.

The term "cultural theory" as applied to cognition usually refers to a

body of ideas inspired by the Soviet socio-historical movement of the

1930s (notably the work of Vygotsky, Luria and Leont'ev). Lately,

that thinking has been enriched by other lines of theorising:

particulalry from within the disciplines of cognitive science (eg.

Suchman, 1987) and anthropology (eg. Lave, 1988). The idea of a

cultural psychology has been most clearly defined by two groups: one

comprises Shweder and his colleagues (Shweder, 1990; Shweder and

Sullivan, 1993); the other comprises various researchers associated

with the Laboratory of Comparative Human Cognition (Cole, 1987; LCHC,

1983, 1986). Their two agendas do not perfectly match (Jahoda, 1992)

but they share a core commitment to the notion of cognition as being

profoundly social in nature. Their perspective on educatonal practice

is, accordingly, one that stresses the socially-organized nature of

the achievement.

In the remainder of the present chapter, I shall concentrate on

capturing the flavour of a cultural psychology. First, I present a

general outline that stresses the central concept of mediational

means; I consider how it relates to cognitive development through the

problematic metaphor of "amplification". Then, I shall identify

central ideas flowing from the application of cultural thinking to

education. A brief qualifying observation is necessary at this point.

My purpose is not to make the following theoretical framework so

convincing that an inevitable agenda for the use of computers in

education will have to be endorsed by the reader. The point is more

to lay the ground for identifying a certain variety of empirical

strategy that is now needed. In particular, I shall recommend

research that clarifies the manner in which this technology mediates

new forms of social interactions among its (educational) users. I

believe the findings of any such research can remain informative

whether or not the theory generating it is judged to be persuasive.

Where scepticism regarding the underlying theory might become a cause

of friction is in relation to how we should best interpret, and

thereby apply, the results of that research.






Many "psychological" perspectives on human nature take biological

themes as their starting point. Cultural psychology is distinguished

by its declared orientation towards the peaks of human achievement -

the practices and artefacts that constitute culture. Of course, this

does not imply a special concern with (high) Culture. In fact, a

reference to usage in biology is quite helpful in clarifying the sense

of this term as intended here.

For biologists, "culture" is the medium in which living material might

be supported. Cultural psychologists orient towards a *medium* for

human activity in this broad sense.  The medium that supports

intelligent human action will comprise artefacts, institutions and

rituals that have acquired their current nature during a long

historical development. This history will be interestingly different

across different communities. The proposal is that any account of

individual cognition and learning must incorporate the nature of this

culture into its conceptual vocabulary. A conceptual vocabulary for

studying cognition should not exclusively refer to structures and

processes concealed within the thinker's skull. It should capture and

express the thinker's interaction with an environment: that is, their

contact with a culture of material and social resources that

everywhere supports cognitive activity.

Given such an orientation, we may anticipate that a cultural theory of

cognition will have a strong contextualist flavour. It will

focus on *situations* for thinking. It will resist suggestions that

the variety of intelligent behaviour can be understood in terms of a

small number of core, cognitive processes. We may also anticipate

that cultural theories of cognition will have a distinctive interest

in the fabric of *socially*-organized life: for social interaction is

surely central to the rich complexity of human culture.

Wertsch opens a recent volume written in this spirit with the

following definition. 'The basic goal of the sociocultural approach to

mind is to create an account of human mental processes that recognizes

the essential relationship between these processes and their cultural,

historical and institutional settings.' (Wertsch, 1991c, p. 6). In

this account, the cognitive attributes of an individual are

fundamentally the outcome of engagement with culture. The analysis

of cognition must invoke a vocabularly that includes reference to

the formats of this engagement - caputing how cultural resources

constrain and enable cognitive activity. The approach invites us to

see the genesis of mental life within our commerce with the products

of a lengthy cultural evolution. Indeed, it is the capacity for

actively exploitating this historical legacy that sets apart humankind

as a species. Cole comments: 'Human psychological functions differ

from the psychological processes of other animals because they are

culturally mediated, historically developing and arise from practical

activity' (Cole, 1990, p. 91). A central concept in understanding

such a perspective is "mediation": a concept now commonly discussed in

relation to the seminal writings of Vygotsky.



The central place of mediational means


Cole's reference to psychological processes in other species echoes a

key point within Vygotsky's development of the concept of

psychological mediation. If we remember that this account was

articulated in the 1920s, the reference towards animal psychology

appears quite natural: much psychological theory of this period arose

from empirical work on animal behaviour. In fact, in early

formulations, Vygotsky seemed particularly concerned to harmonize a

cultural view with the prevailing stimulus-response (S-R) psychology

(Bakhurst, 1990).

Vygotsky draws a distinction between elementary and higher mental

processes. The former define the limits of animal intelligence; they

are biologically based and invite the reductionist analysis favoured

within S-R psychology. They include involuntary processes of

perception, attention, recognition and need. They underpin a basic

repertoire of problem solving behaviour that can be organized in

response to the here-and-now of environmental stimuli. On the other

hand, higher mental functions include all the voluntary and reflective

processes of thinking, remembering and reasoning that we associate

with human mental life. They are not reducible to the elementary

psychological functions. Historically, they arose because human

beings came to turn inward upon their environment - in the sense of

acting creatively upon it to effect certain profound changes in its

relationship to us. It might be said the S-R relation was thus

rendered bidirectional. The resources arising from this creation of

human culture are regarded as central to any account of the nature of


The important sense in which the human subject came to act back upon

nature, and thereby change it, is manifest in the creation of *tools*.

These are at once outcomes of human activity upon the environment

while, at the same time, they serve to organize further and future

encounters with it. Through material tools we gain greater control

over the physical world. This control arises from the mediating

function of these instruments: we act upon the world indirectly, we

act "through" them.

A distinctively human achievement is to have evolved tools realized in

symbolic (rather than purely material) form. Vygotysky, thus,

distinguishes between "technical tools" and "psychological tools".

Historically, it is claimed, varieties of auxiliary stimuli evolved to

have special relevance for controlling the *psychological* world:

notations, diagrams, verbal signals and so forth. The mediation

effected by this class of tools defines the problems that are now in

the domain of cognitive psychology. Through these "signs" (especially

linguistic ones) we come to regulate the behaviour of others. We also

come to exert voluntary control over our own basic psychological

processes: thereby elaborating the activities of remembering,

attention and thought. These artefacts of cultural history are

preserved and made available to each new generation. Thus, they may

serve to support our children's mediated encounters with the social

and material environment. (Indeed, the taking of measures to ensure

this continuity across generations (i.e., instruction) may also be a

uniquely human achievement.)

Vygotsky's initial formulation of this cultural conception of mind was

fairly conservative: it sought to be compatible with orthodox S-R

theory. Perhaps the analogy with technical tools was somewhat

constraining in this respect. Technical tools often have a neatly

circumscribed character; they are visibly self-contained objects

(hammers and so forth). The temptation may be to theorise about

psychological tools that also happen to have this singular character:

icons, maps, verbal instructions and so forth. Such exemplars more

readily take up the role of stimuli. Thus, they might be

conceptualized as "intervening stimuli" located between external

events and behaviour - either in some associationist S-R psychology or

some cognitive theory of human information processing.

Such a simple formulation is apparent in Vygotsky's discussion of

mediated memory and the example of the knot-in-a-handkerchief. This

folk custom is a vivid example of a discrete mediating sign: it allows

organized control of the present by the past (remembering takes place

"through" this device). So a conservative summary of this example

might have the knot-sign function as a class of intervening stimulus:

supporting an association between some past event and some response

that we now make. A more traditional psychological analysis is, thus,


However, as Bakhurst (1990) has documented, this conception was

rejected in Vygotsky's later writing. Vygotsky became dissatisfied

with the implication that signs might be evoked in some S-R manner.

The far-reaching impact of mediation was not well enough expressed by

some catalogue of discrete signs, with their tool-like properties.

Mediational means existed in the form of more complex structural

relations, these having a more elaborate involvement with behaviour.

So Vygotsky became interested in the human capacity for inventing

whole symbolic systems: such as are represented in mathematics,

logical notations and varieties of natural language. What is

significant about engagement with systems such as these is that they

place us in a position of constantly *interpreting* the world, rather

than responding to it. They leave us experiencing the world in

particular ways, reading it in a manner that reflects our own

distinctive history of contact with such systems of mediation.

Some of Vygotsky's own empirical work encouraged this interpretative

theme - by illustrating both a creative and a constraining capacity in

our deployment of signs. Children who were offered pictures to

nmeonically support the recall of a word list used them in

idiosyncratic (but effective) ways. Their use of these signs was not

rigidly related to their physical nature. It is as if the word-item

remembered was actually an interpretative act made possible by the

symbolic device - facilitating the completion of a narrative rather

than eliciting a response. Moreover, adolescents and adults sometimes

seemed less empowered by such devices because, Vygotysky supposed, his

imposed external aids could disrupt their private devices already

internalized for purposes of organizing recall.

In terms of accounting for cognitive development, Vygotsky's changing

theoretical emphasis orients us more to "interpretative practices".

These practices are embodied in the cultural life of a community: the

artefacts, technologies and rituals that it offers. The course of

intellectual growth is, therefore, characterised by gaining access to

a culture's resources of mediational means - as ways of interpreting

experience. In the course of development, children will necessarily

appropriate and deploy whatever local resources constitute their own

opportunities for participating in socially-organized life. They

discover the "designs for living" (Cole, 1990) that have been

historically accumulated within their own culture. These are states

of the world we are born into and, to use an analogy of Bruners, it is

as if we thereby enter onto a stage where the drama and its context is

well established. Our task is to participate in the action and, thus,

to appropriate the mediational devices that can serve to manage

exchanges between ourselves and others.

Three themes arising from an emphasis on mediation


Let us relate these comments about the central place of mediation to

our interests in learning and instruction (leading us, later, to

considerations of new technology). I wish to focus on three

particular implications of the approach being sketched here; each of

them has attracted some empirical support. The first is a novel

definition of cognitive activity in terms of functional systems: a

definition that takes "cognition" to mean more than repertoires of

circumscribed and private mental processes. The second implication is

the "situated" nature of cognitive achievements: what is learned is

ways of acting in particular situations. The third is the profoundly

social nature of cognition. I will briefly summarise what each of

these propositions entails before pursuing their educational

implications in separate sections below.

The first of these points is concerned with how we define cognition

for purposes of analysing development and change. The present

cultural approach is often distinguished by claiming it regards

cognition as a "beyond the skin" phenomenon. A cultural description

of mental activity will typically incorporate reference to mediational

means - resources "outside" of the person, but resources which will be

included in the units of analysis when doing this form of cognitive

psychology. Often such mediational means will comprise artefacts that

reside "outside" in the sense of being clearly visible and external to

ourselves: the maps, diaries, notebooks and filing systems of

intellectual endeavour. Campbell and Olson (1990) propose that such

externally-supported human intelligence describes the most common and

comfortable realization of the activity "thinking". Not that this is

what is captured in popular stereotypes of someone in thought: the

popular image tends to conjure up a solitary, deeply reflective state

typified by Rodin's hunched-up figure. Provocatively, Campbell and

Olson suggest that we naturally take steps to avoid this form of

"inwardly mediated" thinking. Yet, even such contemplative states

need to be analysed with proper respect for the externally-located

mediational means that they will involve: the ways of talking and

symbolising that are appropriated from the thinker's socio-cultural

environment. The solitary thinker's activity is continuous with the

external, socially-constituted environment in this sense. In

summary, this perspective demands that we view cognition in terms of

functional systems of activity integrated by mediational means.

The situated nature of cognitive achievements is the second

implication of an emphasis on mediation. Learning is viewed in terms

of the guided appropriation of mediational means: such change results

in control over the substantive interpretative practices that

characterise a local culture. Remembering, classifying or

thinking are, thus, ways of acting and talking in particular contexts:

contexts drawn from the situations of problem solving provided by that

local culture. So, we become rememberers, classifiers and thinkers.

This is not the same analysis as one highlighting *general* cognitive

resources that underly and support a transfer to new domains of

practice. So, cognitive acquisitions are regarded as initially

situated, in this sense of being tied to contexts of learning.

The social nature of cognition is the third implication of the present

mediational approach. There are actually two senses in which

cognition is being characterised as a social phenomenon. Cognition is

socially located because mediational means are created and evolve

within sociocultural history. The various notations, diagrams,

signals, languages and so on that make up our current systems of

shared signs all embody a history of involvement in human social

interaction: their various contemporary forms will surely reflect this

past. Indeed, such an historically-determined character will serve to

constrain the ways in which they may support our present intellectual

endeavours. In addition to this, cognition is identified as socially

located because these mediational means are commonly encountered in

the course of exchanges among people. This is clearly the case in

early life: children are not left to re-invent mediational means from

scratch. They are confronted with them: this is arranged within the

course of their participation in social life. Thus, when researchers

in this cultural tradition come to consider problems of learning and

cognitive development, they will surely be interested in

problem-solving (in the broadest sense) as it gets coordinated within

arenas of people acting *together*.

I have introduced three implications of a cultural psychologal

approach: cognition as functional systems, as situated and as social.

These ideas are central to my perspective of how computers could best

be deployed within teaching and learning. I shall, therefore, say

more about each below - but stressing their relation to educational

practice. At this point, I will not puruse the link between education

and cultural conceptions of cognition by developing the case of

computer-based learning. Instead, I shall dwell on the more

thoroughly researched case of literacy. Literacy is a technology in

the sense that it involves deploying a symbol system (the written

word) to mediate interactions between ourselves and our material and

social environment. So, we may regard computer-based resources as

more modest parallels to this well-established mediational means.

With this parallel, we may then seek insight from the more extensive

studies of literacy already available.





To capture the force of a mediational analysis of intelligent action,

it is popular to cite illustrations where very vivid prosthetic

resources are involved. A challenge made by Bateson (1972) has been

widely cited to help focus this conception of cognition as something

mediated and extending beyond the skin. Bateson asks:

Suppose I am a blind man, and I use a stick. I go tap, tap, tap.

Where do *I* start? Is my mental system bounded at the handle of the

stick? Is it bounded by my skin? Does it start halfway up the stick?

Does it start at the tip of my stick? (1972, p.459).

Cole's thoughts on this challenge express its general implications for

an analysis of cognition: '..the precise ways in which mind is

distributed depend crucially on the tools through which one interacts

with the world, which in turn have been shaped by one's cultural past

as well as one's current circumstances and goals' (1991, p.412).

Thus, the existence of mediational means invites us to conceptualise

mind as something "distributed" within an environment - rather than as

a repertoire of computational processes constrained to exist only

within our heads.

It is easy to see how the effects of a powerful computer application

(say, a word processor) might be analysed in parallel terms to the

tool in Bateson's example. However, understanding how cognition is

organized and how learning is supported will involve more difficult

analyses than this example might suggest. For one thing, not all

mediational means will be so conveniently circumscribed and concrete.

For another, teaching and learning can not be reduced to initiatives

for merely making new mediational means available to pupils. It is

not as if education was about helping pupils take these resources off

some shelf. That is, the nature of their appropriation will depend

upon the nature of the contexts in which they are encountered - and

to efforts relating to the guided coordiantion of those contexts.

Such claims are elaborated below in three sections. There I consider

issues of teaching and learning in respect of the three themes

introduced above.




(1) Functional systems and mediation: the case of literacy


My aim in this section is to consider how we may best conceptualise

cognitive change as it might occur within the contexts of education.

The cultural approach invites us to analyse this by considering how

mediational means become incorporated within functional systems of

intelligent activity. I will develop the discussion around the

example of literacy, as this is a mediational means that has been most

carefully researched. Thanks to the efforts of Cole and Griffin

(1980), this example also allows us to explore how the achievements of

learning are best conceptualised: in particular, whether to view them

in terms of quantitative processes of cognitive "amplification" or in

the terms favoured here - qualitative changes in functional systems of

cognitive activity.

For the original socio-cultural theorists, spoken language was the

most central of early acquisitions - a view in some contrast to that

of their contemporary, Piaget, for whom language played no powerfully

distinctive role within developmental theory. For theories in the

cultural tradition, speech is seen as an organizer of behaviour.

Speech is instrumental; it helps us to *do* things in the world, to

make things happen. In particular, we learn to effect others through

our speech. It is a small step from this insight to suppose that such

external means of regulating activity becomes, in some sense,

internalized - to become a form of private *self*-regulation (Wertsch,


The concept of internalization is not without its critics. However,

the central feature of this idea remains persuasive: the various

problem-solving experiences of reasoning, remembering, attending,

classifying and so on are - first of all - *activities*. They are

organized within social experience and supported by the resources of

speech. Through participation in social life, the developing

individual is exposed to a set of interpretative practices that may be

appropriated. In a sense then, we have to learn to "become"

rememberers, planners, classifiers and so on (eg., Middleton and

Edwards, 1990). Participation in organized social activity serves to

reveal these powers and possibilities to us. Cognitive achievements

arise as the consequence of entry into particular "communities of

practice". We encounter particular settings where problems get solved

according to specialized practices for the deployment of cultural

resources: resources of discourse, technology, ritual. By virtue of

participation within such communities we become socialized into

possible ways of thinking. Such a perspective on cognition (as

embodied in practice) is in obvious contrast to the dominant

psychological images of, for example, memory or classification as

private cognitive mechanisms.

It naturally follows that the deployment of spoken language within the

various "cultural" contexts of growing up - and the study of its

particular consequences - has been a topic of special empirical

interest to cultural theorists. Speech is the means whereby much gets

done around children and it offers for them particular "ways with

words" (Heath, 1983). It is the means whereby problems are publically

defined and acted upon (Wertsch, McNamee, McLane and Budwig, 1980;

Wertsch, Minick and Arns, 1984; Wood, Wood and Middleton, 1978).

However, there has also been great interest in the developmental

significance of the *written* word. Considering literacy as well as

speech may help further clarify how mediational means support

cognitive development.

Literacy is, evidently, a mediating technology of the kind we have

been considering: it enters into our lives to organize interactions

between ourselves and our material and social worlds. One vivid

perspective on the technology of literacy is furnished by accounts of

its development in *historical* time (eg., Cole, 1991; Goody and Watt,

1968). These accounts make it possible to trace an historical pattern

within which literate practices can be shown to have forced

transformations of human relations on the societal level. The nature

of these historical transformations then offers a seductive analogy

for psychological accounts of development within individual lifetimes.

Perhaps in growing up, our thinking undergoes comparable

transformations as it encounters new mediational means; perhaps such

transformations reflect those documented for whole societies during

periods when they are gaining access to new technologies such as


Luria (1976) reports an early psychological investigation in this

spirit. He studied (during the 1930s) the impact of literacy on

traditional communities within post-revolutionary Soviet society.

That is, he was able to observe the cognitive impact of access to a

radical new mediational means. The sudden drive to develop a literate

population offered an opportunity to evaluate the effects (among

adults) of exposure to reading and writing as it was organized in the

new schools. The inevitably piecemeal nature of the early educational

provision permitted meaningful cognitive comparisons between schooled

and unschooled groups.

Luria reports apparently dramatic effects of even brief exposure to

literacy. These effects were catalogued for intellectual functions in

the domains of perception, classification, reasoning, imagining and so

on. Briefly, the impact of literacy seemed to be associated with a

new capacity to direct thought towards "the words themselves". A

capacity to extricate discursive problems from the immediate context

of a conversational exchange: from the context of expectations and

interpretations that normally guide human discourse. Formal

consideration of words themselves in this way - as the acquisition of

literacy requires - seemed to create for newly-literate individuals a

sensitivity to the hypothetical. Literate individuals become

drawn to reflect on problems that might, sometimes, not actually exist

outside of the (mere) words used to conjure them up.

Developmental psychologists have been impressed by the idea that

access to reading and writing transforms problem solving in this way:

impressed with the idea that writing is a technology with far-reaching

cognitive consequences. For example, Donaldson (1978) appeals to this

possibility in accounting for her influential work on development of

reasoning in childhood. She describes a number of studies revealing

ways in which traditional tests of cognitive development underestimate

young children's reasoning (cf., Cole, Gay, Glick and Sharp, 1971;

Siegel, 1991). She argues that the format of traditional tests are

biased towards reasoning that is most familiar within *literate* forms

of communication.

For example, in the traditional test for conserving number a child

is asked if two lined-up rows of beads each have the same number. One

row is then elongated and the same question is asked a second time.

Answering that the longer row now has more beads has traditionally

been taken to index a structural limitation in childhood reasoning.

However, there may be other factors to take into account;

including some relating to the pragmatics of this situation. Thus,

simply asking the same question twice in such a short period might

lead some children to think they are supposed to change their answer.

It seems that this can happen: children are "more logical" when only

the second question gets asked (Rose and Blank, 1974). The

implication is that children who fail the traditional form of such

tests should not be thought to lack cognitive bits in their logical

equipment. Their problem may involve some lack of familiarity with

the literate form of communication that saturates traditional formats

for testing. Children first mobilise a spontaneous form of reasoning

that reflects their rich experience of thinking in social contexts -

particularly in making sense of other people's actions. Literate

forms of problem demand that children override their expectations of

what the experimenter might be asking of them, override their beliefs

about likely motives and intentions in the situation. Instead, they

are expected to prioritise the language that is being used to define a

problem. This may often be an unfamiliar attitude for children

taking part in psychological tests. To use Donaldson's phrase,

children must cultivate "disembedded" modes of thinking to do well in

these tasks. Their thinking must be disembedded from the matrix of

expectations and interpretations that the context naturally affords

and, instead, submitted to the literal possibilities permitted by the

words themselves.

Donaldson is not alone in believing that schooled contact with the

mediational means of literacy is central to making this happen. This

view has also been championed by Olson and colleagues (eg., Olson,

1986; Olson and Torrance, 1983). Indeed, in the judgement of this

group, much of development in "intelligence" can best be analysed in

terms of variously mastering the literate modes of thinking cultivated

in school.

An important question confronts us at this point. The way we deal

with it has implications for our present concern to conceptualise the

impact of computers - as a further technology that supports access

to new mediational means. The question concerns how we should

conceptualize the *process* underpinning the impact of new mediational

means on individual cognition. The consequences may be clear and

dramatic, but by what mechanism are these outcomes achieved. Two

kinds of response to this problem are evident in the literature. One

is inclined to view the cognitive impact of cultural experience in

terms of acquiring and refining tools of thought (a literate mode of

thought, for example). This is a more cognitive kind of account, one

that toys with the notion of individuals internalizing such

technologies. The alternative is more practice-oriented and regards

the impact of access to new mediational means in terms of a

re-organization of some underlying way of acting in the world. The

distinction is slippery but I believe our attitude to it bears upon

how we understand the educational impact of new computing technologies

with their tool-like properties.

The more cognitive alternative is implicit in Bruner's arresting

metaphor of the cognitive "amplifier" - to describe the effect of

contact with some mediational means. He comments: 'Man is seen to

grow by the process of internalizing the ways of acting, imagining,

and symbolizing that "exist" in his culture, ways that amplify his

powers' (Bruner, 1966, p.320). The parallel is with more familiar

technologies that are said to amplify action - hammers, levers, knives

and so on - except that the symbolic equivalents are "internalized"

(and then deployed to support continued cognitive development). On

this model, cultures might furnish a (varying) supply of basic

cognitive resources (amplifiers) in the form of psychological tools

and symbol systems. Cultures might thereby extend cognitive

capabilities to varying degrees. This collection of mediational

means might now include computer-based resources. However, the

amplifier parallel may need a little more exploration. Cultures are

certainly forthcoming with mediational means, but is an amplifier the

best way of expressing the process of empowerment that follows from

accessing them?

An amplifier is a harmless enough metaphor if it merely implies that

access to a cultural technology can multiply our intellectual

achievements. The products of human activity may indeed be amplified

in this sense and this will be visible in cultures with a rich and

varied supply of technologies. If that is all we intend by it, then

it is both harmless and not very useful. Cole and Griffin (1980) have

analysed the amplification metaphor further and suggest that, often,

we do mean more by it. Furthermore, what we mean in addition might

deserve careful review. This is of particular interest to us here as

the notion of amplification is widely appealed to in discussions of

the cognitive effects of using computers.

The temptation of the amplifier image is to encourage a more-or-less,

or quantitative attitude to the impacts associated with new

mediational means. Yet what may actually be needed is a model of

cognitive processes that emphasises structural change, rather than

quantitative change. This structural analysis would encourage

thinking in terms of functional systems of interrelated components -

rather than singular (amplified) mechanisms. So, in respect of some

cognitive function (such as memory), rather than think of cultural

resources as producing a more powerful mechanism, we would think of a

re-organization effected in the activities that comprise remembering.

Cole and Griffin invite us to capture the controversy here by

reference to the following example. Consider *killing* as a

functional system of activity: one that hunters engage in for the

capture of their prey. The "killing power" of a hunter can be

extended if we supply him with a gun. Thereby, the products of the

killing activity are increased. But that amplification only occurs

when the tool is in his hand. So, the effect of the weapon is best

described in terms of its *re-organizing* the activity of killing, not

in terms of it extending some underlying and general-purpose killing

power. The same analysis can be applied to more culturally familiar

activity systems (perhaps with more vivid cognitive contents).

Consider, for example, shopping. Setting out to purchase a new supply

of goods for the family will be a different kind of activity if we do

so equipped with a pen and paper. We thereby exploit the device of a

list; it will probably help us do this task more efficiently (more

quickly, more thorougly etc.). The incorporation of this mediational

means into the activity system serves, again, to re-configure the

manner in which it is carried out. In this case, the underpinning

activity of remembering has been re-mediated.

Cole and Griffin also refer to the example of memory to express their

point about amplification. They note that, in a test, a child with a

pencil displays a more powerful memory than an undergraduate without

one. It might be said that the child's "memory power" has been

extended. The pencil is a sort of amplifier perhaps: the products of

remembering are increased through its use. This seems a

straightforward claim, but suppose we take the pencil away? Where

does this leave the child's memory in relation to the undergraduate?

The effect of the pencil is to re-organize things we do in relation to

the task of memorizing. This task calls upon a functional system, not

a dimensionalized cognitive power. Cognitive amplifiers may largely

act through re-organizing underlying *activities* (such as might be

involved in remembering) - not by amplifying cognitive powers in some

general-purpose manner that exists as "residue" when the mediational

means are not to hand.

This last point discourages thinking about the impacts of new

mediational means in terms of very *general-purpose* changes in ways

of thinking. The reorganizations effected by access to cultural

resources may be powerful but quite localized, or situated, in their

impacts. Both of these points - cognitive development as functional

reorganizations, and the situated nature of the achievements - are

empirically pursued in cross cultural work by Cole and Scribner


Cole and Scribner studied literacy and its consequences among the

peoples of Liberia. This setting offered a distinctive opportunity

for evaluating claims that access to literacy leads to powerful and

general cognitive changes. In Liberia, several different forms of

written language existed serving different communities and different

cultural purposes. These scripts included Arabic, English and Vai.

For some communities the use of a script was principally associated

with a particular form of cultural activity - religious recitation,

business transaction, schooled instruction, the writing of letters and

so on. With literate and non-literate members of these communities,

Cole and Scribner conducted a series of tests of a kind familiar to

cognitive psychologists: tests concerned with memory, attention,

classification and other traditional cognitive functions. They found

no evidence that exposure to literacy itself created across-the-board

cognitive advantage. Rather the effect of literacy was more

localized. For example, familiarity with the Vai script for purposes

of (postal) communication might confer an advantage on tests of

referential communication skill.

This conclusion is in tension with that proposed by Luria to account

for effects of literacy among the people of Uzbekistan. Cole and

Scribner argue against conceptualizing the impact of literacy in terms

of general and quantitative extensions of cognition - such as might

then be labelled more "rational" or more "theoretical" modes of

thought. Rather, literacy is conceptualized as a technology that

restructures the manner in which we carry out certain cognitive

activities; such as those to do with recalling, classifying, ordering

and communicating. To understand the action of such technologies on

development it is then necessary to study "literate practices".

Writing enters into particular forms of culturally-organized activity

in distinctive ways to regulate interactions among the participants.

The consequence of becoming literate is therefore visible in

situations reproducing particular core activities that literate

cultures typically porvide. So, for example, if a culture's literacy

is mainly for supporting letter writing, then it will most likely

promote a certain sort of cognitive reflection; for example,

reflection about how to specify meaning and intention under

circumstances of limited communicative context. Experience in such

situations will have cultivated practices that are then manifest, for

example, in formal psychological tests of referential communication.

I have pursued the example of literacy in order to illustrate the

general approach that cultural psychology takes towards issues of

cognitive change as it is explored in comparative study: historical,

cross-cultural or developmental. The analysis stresses how we

variously come to think "through" mediational means. Goody, Luria,

Donaldson, Olson and others show how cognition is extended by access

to the particular technology of literate forms. Cole and Scribner

caution against too readily interpreting such mediation in terms of a

general amplification of the way in which we process information.

They observe that the mediational role of the written word may be

associated with circumscribed literate *practices*. It thereby

supports only bounded sets of human activities. (Of course, where

these literate practices are those of "schooled reasoning" then that

bounded set will certainly be a highly prized one for many

technological societies.) In the terms used earlier: the child

becomes socialized into particular traditions of interpretative

practice involving reading and writing. Thus, the key to

understanding the impact of literacy during development will be to

study how the written word enters into children's activity settings -

organizing those settings in distinctive ways.

I believe that we can also understand the cognitive impact of access

to computing technology according to the same agenda being presented

here for literacy. This is an important implication of the present

discussion. It follows from this discussion that computers might be

regarded as entering into certain problem solving enterprises and

achieving their impacts by reorganising or re-mediating the activities

involved. In the end, this is an analysis concerned with

conceptualising cognitive changes associated with learning through new

technology. However, it is not a traditional cognitive analysis. It

is one that dwells upon changes to the structure of activity systems

that a pupil participates in - rather than changes in a pupil's covert

knowledge structures.

Such a view does have important implications for how we think about

ways of using new technology in support of learning. If the

experience of a computer-based cognitive task is conceptualised as

acting to effect some abstract, private cognitive structure (cognitive

"tool" or whatever), then the broader *context* of that experience may

seem less significant. On the other hand, the present mediational

view highlights this context. Pupils should encounter computers as

mediational resources incorporated within suitably rich settings of

activity. That is, settings with authentic goals and purposes for

those pupils, and settings that are explicitly integrated with other

experiences of knowing and understanding as they get organized at

other times. The point about actively seeking integration is one that

I shall return to: it strongly implicates a role for social (teacher)

intervention in support of its achievement. The other point - that

activity settings should be authentic - is one strongly argued by

Brown, Collins and Duguid (1989) in their analysis learning as a

'situated' achievement. This takes us to the second broad theme to

arise from a sociocultural analysis.




(2) The situated nature of cognition


Within the psychological literature, there are two contexts within

which the term "situated" gets used to describe cognition. The first

entails an orientation towards the *outcomes* of cognitive activity -

what is learned. In particular, it concerns a long-standing issue of

how far learning in one situation can be expected readily to influence

the activities of the learner within other, different situations.

This is the issue of transfer of learning. To say that a cognitive

achievement is "situated" is to draw attention to limited

possibilities of transfer: the effects of the achievement are

constrained to its context of acquisition (at least, in the first

instance). Psychologists associated with the socio-cultural tradition

are particularly concerned with this issue.

The second sense of "situated" arises more commonly within cognitive

science. In that discipline there has recently emerged a variety of

theorising distinguished by its situated perspective on cognition.

This perspective has implications for the question of transfer but,

first, it concerns definitions of knowledge itself. It is argued

that knowledge should not be conceptualised as a catalogue of stored

mental representations; instead, knowledge is always created within

the circumstances of interacting with the world - in other words, it

is situated within these interactions. Here, I shall comment first on

the transfer issue as commonly encountered in socio-cultural

theorising and then turn to make a few brief remarks about "situated

cognition" as more typically encountered within cognitive science.

Claims about the generality of learning made within the former

tradition may be made more substantial by theoretical conceptions

developed in the latter. Thus, the two perspectives on cognition as

situated are closely related.

Three groups of empirical observation have encouraged a view of

cognition as tied to particular contexts of acquisition (rather than

general and context-free). The first is research showing that, at a

given point in development, children's thinking may manifest logical

characteristics in some domains while the same characteristics do not

get mobilized in others. Donaldson (1978) summarises some examples of

how the quality of children's reasoning can vary according to the

format of the problem itself. The second is a comparable tradition of

comparative research involving different cultures (eg. Cole et al,

1971): utilization of cognitive resources may vary across cultural

settings according to local familiarity with the terms of the problem.

The third set of relevant empirical observations are laboratory

studies that demonstrate how difficult it can be for experimental

subjects spontaneously to transfer strategic thinking from one problem

(where it has worked) to a new and related problem (Detterman, 1993).

Together these lead to a particular conception of intelligence or

"expertise". Shweder (1990) summarizes this in his outline of

cultural psychological principles:

..what seems to differentiate an expert from a novice (chess player,

abacus user, medical diagnostician, etc.) is not some greater amount

of content-free pure logical or psychological power. What experts

possess that neophytes lack is a greater quantity and quality of

domain-specific knowledge of stimulus properties, as well as dedicated

mastery of the specialized or parochial "tools" of a trade (p. 23).

Vygotsky's stand on this issue was clear; in his words: 'the mind is

not a complex network of [general] capabilities, but a set of specific the acquisition of many specialized

abilities for thinking' (Vygotsky, 1978, p. 83).

Of course this bias towards viewing new acquisitions as being tied to

contexts does bring its own problems. The consequences of a strongly

situated view of cognitive achievements have been considered by Jahoda

(1980). Something like an integrating theory of situations is needed

in order to avoid multiplying accounts of achievements restricted to

their contexts of acquisition. Moreover, the unavoidable fact

remains that things mastered in one domain can be found to serve us

well elsewhere: generalisation of understandings is something we

surely feel does occur.

I will take up this challenge in a later chapter. I agree with

theorists who argue that it is best to start from situations 'and

discover the sources of generality in what are initially context

specific achievements' (Cole, 1990, p.16). However, I shall argue

that what is "in" achievements that affords generalisation is

something that is invariably put there by the social environment. In

other words, this is an account of learning that views new

acquisitions as initially situated, but which recognizes the

possibilities of transfer. Such possibility arises through supportive

interventions of a sociocultural nature.

Evidently, this position carries implications for how we organize

experiences in educational contexts. It suggests that things could

get done and said around computer-based learning (or other settings of

mediated learning) that serve to support the transfer of

understandings between situations. This is a straightforward sense in

which we must come to define a social context for computers deployed

in educational settings. It is part of what we might mean by claiming

a social context for cognition - although this rather specific

argument relating *transfer* to social experience is not the feature

most commonly identified within that claim. In the following section

I discuss the more traditional version of the claim.

What has been said so far in the present section reflects perspectives

typical of socio-cultural theorising. I shall conclude with some

remarks inspired by the second theoretical framework in which

cognition is currently said to be "situated". In this case, the

claims that emerge sit comfortably with cultural thinking about

cognition, but their history owes more to debates within cognitive

science - particularly in relation to the issue of artificial


We may begin by declaring what is not controversial. Human cognition

implicates an agent with some underlying neural organization, and it

implicates an environment from which sensory stimulation arises

and towards which action is directed. The task for cognitive

psychology is to attend to the way in which agent and environment

interact. The intelligent action that, thereby, can result needs to

be described and explained within a suitable theoretical vocabulary.

Controversy surfaces at this point. Traditionally, the vocabulary

preferred by psychologists has invoked a layer of cognitive concepts.

These concepts refer to structures that are somehow instantiated in

the neural organization. They arise as a result of interactions with

the environment; they serve to direct such interactions. In short,

cognition is said to entail stored, mental representations of the

world (knowledge), and mental manipulations performed upon those

representations (thinking). Intelligent action is, thus, driven by

the output from such an underlying mental life.

All of this relates to the development of educational practice,

including what may be attempted with the help of new technology. It

is relevant because educational interventions may be conceived and

evaluated by reference to this framework of cognitive concepts. So,

the issues being identified here will resurface in later chapters:

specificaly, the implications of this cognitive theorising for

practical applications of computers in learning will be discussed in

the next chapter. For the moment, I will just note that cognitive

theories of the sort sketched above will tend to promote certain

ventures at the boundaries of cognitive science and computer science.

In particular, the idea of knowledge as stored representations

promotes the venture of reproducing such a knowledge "database" in

machine form. This would be an attempt, perhaps, to reproduce human

expertise. This cognitive theory might also imply a particular model

of human communication: one in which computers could be programmed to

transfer data in such a way as to simulate communicative processes.

Any such simulation would be of special interest to us here - insofar

as it addresses that special form of communication known as


However, the credibility of these symbolic theories of cognition has

been questioned - along with their implications for computer-based

education (eg. Suchman, 1987; Winograd and Flores, 1986). This

questioning includes the promotion of an alternative conception. The

alternative is sometimes identifed as a "situated" theory of

cognition, a term favoured within the cognitive science community. The

situated view makes a commitment to the idea that knowledge is created

within interaction: it does not exist "behind" that interaction as

mental events that drive it. Thus, the situated approach is

successful in developing the cultural theorist's interest in the

distributed, mediated nature of cognition. The empirical strategy of

the approach is to study cognitive agents in interaction with their

environment - the various contexts of material and social resources

that mediate action.

This alternative to traditional cognitive theorising is not as

subversive as my contrast might imply. Proponents of a situated view

are anxious to stress that cognitive modelling remains a useful

resource. Clancey (1991) characterises the situation this way.

Cognitive psychology has furnished a description of a covert cognitive

space. Essentially, this has involved looking at the products of

human rational behaviour (language, rituals, strategies etc.),

discovering "patterns" therein and then (here is the suspect move)

locating such patterns inside our heads - supposing this mental world

comprises a mechanism that drives the rational behaviour. Clancey

argues that this approach has been useful; but its value is more to

define an agenda - something to be explained - rather than as, itself,

an achievement of explanation. 'Pattern descriptions now serve as a

specfication for how adapted behavior must appear, rather than the

mechanism to be put inside the robot' (1991, p.111).

If the traditional approaches only go this far, what must be

done to construct a more explanatory account of rational behaviour?

It is proposed that knowledge must be conceptualised as an activity,

rather than as a (stored) property of the individual. *Knowing* (in

preference to (*knowledge*) is activity always exercised in relation

to the situations individuals find themselves in. Knowing is a

relationship between the human agent and a material and social

framework that defines the momentary circumstances for acting.

*Learning*, thus, becomes an adaption of the learning person to

aspects of such circumstances, as they encounter them. This has

implications for what happens in settings that we arrange explicitly

to promote learning. What happens to learners in these settings needs

to be expressed in terms that capture a dialectic: in terms that

include features of the environment as they entered into some

interaction that occured. At later times, when learners might be said

to "remember" things, their achievements would be expressed as the

reorganization of earlier ways of perceiving and acting. New actions

(including purely contemplative intelligence (Greeno et al, 1991)) are

coupled to past learning in this sense - rather than through the

mediating intervention of stored symbolic cognitions.

Such theorists reveal an affinity with those sensory psychologists

such as Gibson who have specifically considered the integration of

perception and action. Such ecological theories address, for example,

how we account for an animal's skillful dash through dense terrain.

The account would not be in terms of the animal activating some

underlying cognitive plan that triggers a complex sequence of action,

but in terms of invariant features of the physical environment that

"afford" certain behaviours at the given moment. Greeno et al (1991)

have studied human learning with special attention to the manner in

which a material environment affords problem solving actions in this

sense. However, the thrust of empirical work influenced by the

situated tradition has concentrated on the management of human action

within the *social* "terrain". Here, the opportunity has been

taken to apply techniques from conversation analysis (Goodwin and

Heritage, 1990) to situations where discourse is central to the

learning or problem solving under examination (eg., Suchman, 1987;

Roschelle, 1992). This bias towards studying situations where social

interaction predominates is typical of theorising influenced by the

socio-cultural tradition. It is this theme within that tradition

that I turn to next.




(3) The social nature of cognition


As I stressed earlier in this chapter, there are really two senses in

which the cultural perspective insists that cognition is fundamentally

social in nature. First, it is claimed that all higher mental

functions are entrenched in a framework of rituals, conventions,

technologies and institutional practices: this framework arose in

sociocultural history. Even the most private of cognitive pursuits

will involve us with media and symbol systems that have a social

nature in this sense. Moreover, some are unambiguously social

by virtue of being encountered through the behaviour of others:

particular ways of talking and acting. Second, cognition is social

because the *acquisition* of new understandings is made possible

through participation in certain kinds of supportive social


The influential writing of Vygotsky concerns both of these themes

(Valsinaar and Winnegar, 1992). However, the emphasis of Vygotsky's

own empirical work (and, largely, that of his followers also) was on

the second of them. Thus, many commentators have been led to reflect

only upon the social *interactional* basis of cognition. For example,

in a sympathetic but fairly critical review, Schaffer (1992) appears

to be evaluating present claims regarding the social constitution of

cognition. However, the empirical work cited is exclusively concerned

with cognitive outcomes arising from experience in joint problem

solving (bearing, therefore, only on "social" in the second -

interactional - sense above). So, a perspective does get usefully

reviewed in this exercise but it forms only part of the claim that

cognition is socially constituted.

Some cultural psychologists have been at pains to counter too narrow a

conception of the "social" as it relates to cognition; arguing that

interaction among people does not exhaust the proper sense of social

involvement. Thus, Scribner (1990) suggests that the contemporary

emphasis on interpersonal issues has distracted us from investigating

sociocultural mediation in a fuller sense. A comprehensive empirical

agenda must embrace the influence of interpersonal interactions

("social" themes) as well as the influence of artefacts, technologies

and conventions ("societal" themes, perhaps). In focussing too much

on the former only, researchers have neglected to pursue, for example,

'how cultural communities this world over organize activity settings

for the "social transfer of cognition"' (p.93). Recent literature

indicates that cultural theorists are now turning their attention more

in the "societal" direction (eg., Wertsch, 1991a).

I shall return to such matters in the next chapter, when focussing

more closely on computer-based learning in relation to cultural

thinking. The topic will arise there because the institutions and

practices of formal education do illustrate a societal theme very

well. They illustrate organized activity settings of a kind that our

culture has indeed fostered - for the particular purposes of promoting

the social transfer of knowledge. New technology is an intriguing new

component of such settings. However, I shall suggest that when we do

consider this societal theme, we are still required to attend to

issues of social *interaction*. This is because cognitive change

within educational activity settings may depend upon certain kinds of

coordination achieved for us by the efforts of other people in these

contexts. For this reason, I shall turn next to say a little more

about the typical analysis of instructional interaction that is

associated with cultural psychology.

That central place of social interaction in cultural theory is most

clearly expressed in the form of one key concept - Vygotsky's zone of

proximal development (ZPD). It was conceived to deal with two

educational issues. Firstly, the issue of how one might

satisfactorily assess a child's level of understanding in some

domain. Thus, it addresses the problem of testing. Secondly, it

deals with what goes on during processes of instruction. Thus, it is

about how learning is organized between people.

The relation of the ZPD to issues of testing arises from Vygotsky's

attention to the gap existing between 'actual developmental level as

determined by individual problem solving' and 'potential development

as determined through problem solving under adult guidance or in

collaboration with more capable peers' (Vygotsky, 1978, p. 86). This

leads to an appealing conceptualization of assessment that focuses on

*potential* to learn and on teachability.

The relation of ZPD to issues of instruction arises from what is said

about the nature of productive collaboration as it might be best

organized with adults or "more capable peers". Just how to define

effective interpersonal exchanges within this "zone" has concerned

cultural theorists rather more than the complementary question of how

it might serve assessment purposes. Some conceptions of interaction

within the ZPD will be outlined here, although discussed in more

in Chapter 4.

If we believed that instruction involved only the efficiency of

Intitiation-Response-Evaluation exchanges (of the kind described in

Chapter 1), then the idea of a teaching *machine* would have some

credibility. However, this conception of instruction - one focussed

on the direct transmission and confirmation of information - misses

the rich possibilities of social interaction organized between

individuals of varying expertise. Truly productive encounters between

them will depend on something more subtle than the didactic exchange.

Theorists developing the ZPD concept invite us to view instructional

exchanges more in terms of collaborations.

A popular metaphor to capture what a collaboration might involve

within instructional settings is that of the "scaffold" (Wood, Bruner

and Ross, 1976). To make this work, we assume that the learner is

oriented towards a goal (the completed structure, in our metaphor);

the goal would not be attainable without external aids and support;

the expert's presence serves to ensure such support and thereby

creates an occasion of collaboration. Such encounters do not entail

simple demonstration or direct explanation: they require more

participation on the part of the novice and more sensitivity on the

part of the expert. The encounter is a collaborative one requiring

jointly coordinated problem solving. This image of scaffolding is

helpful but, as a number of commentators have suggested (eg. Newman,

Griffin and Cole, 1989; Wertsch and Stone, 1984) the metaphor should

not be pursued to slavishly. For one thing, it's static and rigid

connotations fail to suggest a real dynamic to activity as it is

jointly organized in this zone of interaction.

A critical commentator will rightly seek fuller definition of this

dynamic: exactly how does the expert's presence in the zone of

interaction serve to create cognitive support? I shall say a little

more in Chapter 4 about the detail of what could actually go on

between participants in this zone. Suffice to say here that I believe

the active creation of socially-shared understandings (between expert

and novice) is an important investment within such instructional

interactions. Tutorial initiatives will often need to build upon a

mutual foundation of that kind. Then, the sense in which such

interventions may become useful - have lasting impacts on

understanding - might be pursued in terms of a further important

concept associated with Vygotsky's account of this zone: the notion of


Vygotsky proposes that all cognitive functions are first experienced

on the *inter*mental plane before they exist on the *intra*mental

plane. That is, our private mental reflections arise from experiences

that have first been organized in the public forum of social

interaction. A much-cited passage from Vygotsky's writing expresses

this well:

*An interpersonal process is transformed into an intrapersonal one*.

Every function in the child's cultural development appears twice:

first, on the social level, and later, on the individual level; first

*between* people (*interpsychological*) and then *inside* the child

(*intrapsychological*)....All the higher functions originate as actual

relationships between human individuals. [Italics in original]

(Vygotsky, 1978, p. 57).

Thus, we are offered a parallel between the external world of jointly

managed problem solving and the internal world of mental functioning.

Cognitive psychology might become the study of an interplay between

these two. In Vygotsky's analysis, a process of "internalization" is

conjectured to allow the social experiences of one to be realized

within the privacy of the other. So, Vygotsky's special interest in

language arises from his regarding it as the mediational means common

to both the inter- and intra- individual world of intelligence.

This account is not without problems. In particular, some critics

have complained that the internalization concept is underspecified.

Even so, this renewed emphasis on learning through the dynamics of

social interaction has proved immensely influential within

contemporary thinking about cognitive development and educational

practice. It is not an easy framework to evaluate empirically

(Schaffer, 1992). However, the evaluation strategy typically

preferred by cognitive researchers - poorly contextualized, short term

studies of outcomes from joint problem solving - is a strategy not

well matched to the scope of the claims. Yet, I believe this kind of

theorising does fit well the experience of practitioners and it fits

well ethnographic descriptions of teaching-in-progress.

Deploying this conceptual scheme here for concrete discussion of

computer-based learning may help to make these claims fully









In this chapter I have outlined one agenda - taken from contemporary

Psychology - for the analysis of cognition, cognitive development and

educational practice. This is the socio-cultural perspective; or that

perspective roughly corresponding to what is now termed "Cultural

Psychology". I argued that the central concern of this approach

was to understand how new mediational means enter into human

behaviour in order to recoordinate it.

Several commitments are entailed by adopting such a theoretical

attitude. Firstly, accounts of intelligent action must now go beyond

the narrow, mental-process vocabulary of traditional cognitive

psychology. Cultural accounts will want to incorporate reference to

the role of mediating technologies as they enter into functional

systems of behaviour. These mediational means will include structural

features of the environment, artefacts, institutionalized

relationships, symbol systems and (most powerful of all) 'ways with

words'. Secondly, the appropriation and elaboration of new

interpretative practices is a situated achievement: it is not best

analysed in terms of the acquisition of generalised cognitive tools or

representations. There are still issues of learning transfer and

flexibility to be addressed - they are central to our interest in

educational practice - but such issues might best be understood in

terms of supportive interventions organized by the socio-cultural

environment. At least, I shall argue along these lines later.

Finally, this cultural approach converges upon a socially-grounded

conception of cognition. Mediational means may be appropriated during

the short spans of individual lifetimes, but they are themselves

resources fashioned over very long periods of cultural history. Most

important, their history reflects their lengthy involvment in human

affairs and this constrains how we may relate to them now. Cognition

is also social in nature because so many of the specific

interpretative practices we encounter during development are made

available to us within interpersonal communication. It follows that

the settings of formal education will be of special interest to

socio-cultural theorists. For it is here that practices of

communication have become especially crafted: refined and concentrated

for the explicit purposes of conveying interpretative practices to

others in the culture. The thesis of this book is that such

traditions of educational "collaboration" should be carefully

evaluated when we contemplate the incoporporation of powerful new

information technologies.

There are, of course, other theoretical traditions addressing problems

of cognition and cognitive development. There are other

traditions, therefore, that might inform the deployment of a new

educational technology. Those that are most influential within

contemporary psychology tend not to put such strong emphasis upon

social processes. By way of acknowledging this, and trying to learn

from it, I shall review these alternatives in the following chapter:

making some contrasts between the approach outlined here and that

associated with two other significant theoretical traditions. I have

chosen these two for their central importance in current psychological

thinking, but also because they both have been influential in guiding

applications of technology to education.