[com-prac] Chapter 4 of The Social Life of Information by Brown and Duguid
- This is a good read, with some good stories, etc. We might want to discuss
the book on this list. (I'm going to run out and buy it immediately!)
--* John D. Smith |(503) 963-8229 | 2025 SE Elliott Ave., Portland OR
--* Politics of Learning | http://www.teleport.com/~smithjd/
--* People tend to overestimate what can be done in one year
--* and underestimate what can be done in five or ten years
From: rre@... [mailto:rre@...] On
Behalf Of Phil Agre
Sent: Thursday, February 10, 2000 6:29 AM
To: Red Rock Eater News Service
Subject: [RRE]The Social Life of Information
[I have heavily reformatted this; the italics and footnotes are missing.]
This message was forwarded through the Red Rock Eater News Service (RRE).
Send any replies to the original author, listed in the From: field below.
You are welcome to send the message along to others but please do not use
the "redirect" option. For information about RRE, including instructions
for (un)subscribing, see http://dlis.gseis.ucla.edu/people/pagre/rre.html
or send a message to requests@... with Subject: info rre
Date: Wed, 9 Feb 2000 16:07:39 -0500
From: "McConville, Sarah" <smcconville@...>
The Social Life of Information
John Seely Brown and Paul Duguid
Harvard Business School Press, February 2000.
Introduction: Tunneling Ahead 1
1 Limits to Information 11
2 Agents and Angels 35
3 Home Alone 63
4 Practice Makes Process 91
5 Learning -- in Theory and in Practice 117
6 Innovating Organization, Husbanding Knowledge 147
7 Reading the Background 173
8 Re-education 207
Afterword: Beyond Information 243
About the Authors 319
Chapter 5: Learning -- in Theory and in Practice
Knowledge management is the use of technology to make information
relevant and accessible wherever that information may reside.
To do this effectively requires the appropriate application of the
appropriate technology for the appropriate situation.
Knowledge management incorporates systematic processes of finding,
selecting, organizing, and presenting information in a way that
improves an employee's comprehension and use of business assets.
We began the last chapter contemplating the trend from business
process reengineering to knowledge management. There, we focused
primarily on the limits of process, which we suggested was an
infofriendly concept, but one that might be blind to other issues.
In this chapter, we take up the other half of the matter and consider
knowledge and learning, again in relation to practice and again
as distinct from information. We do this with some trepidation.
On the one hand, epistemology, the theory of knowledge, has formed
the centerpiece of heavyweight philosophical arguments for millennia.
On the other, knowledge management has many aspects of another
lightweight fad. That enemy of lightweights, The Economist, has
pronounced it no more than a buzzword. We may then, be trying to
lift a gun too heavy to handle to aim at a target too insubstantial
Certainly much about knowledge's recent rise to prominence has
the appearance of faddishness and evangelism. Look in much of the
management literature of the late 1990s and you could easily believe
that faltering business plans need only embrace knowledge to be saved.
While it's often hard to tell what this embracing involves, buying
more information technology seems a key indulgence.
Nonetheless, people are clearly taking up the idea of knowledge in
one way or other. From within organizations come sounds of fighting
between the IT (information technology) and HR (human resources)
factions over who "owns" knowledge management. Similarly, technology
giants have entered a propaganda war over who best understands
knowledge. Elsewhere, the management consultants are maneuvering for
high ground in the knowledge stakes.
In the process, knowledge has gained sufficient momentum to push aside
not only concepts like reengineering but also information, whose rule
had previously looked so secure. To be, in Peter Drucker's term, a
"knowledge worker" now seems much more respectable than being a mere
"information worker," though for a while the latter seemed very much
the thing to be. Similarly, pundits are pushing "information economy"
and the venerable "information age" aside in the name of the more
voguish "knowledge economy" and "knowledge age". There's even a bit
of alternative prefixation in such terms as knobot, which we talked
about in chapter 2, where the buzz of bots and the buzz of knowledge
Beyond its buzz, however, is there any bite to these uses of
knowledge? When people talk about knowledge, are they just clinging
to fashion (as many no doubt are), or might some be feeling their way,
however intuitively, toward something that all the talk of information
or of process lacks? Is there, we begin by asking, something that
knowledge catches, but that information does not?
Twenty-five hundred years of unresolved epistemological debate from
the Sophists to the present argue that we would be unwise to seek
the difference by pouring over rigorous definitions. Moreover,
whatever differences abstract definitions might clarify, persuasive
redefinition now obscures. People are increasingly eager that their
perfectly respectable cache of information be given the cachet of
knowledge. Such redefinitions surreptitiously extend the overlapping
area where knowledge and information appear as interchangeable terms.
Nevertheless, if we check the language of knowledge management at
the door, there do appear to be some generally accepted distinctions
between knowledge and information. Three strike us as particularly
First, knowledge usually entails a knower. That is, where people
treat information as independent and more-or-less self-sufficient,
they seem more inclined to associate knowledge with someone. In
general, it sounds right to ask, "Where is that information?" but odd
to ask, "Where's that knowledge?" as if knowledge normally lay around
waiting to be picked up. It seems more reasonable to ask, "Who knows
Second, given this personal attachment, knowledge appears harder to
detach than information. People treat information as a self-contained
substance. It is something that people pick up, possess, pass around,
put in a database, lose, find, write down, accumulate, count, compare,
and so forth. Knowledge, by contrast, doesn't take as kindly to ideas
of shipping, receiving, and quantification. It is hard to pick up and
hard to transfer. You might expect, for example, someone to send you
or point you to the information they have, but not to the knowledge
Third, one reason knowledge may be so hard to give and receive is that
knowledge seems to require more by way of assimilation. Knowledge
is something we digest rather than merely hold. It entails the
knower's understanding and some degree of commitment. Thus while one
person often has conflicting information, he or she will not usually
have conflicting knowledge. And while it seems quite reasonable
to say, "I've got the information, but I don't understand it,"
it seems less reasonable to say, "I know, but I don't understand,
or "I have the knowledge, but I can't see what it means". (Indeed,
while conventional uses of information don't necessarily coincide with
the specialist uses, as we noted earlier, "information theory" holds
information to be independent of meaning.)
WHERE IS THE KNOWER LOST IN THE INFORMATION?
Knowledge's personal attributes suggest that the shift toward
knowledge may (or should) represent a shift toward people.
Focusing on process, as we argued, draws attention away from people,
concentrating instead on disembodied processes and the information
that drives them. Focusing on knowledge, by contrast, turns
attention toward knowers. Increasingly, as the abundance of
information overwhelms us all, we need not simply more information,
but people to assimilate, understand, and make sense of it.
The markets of the knowledge economy suggest that this shift is
already underway. Investment is no longer drawn, as postindustrial
champions like to point out, to bricks and mortar and other forms of
fixed capital. Nor does it pursue income streams. (In some of the
newest knowledge organizations there is as yet barely enough income
to puddle, let alone stream.) Instead, investors see value in people
and their know-how -- people with the ability to envisage and execute
adventurous new business plans and to keep reenvisaging these to stay
ahead of the competition.
So, while the modern world often appears increasingly impersonal,
in those areas where knowledge really counts, people count more than
ever. In this way, a true knowledge economy should distinguish itself
not only from the industrial economy but also from an information
economy. For though its champions like to present these two as
distinct, the information economy, like the industrial economy, shows
a marked indifference to people. The industrial economy, for example,
treated them en masse as interchangeable parts -- the factory "hands"
of the nineteenth century. The information economy threatens to
treat them as more or less interchangeable consumers and processors of
information. Attending to knowledge, by contrast, returns attention
to people, what they know, how they come to know it, and how they
The importance of people as creators and carriers of knowledge is
forcing organizations to realize that knowledge lies less in its
databases than in its people. It's been said, for example, that
if NASA wanted to go to the moon again, it would have to start from
scratch, having lost not the data, but the human expertise that took
it there last time. Similarly, Tom Davenport and Larry Prusak argue
that when Ford wanted to build on the success of the Taurus, the
company found that the essence of that success had been lost with the
loss of the people that created it. Their knowledge was not stored in
information technologies. It left when they left.
Mistaking knowledge and its sources for information and its sources
can, then, be costly. In her book Wellsprings of Knowledge, Dorothy
Leonard-Barton of Harvard Business School tells the story of one
firm, ELP, taking over a rival, Grimes, primarily to capture Grimes's
impressive intellectual capital. Only after it had paid generously
for the business, machine tools, and research did ELP find that
Grimes's real competitive advantage had lain in the operating
knowledge of its line employees, all of whom had been let go.
Similarly, the sort of blind downsizing produced by business process
reengineering has caused organizations to lose "collective memory".
It's impossible to assess the value of such layoffs. But the
business journalist Thomas Stewart estimated the cost of AT&T's
last round as equivalent to an $8 billion capital write-off.8 In all,
the job of knowledge management cannot involve just the protection
and exploitation of patents. It must include the cultivation of
knowledgeable workers. Focusing on information, however, makes this
kind of cultivation difficult.
Curiously, if knowledge will go out of the door in the heads of people
who have developed and worked with that knowledge, it seems reluctant
to go out (or stay behind) in the heads of people who have not been
so involved. The CEO of the innovative steel manufacturer Chaparral
Steel told Leonard-Barton that for this reason the firm has no problem
with competitors touring their plant. Chaparral, he said, is willing
to show just about everything "and we will be giving away nothing
because they can't take it home with them". Unlike information,
knowledge, as we said, is hard to detach.
While the challenge of detaching knowledge from some people and
attaching it to others may protect some knowledge assets, it makes
management of the knowledge much more difficult. The difficulty has
revealed itself in, for example the struggle over "best practices".
To maintain competitive edge, firms first search for the best
practices either within their own or in their competitors' units.
Once identified, they then transfer these to areas where practices
less good. The search part has led to a great deal of useful
benchmarking. The transfer part, however, has proved much more
Robert Cole of the University of California at Berkeley's Haas
Business School has investigated this difficulty in a recent study of
best practice strategy. He looked at, among others, Hewlett-Packard's
attempts to raise quality levels in its plants around the globe by
identifying and circulating the best practices within the firm. Even
internally, Cole showed, transfer was uncertain. Cole's findings seem
to justify the now-famous lament of HP's chairman, Lew Platt, as he
considered how much better the firm would be "if only we knew what we
know at HP".
Although, as Cole emphasizes, HP works across continents and
countries, failure to transfer practice is not simply a matter of
national or linguistic boundaries. Best practices can have as much
trouble traveling across town as they do across continents. As one
winner of the prestigious Baldridge prize who grappled with this
problem told researchers in frustration, "We can have two plants right
across the street from one another, and it's the damndest thing to get
them to transfer best practices". Similarly, Jeff Papows, president
of Lotus, whose Notes is a widely used tool for knowledge management,
acknowledges that for all the power to communicate that Notes and
similar groupware provide, "spreading the practice has not been
IN DEFENSE OF LEARNING
Circulating human knowledge, these experiences suggest, is not
simply a matter of search and retrieval, as some views of knowledge
management might have us believe. While knowledge is often not all
that hard to search, it can be difficult to retrieve, if by retrieve
people mean detach from one knower and attach to another.
So learning, the acquisition of knowledge, presents knowledge
management with its central challenge. The defense of intellectual
property, the sowing and harvesting of information, the exploitation
of intellectual capital, and the benchmarking of competitors'
intellectual assets are all important parts of the knowledge
management game. But all of these are subordinate to the matter
of learning. For it is learning that makes intellectual property,
capital, and assets usable.
The difficulty of this central challenge, however, has been obscured
by the redefinition that, as we noted earlier, infoenthusiasts
tend to indulge. The definitions of knowledge management that began
this chapter perform a familiar two-step. First, they define the
core problem in terms of information, so that, second, they can put
solutions in the province of information technology.13 Here, retrieval
looks as easy as search.
If information retrieval were all that is required for such things as
knowledge management or best practice, HP would have nothing to worry
about. It has an abundance of very good information technology. The
persistence of HP's problem, then, argues that knowledge management,
knowledge, and learning involve more than information. In the rest
of this chapter we try to understand what else is involved, looking
primarily at knowledge and learning on the assumption that these need
to be understood before knowledge management can be considered.
To understand learning in the context of knowledge management, let's
begin by asking in what conditions do knowledge and best practice
move. In chapter 4, we saw the reps sharing their knowledge,
insights, and best practices quite effectively. These traveled first
among the small group of coworkers and then, with the help of the
Eureka database, across larger groups of company reps. To understand
how these best practices travel, this example suggests, requires
looking not simply from knowledge to information, but (as the idea of
best practice might suggest) from knowledge to practice and groups of
practitioners. For it is the reps' practice shared in collaborative
communities that allowed them to share their knowledge.
As we saw, the reps formed themselves into a small community, united
by their common practice, servicing machines. The members of this
community spent a lot of time both working and talking over work
together. In Orr's account, the talk and the work, the communication
and the practice are inseparable. The talk made the work
intelligible, and the work made the talk intelligible. As part of
this common work-and-talk, creating, learning, sharing, and using
knowledge appear almost indivisible. Conversely, talk without
the work, communication without practice is if not unintelligible,
at least unusable. Become a member of a community, engage in its
practices, and you can acquire and make use of its knowledge and
information. Remain an outsider, and these will remain indigestible.
Two learning researchers, whose individual work we mentioned earlier,
Jean Lave of the University of California, Berkeley, and Etienne
Wenger, a consultant formerly of the Institute for Research on
Learning, explain this sort of simultaneous working, learning,
and communication in terms of both the practice and the community.
Learning a practice, they argue, involves becoming a member of a
"community of practice" and thereby understanding its work and its
talk from the inside.14 Learning, from this point of view, is not
simply a matter of acquiring information; it requires developing the
disposition, demeanor, and outlook of the practitioners.
Like Orr's study of reps, Wenger's study of claims processing (see
chapter 4) showed the importance of the group to both what people
learn and how. Within the group, Wenger's study reveals, knowledge,
traveling on the back of practice, was readily shared.
It may at first seem that group practice and community support are
only appropriate for the tedium of "lowly" claims processing. They
might seem to have little do with the "higher" altitudes of knowledge
work, where the image of the lone, Rodinesque "thinker" is more
common. Yet the value of communities of practice to creating and
sharing knowledge is as evident in the labs of particle physicists
and biotechnologists as in the claims processing unit.15 The
apprenticeship-like activity that Lave and Wenger describe is found
not only on the shop floor, but throughout the highest reaches of
education and beyond. In the last years of graduate school or in
internships, scientists, humanists, doctors, architects, or lawyers,
after years of schoolroom training, learn their craft in the company
of professional mentors. Here, they form learning communities capable
of generating, sharing, and deploying highly esoteric knowledge.
Recently a computer engineer described a group that he led on a
difficult project that, despite the difference in subject matter,
resembles the groups of interdependent technicians and claims
[It] was less than half a dozen people; and the group that did the
software and hardware never did get to be more than about a dozen
people. It was a tiny enough group that everyone knew everything
that was going on, and there was very little structure . . . there
were people who specifically saw their role as software, and they
knew a lot about hardware anyway; and the hardware people all could
program. There wasn't a great deal of internal difficulty. There's
always a little, I don't think you can get even six people together
without having some kind of a problem. . . . There was amazingly
little argument or fighting.16
This description catches central properties of the community of
practice. In particular, it notes how, in getting the job done, the
people involved ignored divisions of rank and role to forge a single
group around their shared task, with overlapping knowledge, relatively
blurred boundaries, and a common working identity. The speaker in
this case is Frank Heart of Bolt Beranek and Newman; the group's task,
designing the core computers for what came to be the Internet. In
all, whether the task is deemed high or low, practice is an effective
teacher and the community of practice an ideal learning environment.
TO BE OR NOT TO BE
Of course, whatever the strengths of communities of practice, people
learn on their own, picking up information from numerous sources
about numerous topics without ever becoming a "member". We can learn
something about Tibetan medicine or racing without needing to work
with Tibetan doctors or become a Formula 1 driver. The critical words
here, however, are about and become. They point to a distinction made
by Jerome Bruner, a professor of psychology at New York University,
between learning about and learning to be. Certainly, most of
anyone's knowledge might best be described as knowledge "about".
Many people learn about a lot of things -- astrophysics, Australian
Rules football, Madagascan lemurs, or baseball statistics. In the age
of the Web, this learning about is easier than ever before.
But, picking up information about Madagascan lemurs in the comfort
of our home doesn't close the gap between us and Madagascan field
zoologists. Learning to be requires more than just information.
It requires the ability to engage in the practice in question.
Indeed, Bruner's distinction highlights another, made by the
philosopher Gilbert Ryle. He distinguishes "know that" from "know
how". Learning about involves the accumulation of "know that":
principally data, facts, or information. Learning about does not,
however, produce the ability to put "know that" into use. This, Ryle
argues, calls for "know how". And "know how" does not come through
accumulating information. (If it did, "know that" and "know how"
would, in the end, be indistinguishable -- build up enough "know that"
and you would become a practitioner.) "We learn how," Ryle argues,
"by practice". And, similarly, through practice, we learn to be.17
Ryle's philosophical argument may have brought us dangerously near
the realm of abstruse epistemology that we promised to avoid. But it
helps explain why the same stream of information directed at different
people doesn't produce the same knowledge in each. If the people are
engaged in different practices, if they are learning to be different
kinds of people, then they will respond to the information in
different ways. Practice shapes assimilation.
The practice of managing a baseball team, for example, is not the
same as the practice of playing on a baseball team. The "know that"
for each job may be fairly similar. Managers and players gather a
lot of the same information. But the "know how" for the two (thus
the way each makes use of their "know that") is quite different.
One's practice is to manage; the other's is to play. Similarly, while
management theorists and managers may posses similar "know that,"
their different practices keep them apart. The two can read the same
books, magazines, and journals, but these don't allow either to do the
other's job. A good management theorist may explain the practice of
management well, but never make a good hands-on manager. Similarly,
an excellent manager may prove an inept theoretician.18
LEARNING IN PRACTICE
Practice, then, both shapes and supports learning. We wouldn't need
to labor this point so heavily were it not that unenlightened teaching
and training often pulls in the opposite direction. First, they tend
to isolate people from the sorts of ongoing practice of work itself.
And second, they focus heavily on information.
Nowhere is this isolation more true than in the workplace. Or
perhaps we should say than not in the workplace. For while many of
the resources for learning to work lie in the workplace, training
regularly takes people away from there, to learn the job in
classrooms. The ideal of learning isolated from the "distractions" of
work practice still influences many training regimens. So let us look
briefly at a couple of examples that suggest some of the limits of the
classroom and the resources of practice.
Limits to Going by the Book
The first example draws on research by two educational psychologists,
George Miller and Patricia Gildea, into how children learn vocabulary.
Miller and Gildea compared learning words in the everyday practice
of conversation with trying to learn vocabulary from dictionaries.19
In the everyday case, they found that learning is startlingly fast
and successful. By listening, talking, and reading, the average
17-year-old has learned vocabulary at a rate of 5,000 words per year
(13 per day) for over 16 years. The children know both what these
words mean and how to use them.
By contrast, learning words from abstract definitions and sentences
from dictionaries is far slower and far less successful. Working
this way, the children in the study acquired between 100 and 200 words
per year. Moreover, much of what they learned turned out to be almost
useless in practice. Despite their best efforts, looking up relate,
careful, remedy and stir up in a dictionary led to sentences such
as, "Me and my parents correlate, because without them I wouldn't be
here"; "I was meticulous about falling off the cliff"; "The redress
for getting sick is staying in bed"; and "Mrs. Morrow stimulated the
Most of us have seen the workplace equivalent of this -- the eager
young intern with all the right information but none of the practical
knowledge that makes the job doable. Similarly, a lawyer friend of
ours recalled how the first days at work were a nightmare because,
despite all her excellent results in law school and on the law board
exams, nothing in the classroom had prepared her for the realities of
having a client on the other end of the telephone.
The Practical Value of Phone Cords
Another colleague, Jack Whalen, showed the power of practice in his
study of learning in a service center taking the calls from customers
and scheduling technicians.20 Sending technicians to fix broken
machines is an expensive undertaking. It is a waste if the problem
does not really require a technician. So the people who take the
calls can save the company money by diagnosing simple problems and
telling the customer how to fix these for themselves. It makes
customers happy, too. They don't have to sit with a dead machine,
waiting for a technician to bring it back to life.
The phone operators are not, of course, trained as technicians. In
the past, however, they learned from the reps when the latter called
in to pick up their next job. The reps would then explain how trivial
the last one had been, and in the process the phone operators could
learn a lot from these mentors. When they next took such a call, they
could offer a solution. As a result of a change in communications
technology, however, technicians no longer pick up their calls this
way. Consequently, operators no longer pick up insights. Their
opportunity for inherent learning has been lost.
The company has tried to replace this kind of learning with the
more explicit support of a "case-based expert system". This is
an information-based system that prompts operators to ask the
customer a series of questions. The operator types the responses
into the system, which then searches for a ready solution. This
alternative has not worked well. As the reps found with "directive
documentation," it can be surprisingly difficult to get a clear
diagnosis and solution this way. Moreover, such a system doesn't
help the operators understand what they are doing. And that lack of
understanding undermines the customer's confidence. It's hard to put
faith in people who are obviously reading instructions off a screen.
As a result, customers will ask for a technician anyway, and so defeat
the whole expert-system strategy.
To overcome these problems, the company contemplated new training
courses with several weeks off site to better prepare new operators.
Whalen and his fellow researchers took a slightly different route,
however. They studied one service center and the quality of
diagnosis its staff provided. There they found two operators who gave
especially reliable answers. One, unsurprisingly, was an eight-year
veteran of the service center with some college experience and
a survivor from the days when reps served as mentors. The other,
however, was someone with only a high-school diploma. She had been on
the job barely four months.
The researchers noticed, however, that the newcomer had a desk
opposite the veteran. There she could hear the veteran taking calls,
asking questions, and giving advice. And she began to do the same.
She had also noticed that he had acquired a variety of pamphlets and
manuals, so she began to build up her own stock. Moreover, when she
didn't understand the answers the veteran gave, she asked him to show
her what he meant, using the service center's own copier.
So instead of training courses, the sociologists suggested
restructuring the phone center. They sought to draw on its reservoir
of knowledge by putting all its operators in positions to learn from
each other. By opening the place up to this collective knowledge, the
redesign effectively created a small laboratory of what Whalen calls
"indigenous sharing and collaborative learning". The new plan also
asked technicians to come in and take calls intermittently. As a
result, operators could learn from them once again.
From these changes, the operators were up to speed in about the time
it took to plan a training course for them and in far less time than
was set aside for actual training. Ultimately, Whalen concluded,
given the amount and level of knowledge already available in the
room, what the operators needed were not so much expert systems or
new training courses, but "longer phone cords". 21 (These allow an
operator taking a call to slide over to the desk and the screen of
a resourceful colleague who could provide the necessary help.) Both
examples, the classroom and the workplace, indicate how the resources
for learning lie not simply in information, but in the practice
that allows people to make sense of and use that information and the
practitioners who know how to use that information. Where in other
circumstances knowledge is hard to move, in these circumstances it
travels with remarkable ease.
To venture cautiously again onto philosophical grounds, the
distinction between explicit and implicit dimensions of knowledge
can help illuminate why practice is so effective. It's possible,
for example, to learn about negotiation strategies by reading books
about negotiation. But strategy books don't make you into a good
negotiator, any more than dictionaries make you into a speaker or
expert systems make you into an expert. To become a negotiator
requires not only knowledge of strategy, but skill, experience,
judgment, and discretion. These allow you to understand not just how
a particular strategy is executed, but when to execute it. The two
together make a negotiator, but the second comes only with practice.
The chemist and philosopher Michael Polanyi distinguished these two
by alking about the explicit and the tacit dimensions of knowledge.
The explicit dimension is like the strategy book. But it is relatively
useless without the tacit dimension. This, Polanyi argues, allows
people to see when to apply the explicit part.22
To take another simple example of this sort of tacit "seeing,"
consider dictionaries again. These are the guidebooks of language
and particularly for spelling. But if you lack the tacit dimension
required for spelling, shelves of dictionaries do you no good. For
being able to use a dictionary (the explicit part) is not enough.
You have to know when to use a dictionary. A good speller will say,
"I just know that doesn't look right". This is the tacit part. Once
it has done its work, you can turn to the explicit information in
the dictionary. The problem for a bad speller, of course, is that if
he or she lacks the tacit knowing that makes words look wrong, then
a dictionary's use is limited. In the end, paradoxically, you only
learn to use a dictionary by learning to spell.
In making his distinction between explicit and tacit, Polanyi argues
that no amount of explicit knowledge provides you with the implicit.
They are two different dimensions of knowledge, and trying to reduce
one to the other is a little like trying to reduce a two-dimensional
drawing to one dimension. This claim of Polanyi's resembles Ryle's
argument that "know that" doesn't produce "know how," and Bruner's
that learning about doesn't, on its own, allow you to learn to be.
Information, all these arguments suggest, is on its own not enough
to produce actionable knowledge. Practice too is required. And for
practice, it's best to look to a community of practitioners.
Teach these boys nothing but Facts. Facts alone are wanted in life.
Plant nothing less and root out everything else. You can only form
the minds of reasoning animals upon Facts. . . . Stick to Facts,
Charles Dickens, Hard Times23
The view of knowledge and practice we have put forward here has
several implications for how to think about learning -- and related
issues such as spreading best practice, improving performance, or
innovating -- as well as training and teaching.
In the first place, it takes us beyond information. The idea of
learning as the steady supply of facts or information, though parodied
by Dickens 150 years ago, still prevails today. Each generation has
its own fight against images of learners as wax to be molded, pitchers
to be filled, and slates to be written on.
Literature about workplace learning is still laced with ideas of
"absorptive capacity," as if humans were information sponges. Indeed,
the idea that learning is mere information absorption may be on the
rise today because it allows for more redefinition. If we accept this
view of learning, then it's a short step to talking about such things
as computers or bots learning, as if what they do is just what people
do. Looking beyond information, as we have tried to do, provides a
richer picture of learning. From this picture, the following features
stand out for us.
Learning on Demand
Learning is usually treated as a supply-side matter, thought to follow
teaching, training, or information delivery. But learning is much
more demand driven. People learn in response to need. When people
cannot see the need for what's being taught, they ignore it, reject
it, or fail to assimilate it in any meaningful way. Conversely, when
they have a need, then, if the resources for learning are available,
people learn effectively and quickly.
In an essay we wrote about learning some years ago, we referred to
this aspect of learning as "stolen knowledge". We based this idea on
a short passage in the biography of the great Indian poet and Nobel
Prize winner Rabindrath Tagore.24 Talking of an instructor hired to
teach him music, Tagore writes, "He determined to teach me music,
and consequently no learning took place". Tagore found little to
interest him in the tedious tasks he was given as practice for these
involved not the authentic activity itself, but only a pale imitation.
"Nevertheless," he continues, "I did pick up from him a certain amount
of stolen knowledge". 25 This knowledge, Tagore reveals, he picked
up by watching and listening to the musician when the latter played
for his own and others' entertainment. Only then, when what was
evident was the practice of musicianship and not dismembered teaching
exercises, was Tagore able to see and appreciate the real practice at
A demand-side view of this sort of knowledge theft suggests how
important it is not to force-feed learning, but to encourage it, both
provoking the need and making the resources available for people to
"steal". We regard this as the paradoxical challenges of encouraging
and legitimating theft. Organizations have become increasingly adept
both at provoking and at responding to changes in their clients'
needs. They need to consider how to do this for their employees as
Despite the tendency to shut ourselves away and sit in Rodinesque
isolation when we have to learn, learning is a remarkably social
process. Social groups provide the resources for their members to
learn. Other socially based resources are also quite effective.
For example, people who are judged unfit to learn to operate simple
tools or who fail to master domestic appliances nevertheless learn to
operate complex machines that present users with hazardous, changing
environments and sophisticated technologies. We refer, of course, to
the car. Technologically, cars are extremely sophisticated. But they
are also extremely well integrated socially. As a result, learning
becomes almost invisible. Consider, by contrast, the triumphal
despair with which people frustratedly boast that they still can't
program their VCR. The success of learner drivers -- with or without
instruction -- should undoubtedly be the envy of many who design far
less difficult consumer or workplace appliances.
The car and the VCR make an interesting contrast. Almost everyone
in our society who learns to drive has already spent a great deal
of time traveling in cars or buses, along roads and highways. New
drivers begin formal instruction with an implicitly structured, social
understanding of the task. Now consider the VCR. Most people can use
their machine to play tapes. What they find difficult is recording,
though that's not a much more complex task. The central distinction
between these two functions is that one is often a social act, the
other highly individual. You might invite a group over to watch a
movie. You are unlikely to invite one over to watch you record.26
Learning and Identity Shape One Another
Bruner, with his idea of learning to be, and Lave and Wenger, in their
discussion of communities of practice, both stress how learning needs
to be understood in relation to the development of human identity.
In learning to be, in becoming a member of a community of practice,
an individual is developing a social identity. In turn, the identity
under development shapes what that person comes to know, how he or
she assimilates knowledge and information. So, even when people are
learning about, in Bruner's terms, the identity they are developing
determines what they pay attention to and what they learn. What
people learn about, then, is always refracted through who they are and
what they are learning to be.27
So information, while a critical part of learning, is only one among
many forces at work. Information theory portrays information as a
change registered in an otherwise steady state. It's a light flashing
out on a dark hillside (to borrow an example from the philosopher
Fred Dretske28) or the splash of a pebble breaking the calm of a still
lake. In either case, the result, as the anthropologist Gregory
Bateson puts it neatly, is "a difference that makes a difference". 29
The importance of disturbance or change makes it almost inevitable
that we focus on these. We notice the ripple and take the lake for
granted. Yet clearly the lake shapes the ripple more than the ripple
shapes the lake. Against a different background, the pebble would
register a different change or perhaps, in Bateson's terms, make no
difference at all. So to understand the whole interaction, it is as
important to ask how the lake is formed as to ask how the pebble got
there. It's this formation rather than information that we want to
draw attention to, though the development is almost imperceptible and
the forces invisible in comparison to the drama and immediacy of the
It's not, to repeat once more, the information that creates that
background. The background has to be in place for the information
to register. The forces that shape the background are, rather,
the tectonic social forces, always at work, within which and against
which individuals configure their identity. These create not only
grounds for reception, but grounds for interpretation, judgment, and
A Brief Note on the "Social"
We emphasize the social side of learning and identity with some
caution. The economist Friedrich Hayek claims that social is a weasel
word.30 Moreover, people readily point out that they can learn a great
deal sitting alone in an office or a library. And you don't have to
go very far with the thesis that learning is significantly social to
encounter the question "What about Robinson Crusoe?"31
Early economists liked to present Crusoe as an example of the homo
economicus, the universal economic man, learning and working in
splendid individual independence. And that's the idea behind this
question. It took Karl Marx to point out, however, that Crusoe is not
a universal. On his island (and in Defoe's mind), he is deeply rooted
in the society from which he came:
Our friend Robinson . . . having rescued a watch, ledger, and pen
and ink from the wreck, commences, like a true-born Briton, to keep
a set of books. His stock book contains a list of the objects of
utility that belong to him, of the operations necessary for their
production; and lastly, of the labour time that definite quantities
of those objects have, on average, cost him.32
Robinson is not just a man in isolation, but a highly representative
member of what Napoleon was to call a "nation of shopkeepers".
It is, of course, not only the British who play to type, even when
alone. The French philosopher Jean-Paul Sartre, in a famous passage,
illustrates how that true-born Frenchman, the cafe waiter, though
working alone, conforms his actions to society's idea of what a waiter
He returns, trying to imitate in his walk the inflexible stiffness
of some kind of automaton while carrying his tray with the
recklessness of a tight rope walker by putting it in a perpetually
unstable, perpetually broken equilibrium which he perpetually
re-establishes by a light movement of the arm and hand. . . . We
need not watch long before we can explain it: he is playing at being
a waiter in a cafe . . . . [T]he waiter plays with his condition in
order to realize it.33
So while people do indeed learn alone, even when they are not stranded
on desert islands or in small cafes, they are nonetheless always
enmeshed in society, which saturates our environment, however much
we might wish to escape it at times. Language, for example, is
a social artifact, and as people learn their way into it, they
are simultaneously inserting themselves into a variety of complex,
interwoven social systems.
Looking at learning as a demand-driven, identity forming, social act,
it's possible to see how learning binds people together. People with
similar practices and similar resources develop similar identities
-- the identity of a technician, a chemist, a lepidopterist, a
train spotter, an enologist, an archivist, a parking-lot attendant,
a business historian, a model bus enthusiast, a real estate developer,
or a cancer sufferer. These practices in common (for hobbies and
illnesses are practices too) allow people to form social networks
along which knowledge about that practice can both travel rapidly and
be assimilated readily.34
For the same reason, however, members of these networks are to some
degree divided or separated from people with different practices.
It is not the different information they have that divides them.
Indeed, they might have a lot of information in common. Rather, it
is their different attitudes or dispositions toward that information
-- attitudes and dispositions shaped by practice and identity --
that divide. Consequently, despite much in common, physicians are
different from nurses, accountants from financial planners.35
We see two types of work-related networks that, with the boundaries
they inevitably create, are critical for understanding learning, work,
and the movement of knowledge. First, there are the networks that
link people to others whom they may never get to know but who work
on similar practices. We call these "networks of practice". Second,
there are the more tight-knit groups formed, again through practice,
by people working together on the same or similar tasks. These are
what, following Lave and Wenger, we call "communities of practice".
Here we sketch the two briefly before elaborating their role in later
Networks of Practice
While the name "networks of practice" helps us to emphasize what we
see as the common denominator of these groups -- practice -- elsewhere
they go by the name of "occupational groups" or "social worlds".36
People in such networks have practice and knowledge in common.
Nevertheless, most of the members are unknown to one other.
Indeed, the links between the members of such networks are usually
more indirect than direct -- newsletters, Web sites, Bulletin
boards, listservs, and so forth keep them in touch and aware of one
another.37 Members coordinate and communicate through third parties
or indirectly. Coordination and communication are, as a result, quite
The 25,000 reps working for Xerox make up, in theory, such a network.
They could in principle be linked through such things as the Eureka
database (though it is in fact not worldwide) or corporate newsletters
aimed at reps. Their common practice makes these links viable,
allowing them to assimilate these communications in more-or-less
similar ways. By extension, the network could also include
technicians in other companies doing the same sort of work, though
here the connections would be weaker, grounds for common understanding
Networks of this sort are notable for their reach -- a reach now
extended and fortified by information technology. Information can
travel across vast networks with great speed and to large numbers but
nonetheless be assimilated in much the same way by whomever receives
it. By contrast, there is relatively little reciprocity across such
network; that is, network members don't interact with one another
directly to any significant degree. When reach dominates reciprocity
like this, it produces very loosely coupled systems.39 Collectively,
such social systems don't take action and produce little knowledge.
They can, though, share information relating to the members' common
practices quite efficiently.
Communities of Practice
Lave and Wenger's notion of communities of practice, which we
mentioned earlier, focuses on subsections of these larger networks
of practice. These subsections stand in contrast to the network
as a whole in several ways. They are relatively tight-knit groups
of people who know each other and work together directly. They are
usually face-to-face communities that continually negotiate with,
communicate with, and coordinate with each other directly in the
course of work. And this negotiation, communication, and coordination
is highly implicit, part of work practice, and, in the case of the
reps, work chat.40
While part of the network, groups like this cultivate their own
style, their own sense of taste, judgment, and appropriateness, their
own slang and in-terms. These things can distinguish members of one
community within a network from others. In networks of scholars, for
example, while all may be from one field, it's often easy to guess
who trained together in a particular lab or school by their style and
In these groups, the demands of direct coordination inevitably limit
reach. You can only work closely with so many people. On the other
hand, reciprocity is strong. People are able to affect one another
and the group as a whole directly. Changes can propagate easily.
Coordination is tight. Ideas and knowledge may be distributed across
the group, not held individually. These groups allow for highly
productive and creative work to develop collaboratively.
The divisions marked by the external boundaries of these groups
have significant implications for the development of organizations,
technologies, and indeed of societies as a whole. Yet they are
divisions that discussions of such developments easily overlook.
For example, discussions of the emerging "network society" suggest
that society is becoming a single, uniform entity. The network
stretches indefinitely, linking the individuals that stand at
each node to one another and providing them with common information.
Communities, organizations, nations, and the like disappear (victims
of the 6-Ds discussed in chapter 1). The network is all, configuring
itself more or less as the vaunted global village.
From the perspective of practice, rather than of process or
information, a rather different picture emerges. From this viewpoint,
any global network has a highly varied topography. While the whole
may ultimately be global, within it there are networks of practice
with lines of reach that are extensive but nonetheless bounded
by practice. And there are communities of practice, with dense
connections of both reach and reciprocity, which again put limits on
extent. These two, networks and communities, produce areas marked by
common identity and coordinated practice within any larger network.
And as a consequence of these areas, information does not travel
uniformly throughout the network. It travels according to the local
Curiously, organization theory suffers from similarly homogenizing
vision. It has been fashionable of late to talk of workplace culture
or organizational culture as if these made organizations internally
uniform. But divisions created by practice produce significant
variation here as well. Within organizations as without, connections
are dense in some places and thin in others. Sometimes these networks
extend across the boundaries of the organization. Elsewhere, they
may confront discontinuities within, where meaningful communication
breaks down. Business process reengineering, in particular, ignores
divisions created by different practices. Indeed, Hammer and
Champy's Reengineering the Corporation seeks to supersede the division
of labor that the economist Adam Smith saw as central to capitalist
production.41 Consequently, business process reengineering fails to
understand the internally varied terrain of organizations and its
fractures and divisions.
Sim Sitkin of the Fuqua School of Business at Duke University and a
colleague revealed similar blindness in "Total Quality Management".
Managers of one large organization, Sitkin and his colleague found,
attempted to implement a single scheme uniformly across a company.
The approach overlooked the different ideas of quality that different
practices develop. Pursuing a common goal in the face of these
divisions made the different groups feel that they were being judged
by the standards of others, and in consequence, fear their work
would either be undervalued or unrecognized. Rather than spreading
quality, the researchers concluded, the scheme only spread dissent and
Failure to read the topography may be at its most damaging as
people try to predict the effects of new information technologies
on organization. One of the remarkable things about these
technologies is their reach. Consequently, they are well suited
to support, develop, and even strengthen the networks of practice
we have described. As these networks readily span the borders
of organizations, their increasing strength will affect those
organizations. Whether networks will grow at the expense of
organizations is a question needing further research, not the linear
assumptions of infoenthusiasts. It seems improbable that they will
simply dissolve organization any more than, as we claimed in chapter
1, they will necessarily damage local communities, which remain
New technologies may, though, spread these communities out more
than before. The growing reciprocity available on the 'Net, while
probably underused at the moment, is helping people separated by space
maintain their dense interrelations. Yet for the sort of implicit
communication, negotiation, and collective improvisation that we
have described as part of practice, learning, and knowledge sharing,
it's clear that there are advantages to working together, however
well people may be connected by technology. Indeed, one of the most
powerful uses of information technology seems to be to support people
who do work together directly and to allow them to schedule efficient
face-to-face encounters. Looking too closely at the progression
from atoms to bits may miss the role the bits play in allowing us
to reinforce the valuable aspects of the world of atoms. Critical
movements in the knowledge economy may go not just from atoms to bits,
but from atoms to bits and back again.