Volume 7(2): 225246

Copyright 2000 SAGE

(London, Thousand Oaks, CA
and New Delhi)

Communities of Practice and

Social Learning Systems
articles
Etienne Wenger
Independent consultant, researcher, author and
speaker

Abstract. This essay argues that the success of organizations depends on

their ability to design themselves as social learning systems and also to
participate in broader learning systems such as an industry, a region, or
a consortium. It explores the structure of these social learning systems. It
proposes a social definition of learning and distinguishes between three
modes of belonging by which we participate in social learning systems.
Then it uses this framework to look at three constitutive elements of these
systems: communities of practice, boundary processes among these com-
munities, and identities as shaped by our participation in these systems.
Keywords: boundaries; communities of practice; identity; knowledge in
organizations; social learning systems

You probably know that the earth is round and that it is in orbit around
the sun. But how do you know this? What does it take? Obviously, it takes
a brain in a living body, but it also takes a very complex social, cultural,
and historical system, which has accumulated learning over time. People
have been studying the skies for centuries to understand our place in the
universe. More recently, scientific communities have developed a whole
vocabulary, observation methods, concepts, and models, which have been
adopted by other communities and have become part of popular thinking
in various ways. You have your own relationships to all these communi-
ties, and these relationships are what enable you to know about the

13505084[200005]7:2;225246;012580
 Organization 7(2)
Articles

earths position in the universe. In this sense, knowing is an act of par-

ticipation in complex social learning systems.
This essay assumes this view of knowing to consider how organizations
depend on social learning systems. First, I outline two aspects of a con-
ceptual framework for understanding social learning systems: a social
definition of learning in terms of social competence and personal experi-
ence, and three distinct modes of belonging through which we participate
in social learning systems: engagement, imagination, and alignment.
Then I look at three structuring elements of social learning systems: com-
munities of practice, boundary processes among these communities, and
identities as shaped by our participation in these systems. About each of
these elements I use my conceptual framework to ask three questions.
Why focus on it? Which way is up, that is, how to construe progress in
this area? And, third, what is doable, that is, what are elements of design
that one can hope to influence? Finally, I argue that organizations both are
constituted by and participate in such social learning systems. Their suc-
cess depends on their ability to design themselves as social learning sys-
tems and also to participate in broader learning systems such as an
industry, a region, or a consortium.
The conceptual framework I introduce here is intended for organiz-
ational design as well as analysis. The questions I ask are meant to guide
the inquiry of the researcher as well as the actions of the practitioner:
what to pay attention to, how to give direction to our initiatives, and
where to focus our efforts. As Kurt Lewin used to say, there is nothing as
practical as a good theory.

Aspects of a Conceptual Framework

A framework for understanding social learning systems must make it
possible to understand learning as a social process. What is learning from
a social perspective? And what are the processes by which our learning
constitutes social systems and social identities?

A Social Definition of Learning

In a social learning system, competence is historically and socially
defined. How to be a physicist or how to understand the position of the
earth in the universe is something that scientific communities have estab-
lished over time. Knowing, therefore, is a matter of displaying compe-
tences defined in social communities. The picture is more complex and
dynamic than that, however. Our experience of life and the social stan-
dards of competence of our communities are not necessarily, or even
usually, congruent. We each experience knowing in our own ways.
Socially defined competence is always in interplay with our experience.
It is in this interplay that learning takes place.
Consider two extreme cases. Sometimes, we are a newcomer. We join a
new community. We are a child who cannot speak yet. Or we are a new
226
 Communities of Practice
Etienne Wenger

employee. We feel like a bumbling idiot among the sages. We want to

learn. We want to apprentice ourselves. We want to become one of them.
We feel an urgent need to align our experience with the competence they
define. Their competence pulls our experience.
Sometimes, it is the other way round. We have been with a community
for a long time. We know the ropes. We are thoroughly competent, in our
own eyes and in the eyes of our peers. But something happens. We are
sent overseas. We go to a conference. We visit another department. We
meet a stranger with a completely different perspective. Or we just take
a long walk or engage in a deep conversation with a friend. Whatever the
case may be, we have an experience that opens our eyes to a new way of
looking at the world. This experience does not fully fit in the current
practice of our home communities. We now see limitations we were not
aware of before. We come back to our peers, try to communicate our
experience, attempt to explain what we have discovered, so they too can
expand their horizon. In the process, we are trying to change how our
community defines competence (and we are actually deepening our own
experience). We are using our experience to pull our communitys com-
petence along.
Whether we are apprentices or pioneers, newcomers or oldtimers,
knowing always involves these two components: the competence that our
communities have established over time (i.e. what it takes to act and be
recognized as a competent member), and our ongoing experience of the
world as a member (in the context of a given community and beyond).
Competence and experience can be in various relations to each other
from very congruent to very divergent. As my two examples show, either
can shape the other, although usually the process is not completely one-
way. But, whenever the two are in close tension and either starts pulling
the other, learning takes place. Learning so defined is an interplay
between social competence and personal experience. It is a dynamic, two-
way relationship between people and the social learning systems in
which they participate. It combines personal transformation with the
evolution of social structures.

Modes of Belonging
Our belonging to social learning systems can take various forms at vari-
ous levels between local interactions and global participation. To capture
these different forms of participation, I will distinguish between three
modes of belonging.
Engagement: doing things together, talking, producing artifacts (e.g.
helping a colleague with a problem or participating in a meeting). The
ways in which we engage with each other and with the world pro-
foundly shape our experience of who we are. We learn what we can do
and how the world responds to our actions.
Imagination: constructing an image of ourselves, of our communities,

227
Organization 7(2)
Articles

and of the world, in order to orient ourselves, to reflect on our situation,

and to explore possibilities (e.g. drawing maps, telling a story, or build-
ing a set of possible scenarios to understand ones options). I use
imagination here in the sense proposed by Benedict Anderson (1983) to
describe nations as communities: it does not connote fantasy as
opposed to factuality. Knowing that the earth is round and in orbit
around the sun, for instance, is not a fantasy. Yet it does require a
serious act of imagination. It requires constructing an image of the uni-
verse in which it makes sense to think of our standing on the ground as
being these little stick figures on a ball flying through the skies.
Similarly, thinking of ourselves as a member of a community such as a
nation requires an act of imagination because we cannot engage with
all our fellow citizens. These images of the world are essential to our
sense of self and to our interpretation of our participation in the social
world.
Alignment: making sure that our local activities are sufficiently
aligned with other processes so that they can be effective beyond our
own engagement (e.g. doing a scientific experiment by the book, con-
vincing a colleague to join a cause, or negotiating a division of labor
and a work plan for a project). The concept of alignment as used here
does not connote a one-way process of submitting to external auth-
ority, but a mutual process of coordinating perspectives, interpret-
ations, and actions so they realize higher goals. Following the
scientific method, abiding by a moral code, or discussing important
decisions with our spouse can all become very deep aspects of our
identities.

Distinguishing between these modes of belonging is useful for two

reasons. First, analytically, each mode contributes a different aspect to the
formation of social learning systems and personal identities. Engagement,
imagination, and alignment usually coexist and every social learning
system involves each to some degree and in some combination. Still, one
can dominate and thus give a different quality to a social structure. For
instance, a community based mostly on imagination such as a nation has
a very different quality from a community of practice at work, which is
based primarily on engagement. I would in fact argue that these modes of
belonging provide a foundation for a typology of communities.
Second, practically, each mode requires a different kind of work. The
work of engagement, which requires opportunities for joint activities, is
different from the work of imagination, which often requires oppor-
tunities for taking some distance from our situation. The demands and
effects of these three modes of belonging can be conflicting. Spending
time reflecting can detract from engagement, for example. The modes can
also be complementary, however. For instance, using imagination to gain
a good picture of the context of ones actions can help in fine-tuning align-
ment because one understands the reasons behind a procedure or an

228
 Communities of Practice
Etienne Wenger

agreement. It is therefore useful to strive to develop these modes of

belonging in combination, balancing the limitations of one with the work
of another. For instance, reflective periods that activate imagination or
boundary interactions that require alignment with other practices around
a shared goal could be used to counteract the possible narrowness of
engagement (Wenger, 1998).

Communities of Practice
Since the beginning of history, human beings have formed communities
that share cultural practices reflecting their collective learning: from a
tribe around a cave fire, to a medieval guild, to a group of nurses in a
ward, to a street gang, to a community of engineers interested in brake
design. Participating in these communities of practice is essential to our
learning. It is at the very core of what makes us human beings capable of
meaningful knowing.

Why Focus on Communities?

Communities of practice are the basic building blocks of a social learning
system because they are the social containers of the competences that
make up such a system. By participating in these communities, we define
with each other what constitutes competence in a given context: being a
reliable doctor, a gifted photographer, a popular student, or an astute
poker player. Your company may define your job as processing 33 medi-
cal claims a day according to certain standards, but the competence
required to do this in practice is something you determine with your col-
leagues as you interact day after day.
Communities of practice define competence by combining three
elements (Wenger, 1998). First, members are bound together by their col-
lectively developed understanding of what their community is about and
they hold each other accountable to this sense of joint enterprise. To be
competent is to understand the enterprise well enough to be able to con-
tribute to it. Second, members build their community through mutual
engagement. They interact with one another, establishing norms and
relationships of mutuality that reflect these interactions. To be competent
is to be able to engage with the community and be trusted as a partner in
these interactions. Third, communities of practice have produced a
shared repertoire of communal resourceslanguage, routines, sensibili-
ties, artifacts, tools, stories, styles, etc. To be competent is to have access
to this repertoire and be able to use it appropriately.
Communities of practice grow out of a convergent interplay of compe-
tence and experience that involves mutual engagement. They offer an
opportunity to negotiate competence through an experience of direct par-
ticipation. As a consequence, they remain important social units of learn-
ing even in the context of much larger systems. These larger systems are
constellations of interrelated communities of practice.
229
 Organization 7(2)
Articles

Which Way is Up?

Communities of practice cannot be romanticized. They are born of learn-
ing, but they can also learn not to learn. They are the cradles of the human
spirit, but they can also be its cages. After all, witch-hunts were also com-
munity practices. It is useful, therefore, to articulate some dimensions of
progress.
Enterprise: the level of learning energy. How much initiative does the
community take in keeping learning at the center of its enterprise? A
community must show leadership in pushing its development along
and maintaining a spirit of inquiry. It must recognize and address gaps
in its knowledge as well as remain open to emergent directions and
opportunities.
Mutuality: the depth of social capital. How deep is the sense of com-
munity generated by mutual engagement over time? People must know
each other well enough to know how to interact productively and who
to call for help or advice. They must trust each other, not just person-
ally, but also in their ability to contribute to the enterprise of the com-
munity, so they feel comfortable addressing real problems together and
speaking truthfully. Through receiving and giving help, they must gain
enough awareness of the richness of the community to expect that their
contribution will be reciprocated in some way.
Repertoire: the degree of self-awareness. How self-conscious is the
community about the repertoire that it is developing and its effects on
its practice? The concepts, language, and tools of a community of prac-
tice embody its history and its perspective on the world. Being reflec-
tive on its repertoire enables a community to understand its own state
of development from multiple perspectives, reconsider assumptions
and patterns, uncover hidden possibilities, and use this self-awareness
to move forward.
The three dimensions work together. Without the learning energy of those
who take initiative, the community becomes stagnant. Without strong
relationships of belonging, it is torn apart. And without the ability to
reflect, it becomes hostage to its own history. The work associated with
each mode of belonging can contribute to these criteria. Table 1 illustrates
how the modes of belonging interact with community elements.

What is Doable?
When designing itself, a community should look at the following
elements: events, leadership, connectivity, membership, projects, and
artifacts.

Events. You can organize public events that bring the community together.
Obviously, these may or may not be attended, but if they are well tuned
to the communitys sense of its purpose, they will help it develop an

230
 Communities of Practice
Etienne Wenger

Table 1. Community Dimensions

Enterprise: learning Mutuality: social Repertoire:
energy capital self-awareness
Engagement What are the What events and To what extent have
opportunities to interactions weave the shared experience,
negotiate a joint inquiry community and develop language, artifacts,
and important trust? Does this result in histories, and methods
questions? Do members an ability to raise accumulated over time,
identify gaps in their troubling issues during and with what potential
knowledge and work discussions? for further interactions
together to address and new meanings?
them?
Imagination What visions of the What do people know Are there self-
potential of the about each other and representations that
community are guiding about the meanings that would allow the
the thought leaders, participation in the community to see itself
inspiring participation, community takes in their in new ways? Is there a
and defining a learning lives more broadly? language to talk about
agenda? And what the community in a
picture of the world reflective mode?
serves as a context for
such visions?
Alignment Have members What definitions of roles, What traditions,
articulated a shared norms, codes of methods, standards,
purpose? How widely do behavior, shared routines, and
they subscribe to it? principles, and frameworks define the
How accountable do they negotiated commitments practice? Who upholds
feel to it? And how and expectations hold them? To what extent
distributed is leadership? the community together? are they codified? How
are they transmitted to
new generations?

identity. A community will have to decide the type of activities it needs:

formal or informal meetings, problem-solving sessions, or guest speakers.
It will also have to consider the rhythm of these events given other
responsibilities members have: too often and people just stop coming, too
rare and the community does not gain momentum. This rhythm may also
have to change over time or go through cycles.

Leadership. Communities of practice depend on internal leadership, and

enabling the leaders to play their role is a way to help the community
develop. The role of community coordinator who takes care of the day-
to-day work is crucial, but a community needs multiple forms of leader-
ship: thought leaders, networkers, people who document the practice,
pioneers, etc. These forms of leadership may be concentrated on one or
two members or widely distributed, and this will change over time.

231
 Organization 7(2)
Articles

Connectivity. Building a community is not just a matter of organizing com-

munity events but also of enabling a rich fabric of connectivity among
people. This could involve brokering relationships between people who
need to talk or between people who need help and people who can offer
help. It is also important to make it possible for people to communicate
and interact in multiple media.

Membership. A communitys members must have critical mass so that there is

interest, but it should not become so wide that the focus of the com-
munity is diffuse and participation does not grab peoples identities.
Including those who are missing can be very helpful in consolidating the
legitimacy of the community to itself and in the wider organization.
Conversely, realizing that the membership is overextended allows the
community to split up into subgroups. Finally, devising processes by
which newcomers can become full members helps ensure access for new-
comers without diluting the communitys focus.

Learning Projects. Communities of practice deepen their mutual commitment

when they take responsibility for a learning agenda, which pushes their
practice further. Activities toward this goal include exploring the knowl-
edge domain, finding gaps in the community practice, and defining proj-
ects to close these gaps. Such learning projects could involve, for
instance, assessing some tools, building a generic design, doing a litera-
ture search, creating a connection with a university doing research in the
area, or simply interviewing some experts to create a beginners guide.

Artifacts. All communities of practice produce their own set of artifacts: doc-
uments, tools, stories, symbols, websites, etc. A community has to con-
sider what artifacts it needs and who has the energy to produce and
maintain them so they will remain useful as the community evolves.

Boundaries
The term boundary often has negative connotations because it conveys
limitation and lack of access. But the very notion of community of prac-
tice implies the existence of boundary. Unlike the boundaries of organiz-
ational units, which are usually well defined because affiliation is
officially sanctioned, the boundaries of communities of practice are
usually rather fluid. They arise from different enterprises; different ways
of engaging with one another; different histories, repertoires, ways of
communicating, and capabilities. That these boundaries are often unspo-
ken does not make them less significant. Sit for lunch by a group of high-
energy particle physicists and you know about boundary, not because
they intend to exclude you, but because you cannot figure out what they
are talking about. Shared practice by its very nature creates boundaries.
Yet, if you are like me, you will actually enjoy this experience of bound-

232
 Communities of Practice
Etienne Wenger

ary. There is something disquieting, humbling at times, yet exciting and

attractive about such close encounters with the unknown, with the mys-
tery of otherness: a chance to explore the edge of your competence, learn
something entirely new, revisit your little truths, and perhaps expand
your horizon.

Why Focus on Boundaries?

Boundaries are important to learning systems for two reasons. They con-
nect communities and they offer learning opportunities in their own
right. These learning opportunities are of a different kind from the ones
offered by communities. Inside a community, learning takes place
because competence and experience need to converge for a community to
exist. At the boundaries, competence and experience tend to diverge: a
boundary interaction is usually an experience of being exposed to a
foreign competence. Such reconfigurations of the relation between com-
petence and experience are an important aspect of learning. If compe-
tence and experience are too close, if they always match, not much
learning is likely to take place. There are no challenges; the community is
losing its dynamism and the practice is in danger of becoming stale.
Conversely, if experience and competence are too disconnected, if the
distance is too great, not much learning is likely to take place either.
Sitting by that group of high-energy particle physicists, you might not
learn much because the distance between your own experience and the
competence you are confronting is just too great. Mostly what you are
learning is that you do not belong.
Learning at boundaries is likely to be maximized for individuals and for
communities when experience and competence are in close tension.
Achieving a generative tension between them requires:
something to interact about, some intersection of interest, some activity;
open engagement with real differences as well as common ground;
commitment to suspend judgment in order to see the competence of a
community in its terms;
ways to translate between repertoires so that experience and compe-
tence actually interact.
Boundaries are sources of new opportunities as well as potential difficul-
ties. In a learning system, communities and boundaries can be learning
assets (and liabilities) in complementary ways.
Communities of practice can steward a critical competence, but they
can also become hostage to their history, insular, defensive, closed in,
and oriented to their own focus.
Boundaries can create divisions and be a source of separation, fragmen-
tation, disconnection, and misunderstanding. Yet, they can also be areas
of unusual learning, places where perspectives meet and new possibil-
ities arise. Radically new insights often arise at the boundaries between
233
 Organization 7(2)
Articles

communities. Think of a specialization like psychoneuroimmunology:

its very name reflects its birth at the intersection of multiple practices.
In social learning systems, the value of communities and their boundaries
are complementary. Deep expertise depends on a convergence between
experience and competence, but innovative learning requires their diver-
gence. In either case, you need strong competences to anchor the process.
But these competences also need to interact. The learning and innovation
potential of a social learning system lies in its configuration of strong core
practices and active boundary processes (Wenger, 1998).

Which Way is Up?

Not all boundary processes create bridges that actually connect practices
in deep ways. The actual boundary effects of these processes can be
assessed along the following dimensions.
Coordination. Can boundary processes and objects be interpreted in dif-
ferent practices in a way that enables coordinated action? For instance,
an elegant design may delight designers but say little to those concerned
with manufacturability. Across boundaries, effective actions and use of
objects require new levels of coordination. They must accommodate the
practices involved without burdening others with the details of one
practice and provide enough standardization for people to know how to
deal with them locally.
Transparency. Do boundary processes give access to the meanings they
have in various practices? Coordination does not imply that boundary
processes provide an understanding of the practices involved. For
instance, forms like US tax returns enable coordination across bound-
aries (you know how to fill them out by following instructions line by
line), but often afford no windows into the logic they are meant to
enforce (following instructions often tells you little about why these
calculations are fair).
Negotiability. Do boundary processes provide a one-way or a two-way
connection? For instance, a business process reengineering plan may be
very detailed about implementation (coordination) and explicit about
its intentions (transparency), but reflect or allow little negotiation
between the perspectives involved. Boundary processes can merely
reflect relations of power among practices, in which case they are likely
to reinforce the boundary rather than bridge it. They will bridge prac-
tices to the extent that they make room for multiple voices.
Table 2 explores how the three modes of belonging affect these qualities
of boundary processes.

What is Doable?
Boundary processes are crucial to the coherent functioning of social
learning systems. A number of elements can be intentionally promoted in

234
 Communities of Practice
Etienne Wenger

Table 2. Boundary Dimensions

Coordination Transparency Negotiability
Engagement What opportunities exist Do people provide Are joint activities
for joint activities, explanations, coaching, structured in such a way
problem-solving, and and demonstrations in that multiple
discussions to both the context of joint perspectives can meet
surface and resolve activities to open and participants can
differences through windows on to each come to appreciate each
action? others practices? others competences?
Imagination Do people have enough What stories, Can both sides see
understanding of their documents, and models themselves as members
respective perspectives are available to build a of an overarching
to present issues picture of another community in which they
effectively and anticipate practice? What have common interests
misunderstandings? experience will allow and needs?
people to walk in the
others shoes? Do they
listen deeply enough?
Alignment Are instructions, goals, Are intentions, Who has a say in
and methods commitments, norms, negotiating contracts
interpretable into action and traditions made and devising
across boundaries? clear enough to reveal compromises?
common ground and
differences in
perspectives and
expectations?

an effort to weave these systems more tightly together. Here, I will talk
about three types of bridges across boundaries: people who act as bro-
kers between communities, artifacts (things, tools, terms, representa-
tions, etc.) that serve as what Star and Griesemer (1989) call boundary
objects, and a variety of forms of interactions among people from differ-
ent communities of practice.

Brokering. Some people act as brokers between communities. They can

introduce elements of one practice into another. Although we all do some
brokering, my experience is that certain individuals seem to thrive on being
brokers: they love to create connections and engage in importexport, and
so would rather stay at the boundaries of many practices than move to the
core of any one practice. Brokering can take various forms, including:
boundary spanners: taking care of one specific boundary over time;
roamers: going from place to place, creating connections, moving
knowledge;
outposts: bringing back news from the forefront, exploring new territo-
ries;

235
 Organization 7(2)
Articles

pairs: often brokering is done through a personal relationship between

two people from different communities and it is really the relationship
that acts as a brokering device.
Brokering knowledge is delicate. It requires enough legitimacy to be lis-
tened to and enough distance to bring something really new. Because bro-
kers often do not fully belong anywhere and may not contribute directly
to any specific outcome, the value they bring can easily be overlooked.
Uprootedness, homelessness, marginalization, and organizational invisi-
bility are all occupational hazards of brokering. Developing the boundary
infrastructure of a social learning system means paying attention to
people who act as brokers. Are they falling through the cracks? Is the
value they bring understood? Is there even a language to talk about it? Are
there people who are potential brokers but who for some reason do not
provide cross-boundary connections?

Boundary Objects. Some objects find their value, not just as artifacts of one
practice, but mostly to the extent that they support connections between
different practices. Such boundary objects can take multiple forms.
Artifacts, such as tools, documents, or models. For instance, medical
records and architectural blueprints play a crucial role in connecting
multiple practices (doctors/nurses/insurers, architects/contractors/
city planners).
Discourses. A critical boundary object is the existence of a common lan-
guage that allows people to communicate and negotiate meanings
across boundaries. This was an important thrust behind the quality
movement, and it was typified by the six sigma discourse at Motorola.
Processes. Shared processes, including explicit routines and pro-
cedures, allow people to coordinate their actions across boundaries.
Business processes, for instance, are not just fixed prescriptive defi-
nitions. At their best, they act as boundary objects that allow multiple
practices to coordinate their contributions.
Boundary objects do not necessarily bridge across boundaries because
they may be misinterpreted or interpreted blindly. Rethinking artifacts
and designs in terms of their function as boundary objects often illumi-
nates how they contribute to or hinder the functioning of learning sys-
tems. An organizational structure, for instance, is often considered as an
overarching umbrella that incorporates multiple parts by specifying their
relationships. But, in fact, it is more usefully designed as a boundary
object intended to enable multiple practices to negotiate their relation-
ships and connect their perspectives.

Boundary interactions
Boundary encounters. These encountersvisits, discussions, sabbati-
calsprovide direct exposure to a practice. They can take different

236
 Communities of Practice
Etienne Wenger

forms for different purposes. When one person visits, as in a sabbatical,

it is easier to get fully immersed in the practice, but more difficult to
bring the implications home because the very immersion into a foreign
practice tends to isolate you from your peers. GM, for instance, has had
difficulty learning from people sent on sabbatical at its more experi-
mental units such as NUMMI and Saturn because their transformed per-
spectives could not find a place back home. When a delegation of two
or more people visit, as in a benchmarking expedition, they may not get
as fully immersed, but they can negotiate among themselves the mean-
ing of the boundary interaction for their own practice, and therefore
find it easier to bring their learning back home.
Boundary practices. In some cases, a boundary requires so much sus-
tained work that it becomes the topic of a practice of its own. At Xerox,
as in many companies, some people are charged with the task of main-
taining connections between the R&D lab and the rest of the corpor-
ation. They are developing a practice of crossing these boundaries
effectively. Of course, the risk of these boundary practices is that they
create their own boundaries, which can prevent them from functioning
as brokers. It is necessary, therefore, to keep asking how the elements of
the boundary practiceits enterprise, its relationships, its repertoire
contribute to creating a bridge and how the community deals with its
own boundaries. And, sometimes, a new practice in its own right does
develop at these boundaries, which is worth paying attention to in its
own terms.
Peripheries. Communities often have to take steps to manage their
boundaries to serve people who need some service, are curious, or
intend to become members. Many communities have found it useful to
create some facilities by which outsiders can connect with their prac-
tice in peripheral ways. Examples of such facilities include lists of fre-
quently asked questions, visitors rooms on websites, open houses and
fairs. Some communities have even established help desks to provide
access to their expertise in a more efficient way. The idea behind many
of these facilities is to provide for some boundary activities without
overwhelming the community itself with the task of accommodating
outsiders demands. For newcomers, some communities organize intro-
ductory events, mentoring relationships, or even formal apprenticeship
systems.

Cross-disciplinary Projects. In most organizations, members of communities

of practice contribute their competence by participating in cross-func-
tional projects and teams that combine the knowledge of multiple prac-
tices to get something done. Simultaneous participation in communities
of practice and project teams creates learning loops that combine appli-
cation with capability development. In these double-knit organizations,
as Richard McDermott (1999) calls them, the learning and innovation that
is inherent in projects is synthesized and disseminated through the home

237
 Organization 7(2)
Articles

communities of practice of team members. The new knowledge can then

be applied and expanded in new projects, and the cycle goes on.
Such a perspective brings up a different way of thinking about these
projects. From the standpoint of the task to be accomplished, these proj-
ects are cross-disciplinary because they require the contribution of mul-
tiple disciplines. But, from the perspective of the development of
practices, they are boundary projects. Indeed, participating in these kinds
of projects exposes practitioners to others in the context of specific tasks
that go beyond the purview of any practice. People confront problems
that are outside the realm of their competence but that force them to nego-
tiate their own competence with the competences of others. Such projects
provide a great way to sustain a creative tension between experience and
competence when our participation in a project leverages and nourishes
our participation in a community of practice.

Identities
As I said, you probably know that the earth is round and in orbit around
the sun. Of course, it is not a flat plate in the way it appears to be at first
glance. You actually want to make sure you know this. It is part of your
identity as the kind of well-educated adult you probably are if you are
reading this article. You may even know that the orbit is not an exact
circle, but a slight ellipse. Chances are, however, you do not know the
exact distance between the earth and the sun or the precise difference
between the apogee and the perigee. This kind of ignorance, your ident-
ity can accept without existential angst because your relationship to the
communities where such knowledge matters is very peripheral at best.
I am not trying to make you feel self-conscious about your knowledge
of astrophysics. There will be no test at the end of this article. (Did I hear
a sigh of relief? No, no, you are perfectly OK just knowing the earth is
round, and many of our fellow human beings have lived very good lives
not even knowing that.) My point is that, if knowing is an act of belong-
ing, then our identities are a key structuring element of how we know.

Why Focus on Identity?

Knowing, learning, and sharing knowledge are not abstract things we do
for their own sake. They are part of belonging (Eckert, 1989). When I was
working with claims processors in an insurance company, I noticed that
their knowing was interwoven in profound ways with their identities as
participants in their community of practice. Their job did not have a high
status in the company (and in their own eyes, for that matter), so they
were careful not to be interested in it more than was absolutely necessary.
What they knew about their job, what they tried to understand and what
they accepted not to understand about the forms they had to fill out, what
they shared with each other, all that was not merely a matter of necessity
to get the job done, but it was also a matter of identity. Knowing too much

238
 Communities of Practice
Etienne Wenger

or failing to share a crucial piece of knowledge would be a betrayal of

their sense of self and of their community (Wenger, 1998).
In the landscape of communities and boundaries in which we live, we
identify with some communities strongly and not at all with others. We
define who we are by what is familiar and what is foreign, by what we
need to know and what we can safely ignore. You are a cello player, but
not the conductor who signals your entry, or the dancer who dances the
ballet you are playing, or the lawyer whom you saw this afternoon about
your uncles estate. We define ourselves by what we are not as well as by
what we are, by the communities we do not belong to as well as by the
ones we do. These relationships change. We move from community to
community. In doing so, we carry a bit of each as we go around. Our iden-
tities are not something we can turn on and off. You dont cease to be a
parent because you go to work. You dont cease to be a nurse because you
step out of the hospital. Multimembership is an inherent aspect of our
identities.
Identity is crucial to social learning systems for three reasons. First, our
identities combine competence and experience into a way of knowing.
They are the key to deciding what matters and what does not, with whom
we identify and whom we trust, and with whom we must share what we
understand. Second, our ability to deal productively with boundaries
depends on our ability to engage and suspend our identities. Learning
from our interactions with other practices is not just an intellectual matter
of translation. It is also a matter of opening up our identities to other ways
of being in the world. Third, our identities are the living vessels in which
communities and boundaries become realized as an experience of the
world. Whenever we belong to multiple communities, we experience the
boundary in a personal way. In the process, we create bridges across
communities because, in developing our own identities, we deal with
these boundaries in ourselves.

Which Way is Up?

Our identities are not necessarily strong or healthy. Sometimes, they are
even self-defeating. In fact, a whole self-help industry has flourished by
offering advice for building healthy identities (Giddens, 1991). Navigating
the social landscape defined by communities and their boundaries
requires a strong identity. Progress can be described in terms of a few
crucial qualities that must coexist to constitute a healthy social identity.
Connectedness. Where are enduring social relationships through which
an identity gains social depth? An identity is not an abstract idea
or a label, such as a title, an ethnic category, or a personality trait.
It is a lived experience of belonging (or not belonging). A strong
identity involves deep connections with others through shared
histories and experiences, reciprocity, affection, and mutual commit-
ments.

239
Organization 7(2)
Articles

Expansiveness. What are the breadth and scope of an identity?

A healthy identity will not be exclusively locally defined. It will involve
multimembership and cross multiple boundaries. It will seek a wide
range of experiences and be open to new possibilities. It will identify
with broad communities that lie beyond direct participation.
Effectiveness. Does an identity enable action and participation? Identity
is a vehicle for participating in the social world, but it can also lead to
non-participation. A healthy identity is socially empowering rather
than marginalizing.
There are potential tensions and conflicts between these qualities. How
big can your identity be and still be engaged as well as effective (not
merely an abstract kind of identification)? Can you really think globally
and act locally, feel like a citizen of the earth without losing your ability
to connect with specific communities? Can you live on the Internet and
still have a good marriage? In other words, it is the combination of these
qualities that matters. Table 3 explores how each mode of belonging con-
tributes to these three qualities.

Table 3. Identity Dimensions

Connectedness Expansiveness Effectiveness
Engagement Is there a community to Is there enough variety Do you have
engage with? How far of contexts and identity- opportunities to develop
back do you go? What forming experiences, socially recognized
kinds of interactions do such as logging on the competences by
you have? What do you Internet and chatting participating in well-
do together? Do you with strangers, going on established practices?
trust and are you a blind date, or visiting a Are your communities
trusted? foreign country? ready to embrace your
experience into their
practices?
Imagination Do you have good Can you see yourself as Do you understand the
conversations? Do you a member of large big picture well enough
talk about your deepest communities, for to act effectively?
aspirations? Do you instance, a world citizen,
listen well? the heir of long-lived
traditions, the pioneer of
a world to come?
Alignment Do you keep your Do you follow guidelines Do you know the
commitments to your that align your actions regimes of accountability
communities? Do you with broader purposes, by which your ideas,
uphold their principles? such as saving energy or actions, and requests
Do you give and receive recycling for the sake of will be judged? Can you
feedback? the planet? convince others of the
potential of a new idea?

240
 Communities of Practice
Etienne Wenger

What is Doable?
To help identities achieve simultaneously high degrees of local connect-
edness, global expansiveness, and social effectiveness, here are some
design elements to consider:

Home Base. Identity needs a place where a person can experience knowing as
a form of social competence. Think of a project-based organization, for
instance, where people go from one project to the next, spending a few
days in-between on the available list. The learning that they do in their
projects does not have a social home, unless they can also belong to a
community of practice. In such a community, they are not only recog-
nized as competent for the sake of a project, their need to develop their
competence is also part of their belonging. Their professional develop-
ment and the development of the practice go hand in hand: the identity
of the community as it evolves parallels the evolution of their own ident-
ity. They can talk with peers who understand the way they look at a prob-
lem, who appreciate the potential value of a half-baked idea, and who
know where the cutting-edge of the practice lies. With such a home base,
people can engage in a diversity of projects and in interactions with other
communities without becoming uprooted.

Trajectories. Identity extends in time. It is a trajectory in progress that

includes where you have been and where you are going, your history and
your aspirations. It brings the past and the future into the experience of
the present. Apprentices in traditional apprenticeship, for instance, are
not just learning skills, they are exposed to possible futures. By observing
and working with journeymen and masters, they develop a sense of tra-
jectory that expands their identity in time (Lave and Wenger, 1991).
Members of a community embody a set of paradigmatic trajectories that
provide material for newcomers to construct their own trajectory through
a community and beyond. In the generational encounter between new-
comers and established members, the identities of both get expanded.
Newcomers gain a sense of history. And old-timers gain perspective as
they revisit their own ways and open future possibilities for others
(Wenger, 1998).
A good way to develop identities is to open a set of trajectories that lead
to possible futures. The engagement of ones identity then incorporates
imagination and alignment: envisioning these possible futures and doing
what it takes to get there. These trajectories can be of various types.
Inbound trajectories invite newcomers into full membership in a com-
munity. Peripheral trajectories allow a person to interact with the com-
munity without making a commitment to becoming a full member.
Outbound trajectories, such as the ones offered by schools, point to forms
of participation outside the current communities.

Multimembership. Identity extends in space, across boundaries. It is neither

241
 Organization 7(2)
Articles

unitary nor fragmented. It is an experience of multimembership, an inter-

section of many relationships that you hold into the experience of being
a person, at once one and multiple. It is not something we can turn on and
off. When we go to work, we dont cease to be parents, and when we go
to the theater, we are still an engineer or a waitress. We bring these
aspects of our identity to bear to some extent in everything we do. Even
though certain aspects of our identities become more salient in different
circumstances, it would be an oversimplification to assume that we
merely have a multiplicity of separate identities. Such a view would over-
look the extent to which our various forms of membership can and do
conflict with, influence, complement, and enrich each other. The work
that we do in attempts to combine, confront, or reconcile various aspects
of our identities has a double effect. It is a source of personal growth. It is
also a source of social cohesion because it builds bridges across practices.
As a result, our identities shape the social structures we live in. The work
of identity constantly reshapes boundaries and reweaves the social fabric
of our learning systems.
Combining concurrent forms of membership in multiple communities
into ones experience is a way to expand an identity. Of course, we can
only combine core membership in a limited number of communities,
but we can also have more peripheral forms of participation, or even
transitory one, such as visits, sabbaticals, immersion, or one-time proj-
ects. Communities that can include in their forms of participation a
large portion of the multimembership of their members are more likely
to engage their whole identity. If I do not have to pretend that I am not
a parent when I am at work, I am more likely to put my heart into what
I do.

Fractals. Identity extends across levels. You are having dinner with your
family, ensconced in an intense discussion of international politics with
your teenagers, livingin the local context of the dinner tableyour
sense of identification with the global environmental movement.
Similarly, you may belong to a local church, but this belonging is usually
an expression of your belonging to a religion that includes many other
people in many other churches. Engaging at the local level of your church
is a way to belong at the broader level of your religion by combining such
engagement with imagination (you can picture many other churches with
people very much like you expressing similar beliefs, even though you
have never met them) and with alignment (in your church you follow rit-
uals that conform with liturgical formats adhered to by all other
churches). Note how the three modes of belonging complement each
other. Engagement is enriched by the awareness that others share the
same beliefs and follow the same guidelines. Conversely, imagining the
whole community and understanding the value of its rituals and norms
gains concreteness by the ability to engage in a local group.
Combining modes of belonging this way creates fractal layers of

242
 Communities of Practice
Etienne Wenger

belonging. More generally, if a community is large, it is a good idea to

structure it in layers, as a fractal of embedded subcommunities. If a com-
munity is large and does not have a fractal structure with local subcom-
munities in which people can engage actively, then it can easily happen
that beyond a small core group various segments of the community feel
disconnected. Subcommunities could be defined regionally as local
chapters of a global community. Some representatives of these local
communities then form a global community among them, whose purpose
is to connect the local subcommunities into one large global one. This is
how some global communities of well engineers have structured their
forms of participation at Shell Oil. Subcommunities could also be defined
by subspecialties as engineering communities are at DaimlerChrysler,
where engineers can join communities specialized in specific compo-
nents (e.g. wipers, seats, or dashboards) but clustered into broader com-
munities defined according to systems (e.g. body or powertrain). With
such a fractal structure, by belonging to your own subcommunity, you
experience in a local and direct way your belonging to a much broader
community.

Conclusion: Participation in Social Learning Systems

The perspective of a social learning system applies to many of our social
institutions: our disciplines, our industries, our economic regions, and
our organizations. This view has implications at multiple levels.
For individuals, this perspective highlights the importance of finding
the dynamic set of communities they should belong tocentrally and
peripherallyand to fashion a meaningful trajectory through these
communities over time.
For communities of practice, it requires a balance between core and
boundary processes, so that the practice is both a strong node in the web
of interconnectionsan enabler of deep learning in a specific area
and, at the same time, highly linked with other parts of the systema
player in systemwide processes of knowledge production, exchange,
and transformation.
For organizations, this perspective implies a need to learn to foster and
participate in social learning systems, both inside and outside organiz-
ational boundaries. Social learning systems are not defined by, congru-
ent with, or cleanly encompassed in organizations. Organizations can
take part in them; they can foster them; they can leverage them; but they
cannot fully own or control them.
This paradox could be bad news because the organizational requirements
of social learning systems often run counter to traditional management
practices (Wenger and Snyder, 2000). The currency of these systems is
collegiality, reciprocity, expertise, contributions to the practice, and nego-
tiating a learning agenda; not affiliation to an institution, assigned auth-

243
Organization 7(2)
Articles

ority, or commitment to a predefined deliverable. But there is also good

news. The knowledge economy will give more primacy to informal sys-
tems. In a traditional industrial setting, the formal design of a production
system is the primary source of value creation. Think of an assembly line
where value derives from the quality of the design of the formal process.
Informal processes still exist, but they produce value to the extent that
they conform to and serve the formal design. In the knowledge economy,
this relationship is inverted. The primary source of value creation lies in
informal processes, such as conversations, brainstorming, and pursuing
ideas. Formal organizational designs and processes are still important,
but they contribute to value creation to the extent that they are in the serv-
ice of informal processes.
This framework suggests two directions for organizations. On the one
hand, they must learn to manage themselves as social learning systems
and develop such systems internally. This means:

giving primacy to the kind of informal learning processes characteristic

of communities of practice and designing organizational structures and
processes that are in the service of the informal;
placing a lot of emphasis on the meaningfulness of participation in the
organization, on the possibility of building interesting identities, and on
community membership as the primary relationship to the organization
(Handy, 1989);
organizing for complexity, working to link the various communities that
constitute the learning systems in which the organization operates;
offering channels, shared discourses, processes, and technology plat-
forms by which local forms of knowledgeability can have global con-
nections and effects; and providing coordination among practices to
create complex knowledge beyond the purview of any practice.

On the other hand, organizations must learn to participate in broader

learning systems in which they are only one of many players. Companies
have learned to participate as one of many players in economic markets
to sell products and services to customers taken as individual decision-
makers. In the knowledge economy, however, they must learn to partici-
pate in learning systems as well. Knowledge production is becoming
more distributed, complex, and diversified, in disciplines and industries
(Gibbons et al., 1994); in regional economies such as Silicon Valley
(Saxenian, 1996); and among consumers who have the potential of form-
ing communities (Snyder, 1999).
In these learning systems, organizations find the talents they need, new
ideas, technological developments, best practices, and learning partners.
The rules of participation in social learning systems are different from
those of product markets. You dont simply compete; in fact, your most
threatening competitor may be your best partner when it comes to learn-
ing together. If you hoard your knowledge in a social learning system, you

244
 Communities of Practice
Etienne Wenger

quickly appear as taking more than you give, and you will progressively
be excluded from the most significant exchanges.
In a knowledge economy, sustained success for any organization will
depend not only on effective participation in economic markets, but, just
as importantly and with many of the same players, on knowing how to
participate in broader social learning systems.

References
Anderson, Benedict (1983) Imagined Communities. London: Verso.
Brown, John Seely and Duguid, Paul (1999) Organizing Knowledge, Reflections
1(2).
Eckert, Penelope (1989) Jocks and Burnouts: Social Categories and Identity in the
High School. New York: Teachers College Press.
Gibbons, Michael et al. (1994) The New Production of Knowledge: The Dynamics
of Science and Research in Contemporary Societies. London: Sage.
Giddens, Anthony (1991) Modernity and Self-identity: Self and Society in the Late
Modern Age. Stanford, CA: Stanford University Press.
Handy, Charles (1989) The Age of Unreason. Cambridge, MA: Harvard Business
School Press.
Lave, Jean and Wenger, Etienne (1991) Situated Learning: Legitimate Peripheral
Participation. New York: Cambridge University Press.
McDermott, Richard (1999) Learning across Teams: How to Build Communities
of Practice in Team-based Organizations, Knowledge Management Review 8
(May/June): 326.
Saxenian, Annalee (1996) Regional Advantage: Culture and Competition in
Silicon Valley and Route 128. Cambridge, MA: Harvard University Press.
Snyder, William (1999) Organization and World Design: The Gaias Hypotheses.
Submitted to the Academy of Management, Division of Organizations and the
Natural Environment, John M. Jermier, College of Business, University of South
Florida, Tampa, FL.
Star, Susan Leigh and Griesemer, J. (1989) Institutional Ecology, Translation,
and Boundary Objects: Amateurs and Professionals in Berkeleys Museum of
Vertebrate Zoology, 19071939, Social Studies of Science 19: 387420.
Wenger, Etienne (1998) Communities of Practice: Learning, Meaning, and
Identity. New York: Cambridge University Press.
Wenger, Etienne and Snyder, William (2000) Communities of Practice? The
Organizational Frontier, Harvard Business Review JanuaryFebruary.

Etienne Wenger is a globally recognized thinker in the field of learning theory and its
application to business. He is a pioneer of the communities of practice research,
and is the author of numerous articles and three books. After working as a teacher
for many years, he was awarded a PhD in artificial intelligence from the
University of California at Irvine, and joined the Institute for Research on
Learning, where he developed a new learning theory centered on the concept of
community of practice. He is now an independent consultant, researcher, author,

245
Organization 7(2)
Articles

and speaker. He works with people interested in developing new kinds of organ-
izational, technological, and educational designs that leverage the synergy
between learning and community. Address: PO Box 810, North San Juan, CA
95960, USA. [email: [email protected]]

246