Going Virtual: Distributed Communities in Practice - page 28


These two CoPs (ESC and EE-AW) were different from WWITMan in that they were more distributed and the practice was extra to the everyday work. However, it was interesting to note that all three CoPs exhibited very similar characteristics, especially in respect of the active cores, how proximity and face-to-face interaction further enable participation, and the importance of a shared artefact. The process of creating and using the artefact was emphasised, and this was further illustrated by the fact the artefact was being designed to be used as a cross-boundary communication tool, not by itself but with a CoP member, thus attempting to retain a participation/reification balance.

Chapter VIII: Lessons Learned

In this final chapter we summarise briefly our journey from KM to our CoP and draw out lessons that can be learned from the issues that have arisen and the insights we have gained. We also consider what practical implications they might have.


In this book we have explored the inner workings of a distributed international Community of Practice (CoP). In order to reach that point we first explored the field of Knowledge Management (KM) to set the context for the study. In doing so, we noted that a large part of KM focuses on the capturecodify-store approach and did not consider less structured knowledge that does not lend itself so well to this cycle. We did, however, note that less structured knowledge is receiving increasingly more attention, but the predominant approach to its "management" appears to be to try to convert it to a form where it can be captured.

Feeling that perhaps a different approach was required, we used the terms "soft" and "hard" knowledge and explored the notion of soft knowledge in some detail, as the management of hard knowledge is already well established. This led to the conclusions that dealing with soft knowledge in isolation is problematic and that knowledge should be viewed as a duality, with all knowledge being both soft and hard, with the balance varying. Harder aspects of knowledge would be easy to capture and codify while the softer aspects would be difficult (or impossible) to capture.

Observing that CoPs are groups where the softer aspects of knowledge are created, nurtured, and sustained, we refined our definition of a CoP, noting that shared artefacts might be an avenue worthy of exploration and embarked on our study to explore how a CoP functions in a distributed international environment. In doing so, a number of issues came to the surface, and we gained some valuable insights. But what can we learn from them?


The "Sharing" of the Softer Side of Knowledge

In Chapter III we explored the softer aspects of knowledge from three different viewpoints, Distributed Cognition and boundary objects, Common Ground, and CoPs. Examples of all three viewpoints were in evidence in the case study, suggesting that the answer as to how the softer side of knowledge may be shared lies somewhere between the three. However, the main viewpoint that was supported was Wenger's (1998) reification/participation duality.

One of the questions that was raised in Chapter IV regarded the form that the artefact might take; that is, what might be in such an artefact? In the study, we saw a strong example of an artefact and how it was used. We were also interested in the possible proportions of the reification/participation duality. It was suggested that in a distributed environment sustaining participation may be more difficult and therefore reification would play a greater role. In the study, we saw that this was not necessarily the case. Artefacts were strongly in evidence and were more important than the groups had realised — they played a more central role than was expected and played a variety of roles that the CoP members had not previously recognised. Participation was shown to still be important, perhaps as necessary as in co-located groups. This was shown by the efforts made by the members to facilitate participation by designing the artefact for participation, by arranging regular face-to-face visits, and even, to some degree, formalising participation by making collaboration with peers in other locations one of the group's stated objectives. The case study demonstrated how maintaining participation might be achieved. We will discuss the issues of participation and reification separately. We saw in Stage One of the case study that the two most important aspects had been shown to be:

  1. the use of a shared artefact, and

  2. the development of strong relationships.

We also saw in Figure 14 that these mapped onto Wenger's duality. In Stage Two and the other two CoPs, we gained further insights as to how this works and how it overcomes the problems of working in a distributed environment.


The duality was very visible in the main case study and was seen initially in examples of shared artefacts such as stories, procedures, documents, and slide presentations. The artefact that provided the best example was the planning document. It was particularly noticeable that this artefact played many roles over and above those for which it was originally intended. A key role played by the planning document was its role as a catalyst. It had stimulative effects, stimulating debate, raising issues and problems for discussion and solution, and acting as a catalyst for collaboration. It acted as a catalyst (as opposed to a vehicle) for the group members to apply their domain and soft knowledge for planning, for reflection, for discussion of issues, and for solving problems. The shared document was not essential to their work, but it played more roles more importantly than they had previously realised. This particular planning document has undergone a further iteration and still plays a major role in their work.

The other roles played by the planning document led to the most interesting insight into the work of this community. It was not the literal meaning of the planning document that was important but rather its role as a facilitator of participation. Furthermore, people were designing it with this role in mind as opposed to its literal role. They were designing a planning document but were using the softer aspects of their knowledge of the community to design the artefact to do something else (which had little connection with planning). This also sheds further light on the question raised in Chapter III as to why the softer aspects of knowledge cannot simply be made hard. It was shown in Chapter III that the Distributed Cognition view of knowledge is predominantly hard and

suggests that softer knowledge could be made hard. The Distributed Cognition interest, therefore, would be in the artefact itself, and it would look for knowledge embedded in the artefact. The example of the planning document shows the soft knowledge that is going on around the artefact in order to make it work. Wenger's (1998) view suggests that the participation aspect of the duality is essential and has been overlooked in Distributed Cognition. The

findings of the main case study support Wenger's view very strongly: it is not the artefact (reification) per se that is important for the community. The members learn and develop knowledge with each other by the process (participation) of creating the artefact and working with it. The process provides a link between reification and participation — it provides a means of participation, and it is through this that members are able to grow their knowledge. This was exemplified in the CoP by the previous unsuccessful attempt to create a planning document. In the current situation, the development of the document was ongoing, allowing continuous participation in the process. The document thus provided a concrete example of the essentiality of considering both sides of the duality and demonstrates how members of a CoP can use documents to make the CoP work.

The planning document also demonstrates how the traditional KM view fails in only viewing the harder aspects of knowledge. The document is a planning document and looks like a planning document, but it is used as something else. When considering the document alone, full understanding of the document would not be possible; it would simply appear to be a planning document — the participation is also essential. If it was observed from a traditional KM viewpoint, a practitioner would consider it easy to reproduce, missing the point that both the reification and participation are necessary


Even though the CoPs were distributed, it was clear that participation was as important, if not more so than in co-located CoPs. The difficulty is in how to achieve and maintain an adequate level of participation. This was illustrated by the fact that the time distance was perceived to be more of a problem than physical distance, showing that interactivity was more important to the members than visibility. The interactivity contributes perhaps to a greater feeling of participation.

Sustaining the participation is an area where the social issues in a CoP play a major role. It was through the social issues that the distributed CoPs managed to maintain the participation element.

It was clear that the motivation to work as a CoP must be present. The members must have that desire, motivation, and will to work towards the common purpose, to learn, and to share. The motivation is one of the factors that sustains the group through the periods of e-communication; that is, it is easier to maintain participation in an electronic environment if the members are motivated to do so. If the motivation is internal, it is even stronger.

All members of the CoPs who were involved in the study emphasised the importance of developing strong working relationships and the importance of face-to-face communication in developing these relationships. Having shared interests (as well as the interest in the community) helped develop the strong relationships, as did regular frequent interaction. During the periods of ecommunication, it was found in WWITMan that the relationship can decay somewhat, but when the relationship developed to a “comfort level' where the relationship reached more of a "friendship"' footing, then the participants were more ready to indulge in regular and frequent communication. The development of a strong relationship meant the members were more likely to maintain the interaction and participation when working in a distributed environment.

Relationships are extremely important to a CoP. Not only do they aid participation, but legitimation in a CoP comes about through the development of relationships. This helps in the development of identity, confidence, and trust in the CoP. This is shown in the "war stories": war stories were a part of all three of the CoPs in the case study, and Orr (1997) talked of how war stories "amuse, instruct and celebrate the tellers' identity as technicians" (emphasis added). When the war stories are taken out of context, the identity is lost and hearers cannot have the confidence in them that they would otherwise have had. This provides further support to the view expressed in Chapter III that softer knowledge cannot simply be made hard — each side of the duality has no meaning without the other.

In the earlier stages of the development of a relationship, when regular, frequent interaction is not so high on a person's agenda, having a task focus can help. If people have a project or a task on which they must collaborate, then they will have to be in more frequent contact. If the motivation is there, too, then the relationship can start to develop.

The task focus can also help when local issues get in the way of the joint work, that is, when people are so busy with local issues that they have little time for considering CoP members who are in other locations (Figure 1).

click to expand
Figure 1: Overcoming Difficulties

The planning document itself provided an excellent example of the need for participation. It became clear that the artefact in itself was not as important as the process of creating it and working with it. People could share in the experience. It provided participation. This experience and what is learned from it are something that cannot be captured, codified, or stored.

Having a task focus can help increase the interaction. It also provides an area for participation. As people work together, the relationship evolves and is hastened by face-to-face interaction where people grow existing relationships and try to extend the web of relationships to include new people. The task focus can help sustain the relationships through e-communication, and face-toface interaction can boost the relationship, but fundamental to all of this is the motivation that must be present. The CoP members must have the desire for the CoP to function. This shows the importance of the human, or social, aspects (Figure 2), which must not be forgot.

click to expand
Figure 2: The Importance of the Human Aspect

The importance of developing the strong relationships to the "comfort zone," that is, to nearer a friendship footing, shows that the inner workings of a CoP are about friendship. The case studies have shown that the capturecodify-store approach is not relevant to a CoP — what we can see at the heart of the CoP is friendship and trust. This has implications for KM, for it shows that it is no longer about managing knowledge as such, but about the human aspect of this process. It shows a need to put people back into the KM agenda, a turn to the social rather than the cognitive or, as Nardi and O'Day (1999) would express it, putting the "heart back into work." Humans are, after all, cleverer than computers — humans have the softer aspects of knowledge, too.

Completing the Duality

Taking the diagram in Figure 2, we can add in soft and hard knowledge to complete the diagram as shown in Figure 3.

click to expand
Figure 3: Participation/Reification Mapped to Soft/Hard

Figure 3 is intended to show that knowledge as a soft/hard duality (as described in Chapter III) maps onto the participation/reification duality. Reification, in the form of the shared artefact, needs the social aspects, too, in order to share the softer knowledge. The degree of softer knowledge depends on the degree of participation or social aspects. If the emphasis is on reification then there will be more of an emphasis on hard knowledge. The important part is to maintain a correct balance for the community.

Are We Really Sharing the Softer Aspects of Knowledge?

The findings from the case study suggested that the softer aspects of knowledge are not so much shared as created, generated, and sustained. The shared artefact in the form of the planning document did not serve as a vehicle for sharing softer knowledge but as a focus for the application and creation of knowledge. It had, for example, Dave's planning expertise and Y2K experience embedded in it. This knowledge was not shared through the planning document, but in working with the planning document, other members had a focus to which they could apply their knowledge. This could also be seen in the different ways the artefact functioned as a catalyst to interaction and participation. It sparked discussion of problems and issues and the members of the community learned from their participation in this process. The planning document also highlighted opportunities for collaboration that provided further opportunity for learning and for generation of new knowledge. This perhaps indicates to us that the softer aspects really aren't shared. It is more the case that people learn from one another and are given the opportunity to develop their own softer knowledge. The parts that are shared are the harder aspects, and the softer aspects are developed through interaction, situated learning, and experience.

Supporting CoPs

What we have seen in the study has supported Wenger's use of the participation/reification duality, as it mapped very well to the new environment. In doing so, we applied the theoretical duality and demonstrated in practical terms how a CoP would map this to its own practice. We have looked closely at the inner workings of an internationally distributed CoP, providing an insight into how such a CoP works. The study has shown that this is essentially about friendship and other social issues rather than specifically about knowledge. This raises the question as to how such things are managed. Clearly, they are not managed in the way KM as a discipline thinks of management. What is seen when we look inside the CoP is how central participants nurture these friendships. This is essentially a human activity at the level of practice, not at the level of the organisation or technology.


It would appear that distributed CoPs in organisations are different structurally from the "virtual communities" that are found on the Internet, but such "virtual communities" are not the focus of this book. The focus of this book is on CoPs that exist in commercial organisations. The CoPs in the studies showed a major difference between the virtual Internet communities and the distributed organisational communities: one example of a virtual Internet community that was described as a CoP was a community that had formed round a MUD (Conkar & Kimble, 1997). The difference between such a CoP and those that were the focus of this book was that, in the case of Conkar and Kimble (1997), the medium was the practice; that is, the community existed only to use the medium. In the organisational CoPs, the practice of the CoPs is mediated by the medium. The case studies demonstrated a noticeable difference in structure between the two types. Whereas the Internet community was "virtual," that is, totally distributed, the organisational CoPs that were encountered were not virtual, but had "active cores" which, in the case studies, were also co-located. Although the initial aim was to explore the knowledge sharing in a distributed Community of Practice, the studies showed an interesting point regarding the structure and development of the groups. From this we can hypothesise that the development or evolution follows the lines:

  1. Communities of Practice seem to evolve, either from nothing or from an official grouping, as a result of the way the members work.

  2. The Community of Practice may then create a link with other people at other locations who do similar work. These people will possibly be members of other Communities of Practice.

  3. This situation can develop even further, in that the Community of Practice might create links with a group of others in another area, possibly abroad, who are involved in similar work and who also function as a Community of Practice. We are then left with a possible situation as shown in Figure 4.

    click to expand
    Figure 4: Community Structure and Evolution

Figure 4 shows the links between a Community of Practice and other individuals who then may become members of the community but not be colocated. It also shows that members may be members of other Communities of Practice and that there may be links developing between Communities of Practice. To some extent this mirrors the networks of organisations developing in the globalised environment, as described by Castells (1996).

These stages of development show us some differences between a Community of Practice in the real world and one in a VE. It also shows that the structure of the CoP in the distributed environment is not totally distributed — there are co-located cores.

As the organisational CoPs had cores with links to other members (or other cores — see Figure 4), it extends the notion of peripherality. The peripherality in Lave and Wenger's (1991) groups was more a social periphery, as newcomers worked on the periphery and gradually moved to full participation. It was couched in terms of legitimacy of membership and practice. While this is undoubtedly true in the distributed group we saw in the study, physical and temporal peripherality were also factors that mediated the practice and the quality of interaction. In particular, physical and temporal peripherality provided further barriers to the functioning of the CoP in that they provided barriers to access; for example, the Japanese member did not have as much access to other CoP members as she would have liked. These barriers were tackled by the big effort that was put into sustaining the CoP — this involved a lot of work in designing the boundary object (the planning document), having meetings about meetings, and arranging the technology.

A further key difference is in the terms of identity. Identity is a central issue for the CoP as described by Wenger (1998) and is developed through participation and engagement. It is more than merely thinking in terms of being a member of a community (as, perhaps, is more the case in the fully virtual community); rather, it is constant negotiation of "ways of being" in the context of that CoP. Therefore, it can be seen how important it is that the members of the CoP work so hard to maintain participation, even going to the lengths of working hard to arrange the face-to-face visits.

Practical Issues

The members of WWITMan were experiencing a number of problems that affect teams, groups, and communities that are operating in a distributed international environment. These problems were the usual ones of differences in culture and language, time and distance. They also encountered the problem of relationship decay. As they met, they developed a relationship, but during

the period of e-communication, they found they experienced a degree of relationship decay. Organisations supporting distributed CoPs would need to consider the following points, which are practical applications of the participation/reification duality as expressed in the terms of social factors and shared artefacts:

  1. First, the participants must be motivated and have the desire to make the community work. In addition to this, it is very helpful if the CoP exists in an organisational culture that encourages sharing, collaboration, and communication. If the culture is wrong, the CoP will struggle.

  2. Strong relationships are essential, but it is important to let them develop and evolve. Working too hard to make relationships will probably fail.

  3. An initial or early face-to-face meeting will probably give the relationship development a good start. In some cases, organisations have run an initial conference from which CoPs have developed. The first face-to-face does not need to be as grand as a conference, however. Simply getting likeminded and interested people together for a seminar or workshop would be a good start. Initially, there might need to be some structure to the discussions. It might also be necessary to continue with some structure, but the importance of the relationship development underlines the need for some unstructured social contact, too.

  4. Some occasional face-to-face meetings (if possible) will keep boosting the developing relationships and will help them move further, faster to enable the participants to reach the "comfort zone" more quickly.

  5. Regular, frequent interaction is important. This will come more easily once the comfort zone has been reached, and in the early stages will be encouraged by having a task focus.

  6. Having a task focus keeps people interacting, increases the frequency of communication, and provides the participants with something in common.

  7. It is important to facilitate the community's existence by the provision of ICT, but the following points regarding ICT are important considerations:

    • Asynchronous and real-time communication should both be supported.

    • A shared space for sharing artefacts during remote meetings or collaboration is very important.

    • An intranet can help with the "harder" side; that is, it can provide shared storage of documents, images, and other files that are used by the CoP. It can also provide discussion forums and message boards.

    • The media should be simple to use (that is, the medium should not obstruct the message). This includes video conferencing, which does not necessarily have to be a full videoconference suite. Cheap cameras with lower quality images can provide context and be perfectly adequate. They are also easier and cheaper to use than a videoconference suite, which also has to be booked in advance.

    • There should be some functionality to help community members be aware of the presence of others so that remote ad hoc communication can be made easier.

    • There should be a choice of media so that people can use the medium they feel most suitable for the task in context — this will be the medium they feel is most appropriate for the task, depending upon the sensitivity or the urgency of the message. Choice will also be affected by context; for example, the medium with which the user feels most comfortable, the time of day, the likelihood of the communication partner being available, and the number of intended recipients.

Software to support communities is now appearing on the market; however, it is important that, when appraising any community software, the principle behind the software should be "connection not collection."

The communities in the case study all found it helpful to have a shared artefact. This can be a range of things: documents, stories, procedures, slide shows, etc. The nature of the artefact will depend to some extent on the practice of the community. The most common ones seen during the case study were a planning document and slide shows, and all were used for communicating across boundaries. It is worth remembering that the shared artefacts were most effective in crossing boundaries when a community member was involved. The process of creating and working with the shared artefact is most important — more so than the artefact itself.

  • Working on the artefact together can be part of the group's task focus.

  • Working with it helps foster regular frequent communication.

  • The process of working together on an artefact may stimulate collaboration in other areas.

If the CoP is distributed, it is also important to use the softer aspects of the members' knowledge, to recognise that the artefact will function as a boundary object and to bear this in mind during the design of the artefact. The members of WWITMan considered themselves to be part of a single CoP spread over a distance. The boundaries that the planning document had to cross were not, therefore, CoP boundaries but physical and cultural ones. It crossed boundaries within the CoP. However, participation, in the form of the relationships that had been developed, assisted the members in designing the document to function as a boundary object. This also exemplifies the softer aspects of knowledge — it is what the members know about the document and about the members in the other core that they use in designing the document to ensure participation. These are the softer aspects of knowledge.

The shared artefacts in the case studies were particularly interesting in that they played different roles from those for which they were primarily intended. In particular, they acted as catalysts for collaboration. This in turn encourages more interaction and participation and helps the relationship develop further.

The Last Word

As KM practitioners have gradually come to recognise the importance of unstructured knowledge, CoPs have begun to receive attention, and there have been a number of attempts to "implement" them. Some practitioners have tried to formalise them — in some cases, organisations have been known to allocate members to a group and inform them that they are a Community of Practice with a specific brief. Further difficulties are encountered if the CoP has to operate in a distributed environment. These difficulties are further exacerbated when the distribution crosses national boundaries. However, as CoPs become increasingly recognised as important, and business becomes more global, the need for international CoPs will grow. The insights gained from WWITMan, ESC, and EE-AW should help practitioners become aware of the problems and issues involved, but more importantly, they show that practitioners need to change their views of the organisation when planning KM initiatives:

Redundancy of Knowledge When planning outsourcing and downsizing, redundancy of knowledge was regarded as negative. The error of this view was seen when important knowledge was lost as people left the organisation. In a CoP, members learn from each other, and there is some sort of knowledge redundancy "built in" to the social distribution of knowledge. This helps maintain the knowledge base in the CoP as members leave and provides a buffer against the problem of knowledge loss. This was seen in WWIT when the members deliberately placed people in their teams together to learn from each other and when the Lotus Notes "guru" was going on holiday.

A Different View of the Organisation The time spent with WWITMan showed us the importance of CoP members "going the extra mile" for each other. This refers to the social relations, the development of member identity, trust and confidence. It could be seen in the way the CoP members worked hard to sustain the community. They organised trips and documents in order to organise and develop their social relations. Practitioners therefore need to explore where the social relationships are and how they can be supported rather than looking at the organisation in terms of where the information is and how it flows (which would be the "traditional" approach). This also applies to document design. It is important to recognise the importance of what goes on around a document during its design and use. When analyzing documents in use, it is not only the document itself that is important, but also the softer aspects of the knowledge that has been applied to it. Widening this view further, it can also be applied to the company itself. The company can be seen as an adaptive and self-generative organisation. The KM practitioner, therefore, when planning a KM initiative needs to look beyond the organisation as stores and flows of information (as would be the case in a capture/codify/store approach) and explore the social networks that exist in the organisation. (S)he needs to concentrate on assisting groups and communities in developing and sustaining these strong relationships if the whole knowledge is to be created, nurtured, and retained.

The importance of social issues, shared artefacts, and the process of working with shared artefacts gives us a practical example of the importance of Wenger's (1998) participation/reification duality. We can also see that it is possibly not exactly a sharing of the softer aspects of knowledge as much as a nurturing and creating. It is clear that the CoP members work hard at the management of softer knowledge — of prime importance is the human work, as opposed to the capturing, codifying, and storage ability of computers. The importance of computers is their support of the human work. The importance of the softer aspects, the participation, suggests that the traditional KM view has been too restricted and that the rush to implement technological solutions follows a false trail without consideration of the softer aspects of knowledge.