WHAT ISSUES DID WE FIND IN STAGE ONE?


WHAT ISSUES DID WE FIND IN STAGE ONE?

The case study was intended to provide some understanding of the work of a physically distributed CoP, and how its members make it work. This aspect was demonstrated very clearly in the vignettes—they were all about making the CoP work. A major point to emerge from the vignettes was the use of documents to make the CoP work. Other vignettes showed technology and how it was used to sustain the working of the CoP, and others indicated what individuals know about other communities and other members that helps put the right people in contact for participation.

A number of themes and issues came to light that provided a comprehensive insight into the workings of the CoP and the interactions between the members.

The CoP

As a distributed CoP, it has an interesting structure. That is, it is not totally distributed—it has co-located cores and has an individual on the periphery. It would appear that the co-located core of a distributed CoP is essential, for a lot of work was undertaken in the co-located core to ensure the smooth operation of the CoP across the distributed environment.

The practice of the community is their normal work. However, the group was not formally created, it came about through evolution.

In the first place, it is interesting to explore the CoP characteristics (as outlined in Chapter III) that were exhibited by WWITMan.

Marginals/Brokers—Knowing the Practices of Other Communities

Star and Griesemer (1989) referred to "marginals" as community members who are members of different communities. Wenger (1998) used the term "brokers" to describe these people. He saw their role as helping the flow of information between communities. Brokers will have a knowledge of the customs, language, and relationships in the different communities and are therefore in a position to assist with mutual comprehension. This could be important for the management of the softer aspects of knowledge; for example, in the cases of boundary objects, a marginal who is a member of more than one community can help when the artefact is being used in the context of another community.

In WWITMan the members were all members of a range of different groups. In addition to being members of WWITMan, their own vertical teams, and WWIT, they were involved in different project groups, some of which may well be CoPs themselves. There were also some CoPs, informal groups that had arisen because of common interests to which the WWITMan members belonged. Stan, in particular, was a member of several different groups; for example the "Internal Home Page Group" and a library project group. In some of these cases, the members learned things in one group that they could use for the benefit of another group. For example, during a visit to Palo Alto, Stan had met informally with a colleague in one of his project groups and found that his colleague was involved in a different group that was developing an application that Stan felt would benefit another of his groups.

Another example of brokering in the case study was to be seen in the way the planning document was used. In creating the document, the UK core used its knowledge of the relationships between how one community uses something and how another does to create the document in a format that would be most suitable for use by the two cores.

Gurus/Wizards

Gurus and wizards in a group are those members who are regarded as "experts" in a particular domain of the practice. They will have a high degree of technical knowledge but will also have developed softer aspects of their domain knowledge, for example, a "feel" for how things work, an intuition when troubleshooting. Other members may learn from working alongside the guru; however, part of the softer aspects of a CoP member's knowledge is simply knowing who knows what and who can be approached for help with different things. When a guru leaves the organisation, it might be possible to replace him/her with a new, technically competent staff member, but the softer aspects of the new member's knowledge will be different.

Gurus/wizards were very much in evidence in the case study; however, it was not so much an exclusive CoP characteristic but more a result of the role of the group. Although the group is a CoP, it does fulfill an official role, and the WWITMan members are specialists in their field. Moving into the wider UKIT group, there were several individuals who had developed a high degree of expertise in a specific area and who were then consulted by other members of the group; for example, Steve was the Lotus Notes expert. This situation was also evident in the USIT part of WWIT. Although WWIT as a whole was not considered to be a CoP, there was an increasing amount of collaboration being undertaken and wizards in one part would assist people in the other part; for example, during one of the e-meetings a problem was encountered, and the Palo Alto people immediately suggested that they had a local expert who could solve it.

An interesting issue that was demonstrated was how the CoP would survive when a guru was missing, and how the CoP dealt with the loss of knowledge on a day-to-day basis. Steve, the Lotus Notes expert, was going on holiday, and UKIT were concerned as to how it might manage while he was not available. The conclusion reached by the UKIT staff was to (a) be briefed by Steve who would provide as much of the harder aspects as possible, so they could maintain the system to a certain degree, and (b) to liaise with USIT as it had more expertise available there. The people in USIT would assist the UKIT people as necessary, where the harder aspects of the knowledge were not sufficient. This was also demonstrated when the UKIT managers were concerned that they might be losing a particularly capable member of UKIT, Stewart. They took a number of steps to ensure that they kept him, but they also identified another member of staff who they wanted to work closely with Stewart. Thus they provided the staff member with access to Stewart so that softer aspects of his knowledge would be shared during general work.

Problem Solving

In a CoP if a problem arises, a key weapon in the CoPs armory is to be found among the softer aspects of the members' knowledge, that is, knowing who has the knowledge to help and being able to obtain the participation of the people with the knowledge.

Within the wider UKIT group and WWIT as a whole, there was clearly a lot of problem solving as the group is an IT-support operation. However, there were also a lot of problems to be solved in addition to these. In some cases, during discussion of the planning document, problems arose or were foreseen. The CoP was either able to solve the problem immediately through the discussion, identify an individual with the relevant experience to solve the problem, or could organise some people to tackle the problem; for example, in the main e-meeting, Stan raised a problem with multi-media publishing tools and found that the US core had encountered a similar problem and solved it. They now had the expertise that was willingly made available to Stan.

Problems were solved within the CoP—both within the two cores and by one asking for help from the other core. If an individual had a problem to solve, (s)he had a good idea of who to ask. The members of the CoP learned from each other through the problem solving. Problem solving within the cores tended to be a result of ad hoc chat, that is, opportunistic communication or informal communication, where if an individual had a problem, (s)he would call a colleague to ask for advice. There were occasional examples of problem solving via e-mail, but in fact one of the main problems (accessing a supposedly shared directory) was not finally solved until the participants were in the emeeting and could demonstrate the problem in real time. This also showed the importance of speed of response in the interchange for some problems. Stan had tried to solve the problem of the shared directory access for several days, over e-mail and voice mail interchange. However, it was not until they were in an e-meeting, and he could say, "Here, let me show you," and actually demonstrate the problem on the screen, that the problem was finally solved.

A shared background and a shared language are part of a CoP and this was useful in solving problems, as people could discuss the problem and work up a solution.

Language

Part of the definition of a CoP is the language shared by its members. The language is an essential part of Common Ground, where some of the softer aspects of a CoP's knowledge are to be found. This was evident in the language of WWITMan, as the members used the technical domain language and acronyms of the wider community of IT professionals (for example, UNIX, NT). In addition there was localised language that would have no meaning to an outsider, (such as organisation-specific acronyms, like "PRs" for Personal Reviews, and group-specific terms such as the names of the servers.

The technical domain language could be seen as being "hard." The softer aspects, then, would be found in the ways the CoP members speak which underlines their attitudes and impressions. For example, when one says, "Is it one of those meetings?," this would only be understandable to another established CoP member. A CoP member might therefore be identified as a member of the CoP by the way (s)he speaks.

Legitimate Peripheral Participation

The participation component was very much in evidence in the importance of the social issues. As Wenger (1998) had identified participation as being the key LPP component to be part of the reification/participation duality, this will be discussed separately along with the importance of the shared artefact, which represents the reification part of the duality. However, as well as being the key to participation, the social issues were also key to legitimation. The legitimation aspect was affected to some degree by the fact that there is a formal aspect to the group with a manager both in UKIT and USIT, and an overall WWITMan manager. However, the members of the CoP did not feel that a CoP is formally created. The members of WWITMan are automatically assumed to have a high degree of relevant knowledge by virtue of the positions they hold, but there was also something more than this. The members of the USIT management team had developed very strong working relationships, and this was key to the legitimation within the group. In a formal group such as a team, virtual team, or project group, the legitimation of the members comes from the formal structure of the group. In a CoP, the legitimation comes from the social relationships that develop. As people get to know each other, they have confidence in the information and knowledge they receive from their partners. This shows the social issues of a CoP to be of major importance, and therefore does not preclude a formally constituted group or team from developing or evolving into a CoP as the members develop relationships, get to know each other well, and go beyond the formal relationships. The members of the CoP felt that key to making the CoP work, particularly when faced with the difficulties caused by distributed working, was the fact that they had developed strong relationships to the point where they would "go the extra mile" for each other. This shows that the essential factor that differentiates a CoP from a team is the "human aspect," that is, the social relationships that are formed in a CoP.

In the case study, there was evidence of a physical peripherality, as there were two co-located cores with a member situated elsewhere in Japan. This member was accepted as a member of the group but did not feature so much in the meetings because of the time difference. She was kept informed of plans and progress, but she was not able to play as full a part as other members. If she wanted to take part in an electronic conference, she had to participate in the middle of the night. She was regarded as being a full member of the group, and the other members had every confidence in her ability, but the physical and temporal distance meant she was, in some ways, a peripheral member. The physical periphery also restricted her real-time access to other members of the CoP, thus reducing opportunities for learning from them.

Learning

Although learning is a key aspect of CoPs, the learning that took place in WWITMan was not a formally stated goal. There were many examples of learning, but where it occurred, it tended to be a by-product. It occurred both within the CoP (WWITMan) and within the wider WWIT group.

Learning could be seen where members of the CoP were keen to leverage from each other. In this case, the knowledge existed in one person who could then help the other who would learn from the exchange. However, the prime aim here was not learning but the avoidance of re-inventing the wheel, which in itself is an important part of KM. This worked as well as it did partly because of the sharing culture of the organisation, and also because of the relationships that had developed between the members. Leverage occurred often through collaboration, which was an important part of the work of WWIT. During the e-meeting that was focused on the planning document, a primary aim was to identify areas where the UKIT and USIT could collaborate and leverage from each other. This was quite common and was regarded as standard practice within the CoP, and was being extended to the wider WWIT group. There was one example where this had not worked, but this was outside of the CoP and the wider WWIT group. Stan's Informatics Team had developed a booking system for audio-visual (AV) resources. This was proving to be a very useful tool in UKIT's armory and was being linked to the room-booking system they had developed. The goal was to be able to book the necessary room and resources anywhere in the organisation. For example, if the UK core was going to visit the US core, Wayne would not have to ask his opposite number to book resources and rooms for him at the US site. From his desk, he would be able to book all the rooms and resources for the visit, in advance. Unfortunately, towards the end to the study Wayne received an e-mail saying that a booking system had been developed in the company, and it was to be adopted as standard throughout the organisation.

Learning also occurred in the collaboration. In this case it was less a case of one person learning from another. Rather, as people worked together on a project they would both learn from the process of what they were doing and gain from experience. The collaboration might be the implementation of some technology where they would gain from the experience or it might be solving a problem, in which case the people would apply their knowledge and experience to the problem and among them would work up a solution. Although the people involved in the collaboration learned and gained further experience, it was not evident that the opportunity was taken to provide feedback to other members of the group. It was more likely that any further sharing of the knowledge would be in a serendipitous manner. Knowing who might have a specific type of experience and who could then be consulted on it would be part of the group's stock of knowledge. In many cases, an important part of the CoP's stock of knowledge was simply knowing who to ask for help. This extended beyond the CoP to the wider WWIT group, as it might be necessary to consult an expert in the other country.

There were two examples in the case study of distributed situated learning using an artefact. In one case, people from Palo Alto demonstrated the Knowledge Base to three people in the UK using NetMeeting. They demonstrated its functionality and took them through how to use it. The other example was in the main e-meeting. The team in Palo Alto was having difficulties setting up the smartboard, and Mike took them through it stepb-y-step.

It was assumed that people within the group would already have a high degree of technical domain knowledge and therefore some sort of shared background. They would also therefore be able to speak the necessary technical language as a pre-requisite to further learning from each other. This was also necessary for legitimation—they each had to have confidence and trust in each other's abilities. Also as a result of the existing level of domain knowledge, much of what people had to learn would not necessarily be of a technical nature, but would be things of a softer nature—how to get things done, who to approach, how to function as a CoP, how things are done, etc. This social learning, however, was not encountered, that is, where members become enculturated in the ways and language of the CoP. The reason for this was simply a problem of time scale. The case study was under strict time limitations and was not long enough to be able to see such learning.

The learning within the group tends to be informal and ad hoc. This is part of the culture of the organisation and reflects much of the communication within the group. A lot of what the group does is serendipitous. However, the group is aware of the importance of learning and is moving towards more explicit support. This is taking place in the wider group (UKIT and WWIT) as opposed to within WWITMan. For example, the members of WWITMan are now trying to consciously get UKIT people to learn from others—they foresaw a problem occurring in the future and put one person (Charles) together with someone (Stewart) who was fully experienced in the field with the aim that Charles would learn from Stewart. WWITMan were also trying to spread this practice to WWIT as they made it an objective to encourage collaboration between members of their teams and members of the corresponding teams in Palo Alto. This was simply a means of trying to support, facilitate, and enable the sharing of knowledge so that if one key person left there would be somebody who was able, to some degree at least, to take over for him or her. The harder aspects of the knowledge could have been written down in procedures, but the CoP preferred to put people together or encourage people to get together so that the softer aspects of knowledge can also be developed.

Being aware of the need for learning, WWITMan could see that there was scope for more learning within the group. In particular, they felt they were not good at learning from the past. In this respect they were probably being too harsh on themselves, as there were several examples of where they had done exactly this. One example was when members of WWITMan were discussing a resourcing issue. This was causing them some problems until the discussion led them into territory that was a similar situation to one they had faced some years previously. Dave remembered this and remembered how they had tackled the situation then. This insight moved them on faster towards a solution this time round. The problem with this is that it is the softer side of knowledge, remembered through experience and triggered by context, and if there had not been someone there who remembered the previous time, then the knowledge would have been lost, and they would have had to go through the whole problem-solving process again. Another example of learning from the past was the construction of the planning document. The CoP members applied their knowledge to the content of the document and learned from participating in the process. Already by the second iteration of the document they felt they had done a better job than the previous year. It was not the document itself that was important, but the process of creating it. The document was not the end product, but became an ongoing, "living" document. Although it was the process from which the CoP learned, at least there was something concrete that came out of it that would serve as a reference— the planning document itself that embodied part of the knowledge of the CoP.

Formal Attempts to Share Knowledge

Knowledge was shared in a variety of ways in the case study. The harder aspects of knowledge—those easily articulable—were shared consciously and deliberately, primarily by instructional means. Documents were circulated, seminars were held, information was exchanged in meetings, and information was placed on the intranet. Two systems were also in development that would aid in the sharing of knowledge: UKIT was developing an asset database, and USIT had developed a knowledge base (K-Base) that UKIT was going to use. These two systems are different and have different purposes but end up doing something similar. The K-Base is to be placed on the intranet and provide a facility that both technicians and end-users can consult with problems. The level of knowledge that a person has will determine the extent to which (s)he can use the system. For example, some entries will be understandable to all, whereas other entries will need a high degree of domain knowledge if the information is to be taken and applied. The asset database simply records facts. It records details of items such as PCs, printers, and similar assets. Mike wanted to develop the system to also provide a history of each item, that is, as anybody did some work on an item, such as replacing a hard drive or a motherboard, then (s)he should record it in the asset database, thereby creating a full history of the item. This means that should a different technician have to work on the item at a later date, (s)he could see what had already been done. It could even be the case that there might be an ongoing intermittent fault, and a technician working on the system might see from the history that a particular mix of components was present that could be causing the problem. Therefore, the asset database could provide a focus for the softer side of a technician's knowledge—different technicians might make different inferences from what was available. Although both systems record hard facts and one was called a Knowledge Base, the softer aspects of the users' knowledge were important. This was recognised by the CoP members, as they could see that different people would get different levels of information. The softer aspects of their knowledge would determine what they could get out of the system. It was also recognised that the K-Base system would be crossing boundaries and be used by people from outside the wider IT group; therefore, the developers were using their expertise to ensure that users from outside the community would also be able to benefit.

The softer aspects of knowledge tended to be shared by interaction in collaboration, either in problem solving where solutions were developed or in collaboration on a project. However, rather than sharing softer knowledge, it seemed to be more the case that situations arose or were created that needed the joint application of softer knowledge—as in the case of problem solving or collaboration.

Distribution/Globalisation

As WWITMan is an internationally distributed CoP, it is not surprising that issues of globalisation and distribution came to the fore. There were the usual problems of time, distance, and culture, but there were also some incidents that showed other problems that can arise when operating in such an environment.

The problems of distribution can be very simple (and therefore easily solved); for example, in an e-meeting Mike (in the UK) attempted to send the chat history (that is, the meeting minutes or action list) to the printer in the American room so that the people there could have a copy. The Americans were quite surprised when their printer suddenly started and then suddenly stopped. The problem was that Mike's page setup was A4 size, and he had forgot that the American printer would require letter-size media.

WWITMan often runs phone conferences or e-meetings. These are generally very good, but there are often problems if more than two or three sites are involved. This can be because of issues of turn taking, but more generally (except where cameras are being used) issues of identity—the participants find it hard to determine who is speaking. This is even worse for the Japanese member who is quite at ease with American accents but struggles with the British accents.

Perhaps the greatest problem experienced by WWITMan was part of the time zone problem. The UK core members have only a short time window when they can communicate with their partners in Palo Alto, and there is no time of day at all when Palo Alto, the UK, and Japan are all at work at the same time. Opportunistic communication is an important part of the culture of the organisation, and the CoP members find it difficult catching their communication partners at the desk. Developing relationships needs regular and frequent interaction. This decreases as one party moves out of proximity and would be helped by being able to engage more easily in ad hoc communication with other partners. This could well be helped by a system that allowed awareness of when the partner is available.

Distribution does not only bring problems, of course. WWITMan (and WWIT) feel they have benefited from access to a wider pool of expertise. The UKIT people do not feel that the USIT knows more than them, but that they know different things. It is becoming more common that if they have a problem they turn to someone in the parallel core. This was demonstrated in the emeeting during discussion of the planning document when an issue was raised. USIT had a local expert who could tackle the problem.

Communication/Collaboration

The Flow Model consolidation of the analysis (see Appendix 1) indicated that collaboration is a major part of the work of WWITMan. It also suggests that most of the collaboration is technical. This is not surprising when one considers the work of the group. The importance of the collaboration is also not unexpected when one realises that it is part of the organisational culture and that WWITMan is consciously trying to extend the spirit of cooperation and collaboration down through its teams, to the extent that members of its vertical teams are encouraged to collaborate on projects with their opposite numbers in Palo Alto. In particular, Dave has written it into the objectives of his individual team members, thus, to some degree, formalising the informal practice.

The collaboration takes different forms. In some cases there are joint projects where people work on something new (and thereby probably create new knowledge). On other occasions the collaboration might be with a view to solving a problem that has occurred, and there are occasions where people get together (or are placed together) with the express aim of one leveraging from the other.

For collaboration to take place, it is clear that communication will play an important role, especially when it must take place over a distance. Both WWITMan and the wider WWIT group are all at home with a wide range of e-media.

Media

The members of the group use a range of different media for their communication purposes. They use standard telephone, fax, and e-mail, but are also at ease with voice mail, full video-conferencing, phone conferencing, MS NetMeeting, and e-meetings using a mix of phone conferencing, desktop video conferencing, MS NetMeeting, and a shared smartboard.

All the media have pros and cons, and each member of the CoP had preferred media for different tasks. The telephone was liked because it was immediate and members liked to hear their partner's voice; however, in many cases the phone was inappropriate because the communication partner was not at his or her desk and the caller had to revert to voice mail. E-mail was similar to voice mail in that it overcame the problem of the time zone, but some people felt the response from their partners was not as fast if an e-mail had been sent as opposed to a voice mail. Some people perceived a voice mail to be more urgent. This was perhaps a result of the fact that many people saw e-mail, too, as an oral and therefore informal medium, not observing conventions that would normally be the case with a written medium. Video conferencing was not generally regarded very favorably. It was not felt that full-blown videoconferencing offered enough extra advantage over a phone conversation to warrant the additional expense incurred. Desktop video conferencing, in conjunction with MS NetMeeting, was starting to be used, and although the members felt the technology was not yet immediate enough, they felt that having an image (albeit a relatively poor one) provided an advantage in the form of context. It provides a social context and gives them something to observe. Because the technology was so cheap, its use was felt justifiable. They only used it in group meetings, however, expressing the view that to use it on a one-to-one basis would be using it as a toy and not a tool. Part of the problem of NetMeeting was the fact that it had a high set-up cost; that is, the technology is not intuitive, and it had taken some time for the members to be sufficiently au fait with the setting up and running of the technology for it not to get in the way of the communication. For the main e-meeting in the case study, it had taken almost half an hour to set up the communication technology. When other people use the technology, WWIT runs the meeting for them so that the participants can concentrate on the meeting and do not have to worry about the media.

It was clear that the medium used was selected for the task. In fact, media selection was regarded as being very important when working cross culturally, as a comment by Stan illustrates: "If you've got it on the wrong medium, especially if it's a different culture (that doesn't understand English humor), you can get it wrong."

In some cases the selection was unconscious, that is, the choice was based on what the user was used to or preferred. For example, Dave is used to e-mail. It is ingrained in the culture and as yet he would prefer to e-mail material ahead of the meeting rather than use NetMeeting. However, there were also occasions where the medium was deliberately selected for the task at hand. For example, Stan chose to wait and use the phone to speak with Linda because he needed to discuss a situation where there was scope for conflict; he wanted Linda to be able to hear his voice so that there would be no misunderstandings and conflict could be avoided. Another example was to be seen in the main e-meeting. Mike was going to send the plan directly through NetMeeting, but Dan in Palo Alto wanted it e-mailed so it would be easier to reach more people. The CoP members expressed the view that they would try to select the appropriate medium for the task and if it did not work, would then select another medium. It was interesting to note that the same medium was not selected for the same task each time – it also depended on context. One case study respondent said, "A key point is the rate at which you receive feedback. Deciding how important the feedback is, or if it's sensitive and needs feedback isn't often conscious. It's habit, experience. Context can change it."

There were some interesting aspects that came to light regarding the selection of a medium. Speed of response was felt to be the key, especially in problem solving. The users wanted rapid interaction. This was felt to be much more important than richness, such as in the form of video. The addition of a shared artefact was also generally felt to be more important than the video image. If the meeting's purpose is discussion, they feel it helps if someone has done some preparation, and they have a document around which the discussion can be centred. They also felt that having a document could be a useful catalyst. This was demonstrated very clearly in the observation of meetings, both colocated and distributed.

Three key aspects that would affect the selection of media used would be:

  • Low set-up costs, that is, easy to set up and intuitive to use, so that the medium does not interfere with the communication itself.

  • Speed of response. This is possibly as important as media-richness. Even if the medium is asynchronous, the possibility of a rapid response is important.

  • The capability to share an artefact. This proved to be exceedingly important. Unfortunately the technology available needed a relatively high degree of expertise in order to use it.

Ad Hoc Communication

Ad hoc or opportunistic communication was shown very strongly in the analysis of the CoP. It is part of the culture of the organisation. There are no offices. Everybody has cubicles with no doors in order to make it easier to "drop in" on a colleague. Coffee and biscuits are freely available around the building, and there is a central café-style area with a range of magazines and newspapers. People are encouraged to chat with their colleagues because informal interaction is felt to be important. Although ad hoc communication is a part of the culture, much of it comes about through proximity. Ad hoc colocated meetings can happen a lot more easily than distributed meetings. In fact, no ad hoc distributed meetings were seen. When meeting in e-media, the CoP members feel that the meeting is much more successful if it is planned. This does not mean, however, that ad hoc electronic communication does not take place; for example, this e-mail from Wayne: "I sent a couple of voice mails yesterday with not a lot in—‘Hi, I've heard about this. Thought you might be interested’."

Ad hoc communication does therefore take place between the cores but not often simple "for your information" communication. It does sometimes happen that technical and collaborative help is obtained from Palo Alto on an opportunistic basis.

However, as a result of proximity and time difference, opportunistic communication primarily takes place within the cores. As it is so important to the group, there is scope to try to improve this situation. Something that could help would be a system that would improve social awareness—having an indication of whether the prospective partner is at his/her desk and available to take calls.

Culture

Culture was an aspect that was highlighted in Chapter IV as being an issue that affects work in the physically distributed environment. Its importance was confirmed in the analysis. Cultural aspects were felt on three levels—the national level, the organisational level, and the group level.

Culture at a national level impacted the work of the group in several ways. It can influence the medium that is selected for communication. If e-mail is used, the group is very careful about using humor as it has been found that the British humor did not translate well in e-mail. The importance of culture made the group feel it was very important to travel to visit the partner core on its home ground. Being in Palo Alto with the USIT staff helped the UK members understand the local culture better. They felt that as they got to know the Californian culture, they would be better placed to work with other Californians. In addition, they also gained experience of work practices within their partner core and could gain insights to their group culture—"the way things are done here."

The organisational culture influences very strongly the working of the group. The organisation has a strong oral and sharing culture, and staff members are encouraged to chat with their colleagues. Much of the communication is opportunistic, and therefore a lot of what the group does can be serendipitous.

The group culture can show differences between the cores; for example, it was pointed out that the American group, USIT, is much larger and can encourage staff to specialise, so the USIT group has a much larger number of wizards or gurus. UKIT is substantially smaller and its members each have to cover a much wider field. This generally means they do not have the depth of knowledge. This does not have to be a problem. On the contrary, it can be a positive benefit as the two sides can be mutually beneficial. In UKIT there is a wider group of people who have some knowledge about something and who can therefore be consulted about it, but if further knowledge is required there is a specialist available in Palo Alto.

Face-to-Face Interaction

We noted in Chapter IV that even in some on-line communities (for example, SFNet) people reach the point where they need to meet face-to-face to further develop the relationships. We also noted that social processes (trust in relationships) are important in a CoP, and the question was raised as to how these might be affected in a distributed environment. Even though the CoP in the study was distributed, it was interesting to see the emphasis that was placed on face-to-face interaction for the development of the social issues.

The members felt that face-to-face interaction was important to the development of a good relationship without which the CoP could not function. They could recall two relationships (in WWIT, not in WWITMan) that had developed purely electronically but both of these were the result of a shared (non-work-related) interest and were felt to be very much the exception. It was felt that the development of a relationship needed "that bit extra. You can give some more details of your personal life, personal interactions," according to one case study respondent.

Although the group members felt the face-to-face interaction was important, they did not feel it was absolutely essential. They felt that collaboration could take place without face-to-face interaction, and that with the range of new media available, it is possible for a relationship to develop over a long period of time. However, they were also of the opinion that the development would be slower without face-to-face interaction and that there would be a limit or a boundary. As a result of their use of the video, they had considered this question very carefully and came to the conclusion that if you are going to work closely with someone, you do need to meet them. Mike felt that:

the big thing about actually travelling is that you actually get to meet people, you get to shake hands and you get to have a curry with them, you get to see where they liveand understand much more the actual culture that they live in, which allows you to pick up a lot more on the relationship you've got with them, to make that work better. And people say that they'd like to do video-conferencing ‘cause they feel they can get better in touch with you as an individual. I'd argue that they can't—it's an illusion. Actually, the best way to get in touch with someone is to actually go and see that person or have them come and see you and then once you've built that relationshipthen you can continue that relationship, using that as a basis, on either video or audio conferencing.

The strong personal relationship appeared to carry the community through the periods of e-communication. Knowing each other gave them a greater feeling of unity and common purpose, or, as one of the respondents put it, "you need the personal relationship if you are to go the extra half mile for someone." The strong personal relationship was also felt to help with issues of identity—the members of the group knew who they were communicating with, even if it was via e-mail. Because they felt they knew their partners so well, they also had confidence in what they were receiving from them. This point of confidence also has a bearing on the legitimation. As members get to know each other, have confidence in each other, and trust each other, they gain legitimation in the eyes of each other. The trust and confidence that the members had in each other was evidenced in the fact that the UKIT members would trust USIT to test some of the UK systems for a millennium problem. They would be willing to do this because of the confidence they had in the US team—as a result of the relationship that had been developed. The UKIT members felt they knew the US team so well that they could trust them with the testing of their systems. The face-to-face interaction itself does not develop trust and confidence; rather, it facilitates the more rapid development of strong relationships that allow trust and confidence.

The group felt they were developing relationships using face-to-face visits and then trying to maintain them over e-media. This worked well, but there was an element of decay to the point where they would meet again face-to-face in order to refresh the relationship. They were keen to point out however, that the relationship did not decay to the same level each time; that is, there was an upward "trend curve" in the relationship. After a period of time they felt they reached a "comfort zone" where they were compatible with their colleagues. This would be the point at which the element of trust was present. They considered there to be a "hierarchy" of media that affected the speed at which this state could be reached, and at the top of this was getting to know people through personal visits. If a relationship were conducted solely through emedia, it would be restricted. It would be possible to get to know someone through the e-media, but it would take longer and there would be more misunderstandings. The feeling was that a cushion was needed—a basis on which the relationship can build. People need to meet physically and build up the relationship. It is not necessary to meet physically to collaborate. They felt that collaboration can take place perfectly adequately over e-media, but if more of a relationship is needed then meeting face-to-face can take it further, faster.

Importance of Shared Artefacts

A most striking finding from the case study was the importance of shared artefacts. There were several shared artefacts in evidence—a knowledge base, minutes of e-meetings in the form of a chat log, and a planning document. The planning document was the most striking, and it proved to be more central to the operation than the group realised. Although the document was designed initially to help with the planning and coordination of the group, it was also deliberately used as a communication tool. In addition, the document played a number of other, unintended roles—it stimulated discussion, problem solving, and reflection, and acted as a catalyst to collaboration. The most interesting feature, however, was that it was designed deliberately to have this effect; that is, it was only literally a planning document, but it was also used as a way of ensuring participation, and was even designed with this use in mind—not merely as a planning document.

Another striking aspect of the planning document was its stimulative quality. It stimulated both discussion and collaboration. The planning document was used to drive meetings and also was the focus of meetings. During discussion around the artefact, it would often trigger other discussions. Issues would be raised and be discussed, but the document would remain the focal point, like an anchor, as discussions moved around it, away from it, and returned. Some of the issues raised for discussion were problems that were being experienced by one of the cores. In that case the members would apply their knowledge to the problem that had been raised. In some cases they could solve the problem there and then, or they could provide the name of a local expert who could help. Alternatively, one core might have already tackled the problem, and it would make arrangements to get together so one core could help the other.

The document also functioned as a catalyst for collaboration. In the same way as discussions were triggered, the members of the CoP, through discussion of the document, would come across areas where they could usefully collaborate. As in the above example, they might see areas where one side could help another with a problem. Similarly one core could possibly leverage from the other and thereby avoid having to "reinvent the wheel." Through discussion of the document contents they were able to identify areas where they could work together, that is, projects where they could usefully collaborate, or areas where one side had a guru who the other could use. In addition to identifying projects already listed on the document, ideas were stimulated as to further areas that would provide useful collaboration opportunities. The most obvious area for collaboration was that Dan in the US wanted to have the opportunity to examine the document more closely, with a view to combining the UK and US versions and making it a single document.

Other uses of an artefact were shown by other artefacts. In e-meetings the members noted actions on an "action list" in the chat log. This functioned as the minutes of the meeting, thereby providing a "history." More importantly, this means of providing the minutes gave the members more confidence in their accuracy as they were visible to all as they were being written. The members felt that this was a great improvement over the traditional method where, as one case respondent put it, "sometimes you think the guy that wrote the minutes was at a different meeting."

There was also an occasion where an artefact was used for demonstration, a sort of situated learning. This was an e-meeting where UKIT wanted to learn from what USIT had done with its Knowledge Base. Rather than simply telling UKIT about it, the people in USIT set up a meeting over e-media and demonstrated it and took the UK people through the system.

It was, however, the planning document that was most in evidence during the period spent with UKIT. It was not the artefact per se that was important, but the process of creating it and using it. This was illustrated by the fact that the UKIT had developed a planning document the previous year, and the document itself was seen as the aim. Once the document was finished, it was rarely, if ever, referred to again. The members had learned from this and were much happier with the current situation. Mike pointed out that they had realised eighteen months previously that they needed to go through a planning cycle. They created a document and then put it away. They learned from that, and now have a continuously rolling document and intend to revisit it monthly with the aim that it will be a mature document by the following year. This will be beneficial, not because they have a document but because of the continuous process.

The process of creating the document meant the members were interacting around the document and were applying their knowledge and expertise to its creation. As they applied their knowledge, so it came to be embedded in the document. For example, as Dave worked on the planning for the Year 2000 problem, his planning expertise was reflected in the contents of that section of the document. It did not appear to be the case, however, that the soft knowledge was captured in the document and then transmitted to the other group. Rather, it provided an opportunity for the sharing of knowledge through stimulating collaboration and interaction.

The process of creating the document provided a further interesting aspect: Although the initial aim of the document was as a planning tool, it was also intended to be used for communication with the members of the other core. It was interesting to see the use of the shared artefact to aid communication with distributed members and at the same time serve as a catalyst, focal point, and embodiment of soft knowledge. In aiding communication with the peers in the US, the document had to cross boundaries. It was therefore deliberately tailored for the purpose. This aspect was the subject of both formal and informal discussion within the UK core. They were able to tailor the document partly by using an existing document that had already been prepared by USIT. This meant they could adopt a similar structure. However, the members of the two cores had already developed strong relationships with their peers in the US and felt they knew them well. Knowing them meant they could tailor the document to their intended audience. The group also had a shared background and a shared language, which would help to reduce possible misunderstandings.

As the document was also intended to be used as a communication tool, it was therefore going to be deliberately used as a boundary object. In this case the boundaries it would have to cross were cultural and physical but were within the same CoP. Interestingly, the document was also used to communicate with the vertical teams, thus crossing the boundaries between groups but, in this case, not having to cross cultural or physical boundaries. The document was not only used to communicate with the members of the vertical teams—they also had input to the document. The team managers (Dave, Mike, and Stan) consulted their people in the early stages and incorporated their views in their initial documents, which were then melded to create the main document. We see here how the representations were represented across different states and media; for example, Dave received written input from his team via e-mail and on printed paper. Stan took the four documents and merged them. The resulting document was printed and discussed. Stan transferred it between his PC and his laptop computer for working at home. This was repeated several times and then the document was placed in NetMeeting for the e-meeting where it was discussed with USIT.

The planning document also provided a vehicle for the application of the softer side of knowledge. Softer knowledge was not captured and codified, but was applied to its creation and perhaps embodied some of the softer knowledge of the people who created it. It was when it was used that it fired discussion of issues and people talked about the issues and applied their knowledge to the problem that softer knowledge was applied. Thus the document became a catalyst and a vehicle.