Delta Partners Management Consultants
Your trusted advisors.

Should a Public Service 'Community of Practice' be open to everyone in government?

Ellen Godfrey

Ellen Godfrey is a Senior Consultant with Delta Partners whose recent work has brought her to engage in some serious review of the concept of “Communities of Practice”.   For those unfamiliar with Communities of Practice (CoP), the basic premise is that a group of people with a common interest around any particular topic will coalesce to share knowledge and gain information on that topic.  Much discussion has taken place recently regarding the use of on-line CoP’s as the Public Service struggles to find ways to maintain institutional memory in the face of a massive wave of Baby Boomer retirements.

As she followed this line of research, she has developed some ideas that the internal wiki used by the Public Servants in Canada, GCpedia, might not be an appropriate venue as it does not provide for private communications between sub-groups.

To delve deeper into this topic, Ellen emailed her friend, Dr. Richard Smith, Professor of Communications at Simon Fraser University. Dr. Smith strongly disagrees and an enlightening discussion follows.  Both have agreed to allow the full text of their exchange to be published here on our blog. 

Enjoy – and please share your thoughts on this most interesting topic in the comments!

- Alcide



Hi Richard,

Hope you are enjoying this beautiful summer weather and that all is well with you. We are just beginning our Knowledge Transfer Project with a federal government department here in BC. I need to get a better sense of whether or not GCpedia can be of use for a community of practice -- as one tool for knowledge transfer. I am doubting it, as I hear that it can’t be partitioned. That is, if the group we are working with wanted to create their own CoP, it would have to be open to the entire civil service.


Ellen, Wiki’s don't partition. The concept doesn't exist in a "normal" wiki, but has been important in some situations, I guess, from business culture.

Your client might ask themselves why they want to exclude people from their community of practice. The answer to that question might reveal much that is hindering their progress from becoming one.



Not sure what you mean here. If you want to create a ‘safe place’ for folks who work in one section of a huge federal civil service, so they can share ideas about their specific practice, what’s wrong with that?

I’ve just read Etienne Wenger’s book on Community of Practice, which if you can get beyond the theoretical abstractions, is very informative. He talks about the key elements of a community of practice. One is boundaries (pdf).  Another is identities. These are old sociological concepts you will be very familiar with. To get a true community going, both of these elements naturally arise. Especially in a workplace where people’s natural inclination is to not participate. These folks need a safe place to open up and to feel they are sharing with their group. I don’t yet know how many of the people in the unit are aficionados of the openness that those who flock to social media enjoy. Many are older, and work in an environment where change can put at risk their ability to do their jobs.


Richard-smith-headshot-112x128Not saying it is wrong, but questioning the premise. How is it unsafe to allow other members of the public service to see/edit your wiki?

If you want to talk boundary, put a name on the wiki. That can be the boundary. Canada has a boundary but it isn't impermeable. I think many people would regard permeability of boundaries as one of the best aspects. Would France be interesting if you couldn't go there?

I don't doubt that they might think that it would be safer to lock others out of their digital world, but in reality their physical world has plenty of permeability.

And, it may be true that nothing will happen if the boundary isn't created. The community may never thrive. But it isn't necessary to start out with that assumption.

The reality, in wikis, is that no one is going to notice or care if you start a page or pages. It is more like a park, with lots of green space. If you and your family set out a picnic and play some volleyball, it is almost a surety that no one will bother you. It isn't necessary to draw a big line around your picnic site. And, if someone should drop by, or if you happen to see a friend you'd like to invite into the game, wouldn't it be nice to just have that happen.

Again, I am not saying you have to drop the boundary, only that it may be the case that you can start without a boundary. Or try, anyway.

Suggest to those who need a boundary that they just try without one for a few days/weeks. See if anything bad happens. It won't. If it does, they can always put up a boundary.



You make very good points and I’d like to reflect on them. As you know, one of the biggest challenges for online Communities of Practice is getting sustained participation of enough members of the community so that the online site functions in a way that is truly useful to members of the work groups meant to benefit from it. And so that they stay interested. In addition, since the CoP is meant to aid Knowledge Transfer, we will  want to design it so that it incents very busy people (many of whom do not have the habit of sharing on line) to put useful info on the wiki.

 Many of the Communities of Practice Wikis that I have seen, started by enthusiasts within a work group, never spread beyond the founding enthusiasts.  My experience, the literature, and a scan of gov’t online wikis shows that more die out within 2 years, without ever gaining the support of any beyond the original 4 – 6 founders than survive to spread. 

We are particularly keen to gain the adherence of the 40 – 60 year olds who are leaving the organization, and contain valuable institutional know-how. I have some ideas on how to do this and should know more once we begin the interviews and learn the viewpoints of the people there. During the interviews, we will find out if people will feel uncomfortable posting if they think the entire civil service can read their posts and make judgments about the excellence (or lack thereof) of their knowledge. If there is a culture of not sharing knowledge, it will be a harder sell if the wiki is open beyond the work groups involved in the project.

My experience suggests that telling people few will actually look at the wiki, won’t help.



A few hypotheses (testable)

If numbers matter, then having it open may mean more numbers.

If quality matters, knowing others will read it may mean better quality.

If getting senior members to participate matters, then something that allows them to share for posterity and not a small group will improve their participation.

Senior members of the public service have more friends and colleagues outside their division than inside. Think about it - as you rise up, there are fewer like you until you are alone in your group - but there are others at your level across the public service. And so you network with those folks.

The reality is, most of all, NO ONE will come and see these pages except the people who care about them.

Second, ANYONE who messes with them is tracked. This is not anonymous; you have to have an ID to get in.

I would hold off for at least a year or two any expectation of "community of practice" or some sort of revelations of "how things really work around here." Instead, try to get useful but mundane things up there. Phone books, meeting meetings, agendas. Work your way up to meaningful, and knowledge. Start by getting information.

In an email a few hours later Richard added

I am irrepressibly and unremittingly optimistic about the positive side of people.  I think, unless AND until there is a proven downside you should assume the best. In this case you can't really have much of a downside, given that the whole site is already protected.


Ellen Godfrey holds an MA in Conflict Analysis and Management and is a senior consultant at Delta Partners. She is currently on the board of the Vancouver Island Health Authority where lean design initiatives are profoundly improving health outcomes. Ellen brings 30 years of experience in leadership in the public and private sectors – having won a Canada medal for innovation in working with a worldwide, Fortune 100 clientele – to her views of change management.

Richard Smith is a professor at Simon Fraser University, in the School of Communication, and is publisher of the Canadian Journal of Communication. He is one of Canada’s leading voices on the subject of technology and society.  His thoughts are available through many channels, including his blog or follow @smith on twitter.  Full contact and biographical information is available here.


In my dept. we are launching an internal Wiki, at a cost to users, which can be partitioned and I have been arguing for months that excluding individuals defeats the purpose. Thanks for agreeing with me on this Richard and giving me some more valuable ammunition.

By Steve on 2010/07/30


I don’t know if my views amount to ammunition, but hopefully you can use my questions to get people to consider their assumptions.

Community relies on trust and trust is built up over time. It can’t be mandated or assumed. Online it is made from participation and curation not by exclusion. Nurture the people and the contributions of those who choose to visit your community.

Most real communities were founded by people who welcomed other people in, who built something valuable over time. Real communities that put up impenetrable barriers (Jonestown, Waco) are not models to follow.

The biggest hurdle for online communities is not being polluted by “outsiders” but being irrelevant / dead.


By Richard Smith on 2010/07/30

Having started and participated in numerous Communities of Practice, both online and off, and having read Wenger way back when, I have to agree with everything Richard says!

Richard - I congratulate you for your continued optimism & faith in the human race.

For context: I am co-leading (what one could call) a community of practice using a wiki as our main tool. We use it for project management, document development, sharing documents and to communicate what we are working on. Having used what was sold to us as community of practice software in the past, I have to say, I quite prefer this approach.

I had my doubts when we started with it, but because the group was comprised of 40 individuals who work at 20 different departments, it was the best option available at the time. With a little guidance, practice, support and training, I now love it!

Ellen - One of the most important pieces of advice from Wenger is that a community must be built through trusting relationships. Usually that means building trust through face-to-face encounters although I see that changing as people become more used to building eRelationships. However, without trust, I don’t think there can be true collaboration/sharing.

I think another key success factor is having, knowing & agreeing on the goal or purpose (the “what”) of the community and ensuring the entire approach (the “how”) supports that goal. Just coming together and hanging out is rarely enough to make people take time out of their day to learn wiki mark-up, upload their documents, or comment or edit someone else’s document. 

That said, it *can* be enough if the community is made up of people who feel alienated and crave the benefits that come from working together with others who are passionate about the same topic.

If you need something more hands-on than Wenger, I recommend “Leveraging Communities Of Practice For Strategic Advantage” by Hubert Saint-Onge as a more practical follow-up to Wenger’s more theoretical book.

Hope this is helpful. I enjoyed reading your back & forth.

By Laura Wesley on 2010/07/30

I, too, find a wiki an easy tool for CoP-related things. It just feels “natural”—once you get the hang of it. As a trick to getting the hang of it - get people to build something small and simple, like a glossary, or a list of suppliers or a “tips and tricks.”

And I echo Laura’s implied suggestion to start with face to face. In fact, I’d suggest “face-to-food-to-face”... in other words, eating together. It has a remarkable effect.


By Richard Smith on 2010/07/30

Here’s a question for you supporters of openness. My experience with many folks in the middle of large hierarchies is that they are very sensitive about posting any kind of information or opinion in an open forum—even one ‘inside’ like gcpedia.

There are many rational reasons for this, and at the end, it means that a key cadre of people will not share the treasured pieces of information they have spent a career garnering, because of concerns about being misunderstood or antagonizing someone. As we look to Communities of Practice to enable knowledge transfer, such reservations are of serious concern beacuse we may lose the input of key knowledge holders. Do you openness advocates think that these folks will get comfortable and put aside their concerns?

By Ellen Godfrey on 2010/07/30

I wonder if your experience is current, or is it from a few years ago? People may have changed. Who would ever have imagined that people would use a tool like facebook, but they do. In fact, in the millions. And for a simple reason: it is useful.

I am trying to imagine what comparable things exist or have existed that you could compare to gcpedia and form the basis for this observation. As far as I know, there are no “open forums” for information and opinion - online or otherwise - and that is why gcpedia exists.

Instead of comparing how people may or may not have acted at some time in the past on a system that may or may not have been in some way similar to gcpedia, perhaps we should ask someone from a department that IS using gcpedia, right now. Has it worked? If so, what was the secret of the success? What do they use it for? Is it just young nerds?

I also have to wonder why someone “in the middle of a large hierarchy” would ever compose a message or wiki posting that stood a chance of being misunderstood or antagonizing anyone. Surely that is how you get to the middle of a large hierarchy - knowing how to play the game. In fact, they probably know how to write forcefully, succinctly, cogently, and perceptively but still retain a civil tone. They are, after all, civil servants.

How is it that “openness” is equated with irreverence or mischievousness? What is it about members of department A that would suggest they are somehow out to “get” members of department “B?” The openness that I am advocating - and really, I am not so much advocating openness as questioning “closedness” (in other words, what is the harm that has been demonstrated from a “non partitioned” wiki?) - is merely taking a stand for the *possibility* that wiki pages could be visited by others, and “others” is already people who are logged in members of the civil service. Surely this is not somehow akin to (or imagined as) being thrown into 4chan or YouTube comment stream.

Furthermore, why do we think that key nuggets of knowledge need to be or will be passed on with spite or venom? I don’t see any evidence that civil servants will somehow become incivil when they have access to a wiki. There is considerable research to suggest that a combination of moderation (as in having a moderator) and real names (as this wiki - gcpedia - requires) results in a community in which people are exceedingly civil.

Are there really secrets that can only be confided to other members of the “tribe” in this department? If so, perhaps those sorts of things are best passed on at the water cooler and should never be typed up and put on a computer in the first place. If the security of a partitioned wiki in Canada is the security that you need, then the recent Wikileaks incident might provide some sober second thought on that approach.

By Richard Smith on 2010/07/31

It seems to me that one’s preference for an open vs. a closed wiki has a lot to do with why you are participating at all. I expect that a senior public servant who genuinely wants to impart to more junior personnel the benefits of 25 years of hard work and valuable experience will want to be as open as possible, but those in the “middle of the pack” as you call them, will surely have a different attitude. Like it or not, these middle managers really are competing with each other for positions at the pointy end of the pyramid. To those of you who are in academia, how much open sharing really goes on among faculty in the same department who may all be competing for the same research grants?

By Doug Griffin on 2010/07/31

Etienne Wenger’s writings were very influential for me a decade ago when I entered the realm of knowledge transfer in the federal government. As the manager of a pilot project involving the imminent retirement of inspectors at Transport Canada, in concert with a half dozen other federal departments and agengies, it was a fascinating study of how one attempts to address the transfer of explicit and implicit knowledge back to the organization. This project was subsequently awarded a best practice by then Clerk of the Privy Council, Jocelyn Bourgon.

Fast forward to today, where many of we Baby Boomers are on the cusp of retirement. Reflect on the bigger picture of those working in the private and not-for-profit sectors in North America. What are the consequences for global competitiveness and policy development as knowledge drains steadily from organizations?

It’s easy to talk about concepts, the abstract and the hypothetical. That’s why we have academicians. When we talk about communities of practices, this is not some abstract theory; it’s very real. CoPs are an integral element of how organizations will address the looming massive outflow of experienced knowledge workers; the challenge is to determine (as back in 2000 with my project) what is corporate value-added knowledge, and what can be discarded.

But the most critical challenge facing those in managerial leadership positions is how to constructively address the inter-generational divide when it comes to sharing and transfering information and knowledge.

By Jim Taggart on 2010/07/31

Fascinating discussion. I was the gcpedia lead for the first couple years of its existence and mostly agree with Richard. I must have fielded the question of private spaces two or three times a week for that period. I would would always ask about the reasoning for a private group and usually the answer was that someone thought someone would feel more comfortable in a private space. To this I usually replied that it was time to step out of their comfort zone. The purpose of gcepdia is to encourage sharing. If you don’t want to share, then you should not be on gcpedia. Go get yourself a closed sharepoint site. If you think sharing is important then do it in the open where others can accidentally discover your work and build on it, maybe avoid re-inventing a wheel or two.

I could go on, but I think Richard and the other supporters have said it well.

That said, closed spaces do have a role. If the topics being discussed are sensitive or have political or procurement implications then a closed community is appropriate in my mind, but they must be the exception, rather than the rule.

Some observations:

Mass collaboration takes mass (yes numbers matter)
Innovation takes diversity and openness
The creator of content is the last one qualified to judge who might find value in it
Wikis like gcpedia represent a new way of working,
Change will take time
Perhaps most importantly; working openly is not natural for some - force the issue if you can by making it hard for silos to form

My opinion.

By Thom Kearney on 2010/08/01

Thom rightly points out there is a place for private discussions but mostly these can be managed as the exception or taken offline.

The comment about wanting to protect other people reminds me of an an evaluation I recently made of another academic department. We interviewed 12 people and quickly determined there was a (not that serious) problem they had to solve. In every single interview, the problem was identified and (almost the same) the solution was proposed. The barrier: they all felt that no one else would agree to do it. In fact, every single person independently proposed the same solution but suggested that “other people” would never let it happen.

Before we impose privacy for other people we might ask ourselves, why not let them choose for themselves? And, do we really want people to share information that they are uncomfortable sharing? Better to focus on that which people are proud to share. People are willing to share quite a lot, especially when it is things they are proud of. And that is the best place to start.


By Richard Smith on 2010/08/01

Thank you for asking me to contribute to this discussion.  We’ve had the opportunity to develop a number of wikis within the Government of Canada including ones for DFAIT, Finance Canada, HRSDC and CIDA, and I always find the interpretation of why a dept needs a wiki and what it will do with it fascinating.

In my experience, the question of whether to go with an open or partitioned wiki often comes down to corporate culture.  If a department has a strong culture of open communication and collaboration, then an open wiki appeals widely and is quickly adopted by users. However, if the departmental culture is more secretive or academic (an intellectual-property focused culture often found in the sciences or policy disciplines), then it is often easier to foster adoption by offering partitioned areas. 

My observation is that human nature is too complex to apply universal approaches.  Rather, one is better to try to understand the value that individual users within a department will find in contributing to or using a wiki for their community of practice and choose the approach accordingly.

By Jennifer Savage on 2010/08/02

My experience in federal government efforts to create CoPs is that they flare up, flicker and fade away. Partly this is due to a lack of conviction on the part of the joiners. But another cause is, I think, a corporate culture that has determined knowledge is power. Not all people in the latter part of their careers wish to share what they know. Many wish to hold on to what they know and offer it out post-retirement for a consultation fee ...

By Julie on 2010/08/03

Based on some “back channel” communications I have received, I would like to weigh in on this discussion.  I think important points are raised that have not yet been tackled, so I hope that the commenters are receiving updates to this thread!

The “elephant in the room” that has been alluded to, but not addressed head on:

Many middle managers are under strict orders from their superiors not to make any public statements at all outside the confines of their immediate team.  This often includes discussing matters that, to the average person, would appear to have no political or controversial aspects to them at all.

Sometimes there is good reason for this - in one case the discussion of nuts and bolts matters related to policy implementation would be carefully scrutinized by potential private sector bidders who hoped to get a contract flowing out of the policy.  Anything that could be spun as responding to public pressure as related to the execution of the public servant’s duties could be fodder for a lawsuit and/or could derail the necessary procurement.  As a result, discussion of practical matters relating to the policy was really under lockdown.

The consultant on this project was surprised to find that many matters could not even be discussed in email.  And, as a fan of transparency, the consultant would often not see the political challenges lurking in the technical discussion, and would unknowingly send an email that could, by the chosen terminology, be misinterpreted to cause problems.

This strict message control is a fact of life in the minefield that many civil servants navigate every day.  Opponents are everywhere, and they are not just external to the public sector.  Many internal rivals seek items that can be misinterpreted or misrepresented to scuttle or otherwise impact policy and programs.

Wenger posits that Communities of Practice are built on trust.  Is it possible to create an environment of trust between community members who live their professional life in a sea of distrust?  Can these individuals be convinced to share “team” knowledge when it is available to anyone within the larger institution?

Look forward to your thoughts.

By Geoff Schaadt on 2010/08/03


You ask a very provocative and deep question. In an ideal world, which is what Wender sees, people unreservedly share their knowledge. While I have a few months until I retire from the federal government, I’ll discretely refrain (for now) from airing my perspective. However, I would strongly recommend your reading, along with the reeaders of this blog, Donald Savoie’s excellent 2008 book “Court Government and the Collapse of Accountability.”

Read in particular the chapters dealing with the Values & Ethics initiatives of the later nineties and early 2000s (of which I was heavily involved). Savoie, as the dean of political analysis and regional development in Canada, pulls no punches. A must read for those serious about understanding the workings of Canada’s public service and political system.

By Jim Taggart on 2010/08/03

I must say that I do not agree with Julie’s cynical view on the sharing of knowledge by those on the cusp of retirement. I won’t use myself as an example of someone who has always shared openly during his 28 years in government, even with retirement only a few months away. What I will do is reference my previous post where I briefly noted a knowledge transfer project I managed several years ago. The inspectors are Transport Canada who were planning to retire were very excited to think that the organization was interested in their tacit knowledge, and willingly volunteered for the project.

I’ve experienced this type of attitude since then at other departments, where given the opportunity imminent retirees want to share what they know, including being part of mentoring programs.

Julie, I believe, has made an assumption, but one that needs to be validated through some empirical work.

By Jim Taggart on 2010/08/03

Regarding public statements (Geoff), I think this might be a misunderstanding of the context. What we are talking about - at least so far - is a system akin to gcpedia wherein the discussions are strictly for government of canada employees. The question of “partitioning” was whether or not to further sequester the discussion to a smaller group (department or even sub-department).

I think it should go without saying that confidential matters - whether relating to bids or anything else - should not be on public servers. Those should stay in the realm of face to face or telephone conversations or (I guess) email on secure channels.

There is, of course, another discussion about opening up government even further. The so-called “Gov2.0” discussions certainly contemplate that, and there are some benefits to be sure. But that isn’t where we’ve been focusing up to now in this thread.

By Richard Smith on 2010/08/03

Culture needs to change. The tools we choose to use, and the way we use them can either promote positive change or reinforce existing behaviours.  By choosing to reinforce the status quo we miss an opportunity to improve.

By thomkearney on 2010/08/04

Further to Thom’s comment, the May issue of HBR has a great article by Tamara Erickson on Generation X - The Leaders We Need Now.  This new group of leaders will bring a new approach. Our workplaces are unpredictable and in a state of constant change, and this requires skills that are inherently natural to this cohort.  Gen Xers are attached to and comfortable with their networks and readily adopt collaborative tools.  They have the ability to frame challenges in ways that seek broad participation to find answers.  Further, they have been buffered by disruptive events leaving them in anticipation of a future of change.  These formative experiences have left them with strong values and a resolve to protect their work/life balance.  This consistency, the article states, is a key leadership skill in developing organizational identity and meaning.  All these skills are necessary going forward; the status quo must give way.  The article, unfortunately, also reveals that Gen Xers are becoming more frustrated and there is a risk of losing them “in droves.”

By Diane Thompson on 2010/08/04

In my experience the free flow of information is directly related to how we measure results and reward people. If individuals are measured and rewarded for ‘being ahead of others’ or the ‘most effective department’ the ‘group who saved the most money’, then why would those employees share their unique advantage and routes to their successes? If, on the other hand, they have a manager or senior colleagues who models the concept of sharing; who shares things freely so others are privy to new information, makes a point helping others glean insight into alternative approaches, then those employees know they have gained from that action and generally they emulate that behaviour. When people are asked on a regular basis who they have helped, or what have they found useful and who have they passed that along to, you get a real shift in the culture, especially when that employee receives recognition for their role in passing along valuable information. Soon it becomes second nature and an organization dedicated to both the individual’s growth and the organization’s learning is born.

By Heather Hughes on 2010/08/05

I just came across an interesting quote from Don Tapscott that reflects directly on the new reality that all of us - not just the public service - will be increasingly faced with as we move forward into the new reality:

Paradigm shifts involve dislocation, conflict, confusion, uncertainty. New paradigms are nearly always received with coolness, even mockery or hostility. Those with vested interests fight the change. The shift demands such a different view of things that established leaders are often last to be won over, if at all.

By Geoff Schaadt on 2010/08/24

Very true. We are undergoing a major transition, not only in the Public Service but in society. The established leaders will have to find new models, or retire. In the meantime its going to get messy.

By Thom Kearney on 2010/08/24

It all depends on context. Wikis aimed at the civil service should remain as such. Access to information and its open availability is merely a concept, one that certain principles should be based on practical and operational needs.

By Jim Taggart on 2010/10/17

It would be most helpful if people shared examples of specific types of information that isn’t being shared for reasons that don’t stand up; and also specific types of information or conversations that in fact really cannot be shared for legal or privacy reasons. There are certainly types of information that must be held closely by people who can ensure legal and privacy requirements are met, but it seems we need to do better in avoiding what I might call ‘privacy creep’ where privacy trumps the benefits of open-ness inappropriately. The devil is in the details and only when we can share these specific examples and debate them, can we induce principles that will help us in the grey areas.

By Ellen Godfrey on 2010/10/18

Add a Comment

Notify me of follow-up comments?