Stephen Downes

Knowledge, Learning, Community

May 31, 2004

Notes from the International Centre for Governance and Development's workshop, Building Distributed Communities of Practice for Enhanced Research-Policy Interface 28-31 May, 2004, Saskatoon, Saskatchewan.

Day One

Understanding the Demand for Distributed Communities of Practice

Harley Dickinson
University of Saskatchewan
An Academic Researcher's Perspective

My interest in this area has emerged from a long-standing interest in knowledge transfer in the health sector and especially health policy.

The main thesis is that distributed communities of practice move us only part of the way toward democratic science. What I try to argue is that sometime around the 1970s in sociology the event was announced by Daniel Bell's book: the most advanced nations of the world have moved from industrial economies to post-industrial or knowledge societies. In this shift, the production, transfer and use of knowledge has become increaisngly important for policy-making. This transition has been the defining feature of national and global development.

An idea emerging from this is the idea that we are entering into a knowledge system, a purposful creation, transfer and storage of reserach knowledge and knowledge processes: these become intentionally planned and organized for this purpose. We have a managed knowledge system. One of the challenges of this field is that we never got to plan knowledge transfer before, and one of the challenges is to reorganize in order to be able to plan, to reinterpret organizations from the perspective of the knowledge system

The nbotion of community of practice emerges as one of the key themes. In the wake of system terms comes the imperative to manage that system. When you see it in system terms comes the idea of strategies for managing that system. The notion of the community of practice was one of the key themes to emerge in this respect. It seems to have come out primarily from business schools and administrations.

The functions of the knowledge systeminclude:

  • knowledge mandating - interface between policy-making and other mechanisms, eg. pressure from the government for researchers to do their research in new ways via funding councils, etc. - a shift in decision-making from the reserachers to a wider community. For many academics this is a bit of a rude awakening.
  • Knowledge production
  • Knowledge storage, retrieval and transfer
  • Knowledge application - when you're looking at policy, the knowledge produced by research feeds back into policy, and so it becomes a system
  • .

There are three associated knowledge sub-systems: innovation systems, public policy system, and the cultural systemm. The first is the production of knowledge to generate new economic opportunities; universities have here risen to a position of importance. The public policy system gathers knowledge from research and applies it to governance, for example having social scientists involved in the production of knowledge for policy formation. It is the brokering and exchange and formation of ideas. The third subsystem primarily involves the discimpolines of the humanities and the fine arts, trying to create and transmit a cultural heritage. This third subsystem had been predominate in universities, but has slid in terms of resources and prestige into third place.Knowledge system management may be understood as two approaches. The first sees the knowledge system as an information system. There is a huge amount of effort put into developing, deploying and training people in knowledge technologies in order to make use of the huge amount of knowledge available. This effort involves, for example, the codification of knowledge. The second subsystem is the community of practice approach, which looks at knowledge not so much as data and information but knowledge as a learning process, focusing on knowledge as tacit, non-codification, for example procedures and the way things work. It takes the form of unstated non-codified know-how and know-who. Communities of practice emerged in the business community as a way of harnessing and mobilizing this tacit knowledge.

It's not a large step to combine this, and in my view distributed communities combine the ICT approach and the communities of practice approach. For example, consider the workshop brochure: groups of geographically dispersed professionals who share common practices and interersrs. You can see the two dimensions. They have difference (knowledge) interests, and they are distributed geographically. It doesn't go far enough, though - it is restricted to professionals, but not the public, policy-makers, NGOs, etc. So while this provides a solid foundation for producing interdisciplinary knowledge, as a policy mechanism it is insufficient.

Related concepts:

  • Advanced collaborative environments (Wong, et.al. 2003)
  • Collaborative research networks (Camargo and Simpson, 2003)
  • Virtual labs or collaboratories (Forest and Koopman, 2003)

The advantages of distributed communities of practice include boundary crossing (producing new knowledge in new ways), communication, and research participation. It enages or allows to be engaged researchers that might be isolated, and allows for the rationalization of knowledge production by elimination duplication. Challeges may be found along the same dimensions: boundary crossing, for example, creates tensions between different groups, different work habits. In communication, there is the possibility of misunderstanding, usually hidden behind the fact that you think you understand each other. This is exaggerated by 'mediated relationships' and usually requires face-to-face to establish something like social norms. I heard mention of a rule of thumb that 80 percent can be done by distance.

The distributed community of practice is an important concept but it doesn't take the communication out of the usual group, such as non-researchers. There are conceptsd to address this: Mode 2, Triple Helix, Post Normal Science, for example. These three models are socially distributed knowledge production. The production takes place in different settings - it's not reserachers in universities interacting, it's researchers plus others, and it's the 'plus others' that's the critical piece. If you want to have effective knowledge production for policy-development, you have to have the policy-makers involved in the research process as early as possible. It's pulling the research and putting it into various contexts of application.

The consequences of socially distributed knowledge production include new types of knowledge - not just factual knowledge, technical knowledge. Once you involve community partners and policy makers the process integrates the production of facts with political and ethical knowledge. It is the integration of factual knowledge with political and ethical consequences. Science is not value-free, it is value-laden, and this has to be part of the production of science, not just an add-on. Many of the protests in universities have to do with this shedding of the cloak of value-freedom.

Another consequence is thevalidation of knowledge. Political and ethical issues come to be legitimate standards for addressing the validity of knowledge claims under the new model.

My comment: while much of what was said in this talk makes sense, what the author is doing is taking new concepts of knowledge production and exchange and applying it to existing policy structures. The assumptions is that these policy structures remain more-or-less unchanged - that there will continueto be, for example, policy-makers. It's not clear that this is a valid assumption, because the creation of distributed communities of practice results ingreater empowerment, greater self-government.

If this is the case, then one wonders (a) how to describe the merging of non-scientific knowledge with traditional knowledge, and (b) what impact the planned integration of policy by (today's) policy makers will do to knowledge production in general. If policy devolves to the individual, then the political and ethical framework throughwhich knowledge is interpreted will be the practitioner's own. But if we at this time insert instead the political and ethical imperatives of policy-makers, we are in effect, prior to the impact of the network, substituting these policy makers' politics and ethics for individuals, in essence, disempowering people at the moment of empowerment.

It is likely the case that the screams of anguish - as the author styles them - emanating from the knowledge community are not based on their no longer being able to adopt a value-free stance, but rather, it is from the imposition of external - and in many ways alien - values into their own work and their own lives.

We can draw this out more clearly by questioning the supposition that knowledge systems ought to be managed. If, as I argue elsewhere, the proper model for understanding such a system is a network, then by its very nature such a system will resist management; it substitutes a homunculous, if you will, for the natural self-governing processes inherent in networks. While it is possible to impose an administrative structure over a network, it is not clear that it is desirable, and the effort is very likely toundermine the benefits that a knowledge network could produce.

John Lobsinger
Policy Branch, Canadian International Development Agency
A Policy Practitioner's Perspective

I have spent most of my time at CIDA as a program developer in the Caribbean, in Africa, in Bangladesh. The last seven years I have moved into what may be called the knowledge part of CIDA. I think tht CIDA has done a wonderful job of bringing together a number of parts of the knowledge economy. I think that we need to advance the knowledge economy, and to push that in development. But I am here as part of the problem: it isn't nearly as pretty on our side of the picture. Think of me as the least explicitly organized in this task, but very much searching for answers.

Governance at CIDA is a big deal for us, it's 375 million dollars a year, hundreds of projects, dozens and dozens of organizations, with about 10 'experts' working to bring knowledge to do them. From the university's perspective, anything that isn't university is policy (chuckkles), but from the policy side of it it's different. I am talking explicitly about policy themes and products.

Let's look at the themes first:

  • Human rights and social values, pluralism, property rights and other rights-type questions not at the level of international committments, children and gender issues
  • Democratic governance, elections, the media, civil society in building a democratic government and culture
  • Public sector development, leadership and coordination, economic management, environmental management, delivery of social services, security services, departmental management, auditing and accountability, and transparency
  • Local governance, how cities and small communities are run. The relationship between the levels of government.
  • The legal system, which is essential for democracy: applying laws, making judgements, systems that make sure you have systems and that people can make us of them.
  • Conflict and peace-building, conflict prevention, domestic security and the international infleunce on that, for example, how to cut off the flow of small arms, fragile states such as Afghanistan or Sudan that may produce instability elsewhere, rebuilding countries after a conflict, DDR (demobilization, disarmament, rehabilittation - you don't just send them home, as we learned in Iraq)
  • Security sector reform, policing, etc.

So what do we do? I thought we wrote policy papers. Nope - I haven't written one yet, and I've been there seven years. That is the apex of a policy-makers life - being able to come up with a generalized statement, usually short, of what we should do. These are presaged by strategies and action papers, and in turn supported by practices and policy manuals. We are often called upon to give advice. A variety of people might like to know what a policy says, what it means.

We also have stewardshio responsibilities: helping people live with the policy and explaining what it means to them. For example, networking to bring together colleagues. And in a similar way, practice-to-knowledge products. We work in areas where there may not beknowledge per se, and that it may be best developed in the field. We also provide training - a way to take knowledge to people who have to apply that knowledge. This also involves defining success indicators and generic results statements, helping them to find sources of knowledge, to write up generic terms of reference, helping them do their jobs faster and better.

In addition to working in our own organization, we also like to think we influence others. So CIDA works with foreign affairs and National Defense, along with others in Canadian society, parliamentary offices, etc. We think we may influence the academic community, and we dream of being able to infleunce large organizations (but the World Bank is the World Bank, a big solid knowledge-based entity.

Now I have to say - what I've described - that's one vision. If you talked with a person more involved with programs, she would give a more concrete specification, still knowledge based but different. A representative from the UK would slice the cake differently. This is just one way of looking at reality; someone else might say, 'I am interested in anti-corruption'. All of this is intended to be applied to a problem, and we look at it from the point of view of the problem, with these as dimensions.

But it is clear, no matter how you look at it, that knowledge must be correct, timely and in the right form.

Now: what is knowledge?

We need to look to begin with at the distinction between knowledge for production and consultation. If we look at new governance models, we consultancy applied in, for example, ways of connecting with citizens. But that's a different kind of input, a different form of knowledge - it's more your reaction to what we will do rather than informing us of your world view. Still, both sorts of knowledge involve ways of connecting the right supplier with the right receiver.

This is the most important question: we are never clear about what we think knowledge is. I know that in my business where we go out and get countries to change how they do the things they do it's about learning new skills, it's about learning new values, it's about sharing the product of society more widely, and above all it's about building institutions. I could spend the day talking about historical versus theoretical knowledge. We have lots of knowledge describing what we used to do. But what have we learned from it.

We have to ask, what is it that we bring to developing countries. If we do it badly, we will bring a solution. If we do it well, we will bring options, that people can adapt to their circumstances. What we take to a developing country is knowledge - and that is, in an important sense, the ability to predict an outcome. And because the world is a complex place, to manage the context in which it is applied - if it worked in South Africa, will it work in Mongolia?

I think we do something unique. It's well motivated, but it's enormous. We're winging it have of the time, we're winging it most of the time, because we're way ahead of what the world really knows. We have tended to work with little tightly controlled projects, but we're learning that thisn't the way to do it. We're learning that it works better if you get all the suppliers and all the demanders around the table and find what is a reasonable objective for now and a reasonable plan to accomplsih this. But there isn't one project in government where we can use that word 'comprehensive'. Look at legal knowledge, for example: I met with very many Canadians who were very smart, but smart in one area. They had never put up on a blackboard everything that describes what it is that a legal system is. We don't do these things and it's an enormous area for future growth.

Where do we get knowledge? Make or buy, broker or inspire? The ways that we do it make a difference. Very few agencies make it, because it's too expensive, so we buy it - but this assumes that the knowledge is there, that's it's complete. So we go to various agencies - government departments, consultants. Each has its own perspective on knowledge, using it, producing. For example, motivations in a university are quite different from a consulting firm, and what we do with it. And how we do it feeds into the process: you can't give advice if you haven't eaten, and to eat you need a contract.

We have to be encouraging the creation of knowledge - but by whom? Each place has a different perspective, a different set of practices. Aid agencies are very big on lessons learned, for example - but we have never stated what a lesson is. A lesson depends on what you have learned, but that depends on your background knowledge. What different people see from the same situation varies. This is the broader sense of the knowledge management challenge. Think about incentive, timing and sceduling.

Incentives: modern institutional approaches to economics. As a bureaucrat, it has dawned on me thatg people in universities work on a different set of incentives. You have peer review - and as long as you stay in touch with that community, you're good. But for us, knowledge is that which allows us to successfully predict - to go in, try to make some changes, and actually have that happen. I suspect that if we're going to work together then in addition to getting our culture and working rules aligned, we need to get our containers aligned, to related out actions to a bigger whole. Back to the legal system: unless there's a model of what that bigger whole is, there is no way to fit into that big framework.

The core question: what do we do to relate our individual activities. How do we take what we do and fit it into the larger whole. Usually we think of market economies - but markets fail sometimes. Perhaps we should be thinking of brokerage models that allow us to intervene.

I think there's room for a more considered approach than we have. It leaves to much to chance, to the innovation chance. How do we do this? Do you organize a 'Saskatchewan' model, a national model? Should we all buy into a World Bank model? That's an essential but a complicated issue.

My Comment: the search for an overarging model, while necessary from the point of view of grounding activities, is misled in the sense that there is not, and never will be, a single over-arching pictures. To understand how to approach this, we need to be able to deal with multiple simultaneous comprehensive models, a way to shift world-views on the fly.

Mark Nelson
World Bank
Sharing Global Knowledge for Capacity Development

We have spent a great deal of money on knowledge management but the results have been disappointing, so we are looking for new models of knowledge organization. This whole push for this came as a result of the collapse of the Berlin Wall and the crises that came about in the 1990s in Asia. We discovered that we had knowledge, and it was present within the institution, but it wasn't available to the people who were in crisis and needed it. I think we've made quite a lot of progress but have a long way to go.

Financial resourcs are needed less and less from the World Bank, the private sector is now the major player here. But the Bank remains an important knowledge resource. So this leads to the whole vision of the knowledge bank, and not just the investment bank.

The business case for investiung in knowledge sharing is that we want to do a better job on the ground, deal with changes more rapidly, etc. We have found that decentralization of our operations into the field is a more effective way of working, and that increases the need for sharing. We are also involevd in more partnerships, such as CIDA, and that again pushes the need for sharing (in fact, CIDA was a major driver behind the knowledge sharing agenda). And we need to help clients develop their own knowledge, which we may not have.

For example: transparency and citizen oversight - educational reform in Uganda. This illustrates why this matters. We had been investing in educational reform in Uganda for many many years. In the early 90s Uganda decided to massively increase its expenditure on early education financing, an enormous increase in funding. But the amount of money that actually reached students hardly increased at it all. It went on and on, and people called for audits and reforms, etc., with disappointing results. In 1998-1998 they decided to post how much money each school should receive on the school door - and in this one year there was a huge change in how much the school re`ceived. It was the knowledge on the part of the citizen that prompted this. Now what's really interesting is what led people to conclude that that was the thing that needed to be done, and how they reached that decision.

We really are't able to take this knowledge that we have, in Washington say, and make it usable for people on the ground. What we have seen is that disrtributed networks is probably the best way of doing that. All the knowledge was there, we had big warehouses and people on skates, but when the crisis came, we just couldn't find it. We need to make it part of your day that you are dealing with this issue, to make sure you are making a contribution to something that can be used by someone else. That cultural change in a big organization like the World Bank can be quite complicated.

We have a series of organizational structures now. For example, we have about 100 thematic groups that focus on specific issues; these all have their knowledge managers, established systems for organizing activities, a website; we have 24 advisory services at the bank, like call sectors; we have sector knowledge collaborations. We have debriefing programs for people who go on a mission - they are interviewed by someone to try to extract what they may have learned. We have a system that tries toi capture ingigeneous knowledge.

The communities of practice we have at the bank are mainly organized by sectors, though also some are organized by region. The sector boards that deal with these issues fund these issues; this relies very heaviliy on partnerships with other organizations. Wehave tried to be very flexible about the kinds of technology we use and to help the countries invest in this technology.

The governance work at the World Bank is one area where we have been able to measure some sort of change. his has workled as a result of long-term sustained engagement, both with governments but also to networks and ongoing collaboration - email, videoconferencing, and the like. That has been a very big program. We have found the most effective approach is where a country explained how they approach a problem to another country - for Haiti, for example, we are going to use Sierra Leone - we will connect them by videoconference, have some missions, and help Haiti develop some public sector reforms that will be funded by the bank. We t won't find something unless it needs to be done, and what usually needs to be done is a national collaboration, between government and non-government agencies.

We measure governments by aggregate indicators: voice and accountability, political stability, government effectiveness, regulatory quality, rule of law, and control of corruption. This turns out to be very useful, looked at by governments, media, advocacy groups.

The key enablers of knowledge sharing, from our perspective, are: leadership, communities, financial resources, culture of incentives, communications, technology and measurement or data.

On the leadership side - it was our president that pushed this, and he did it very well, because he just kept pushing it. We invested in a whole series of technologies - you may have heard of all the interactive technologies that are available at the Bank, WebCT, a VLT developed in Russia. We tried a lot of things.

What we know so far is that this stuff works when it is driven from the center, when it is demand driven, not supply driven. The demand sometimes have to be developed and clarified, but there's no question that countries want and need knwoledge that they don't have at their disposal. For us it was really key to integrate this into the lending program: how the money that was taken into the country and used. We also had to make it part of the evaluation system of the Bank itself: staff is evaluated on cleint orientation, drive for results, teamwork, learning and knowledge sharing. This has had an impact. The same criteria are built intlo operations evaluation, how products are evaluated.

As a result, we've gotten some recognition. For exmple, the Transport Sector has recognized the Bank as leading edge, we have received all kinds of awards for this work.We have tried also to do some internal benchmarking. Right now the focus is on trying to improve the way we apply our tools to our business process and improve service to clients. We asked the thematic groups what kind of things they are doing, what kind of knowledge sharing they are doing, and provide numbers: they had to go through their records and find out. Then we went to get some subjective commentaries to see what was the most effective of those activities. It still shows that face-to-face evbents are very important, second only to support to task teams. Things like the help desk and research, though, rated fairly poorly. But it's the mix that beocmes the most effective.

Finally, we need to inbvolve the client, to develop the knowledge locally. We're doing a etter job when we do involve the client. When we look at how the client evaluates the Bank, though, it shows we have a long way to go. The Bank is doing relatively badly in the things that are of the most important to clients - being able to give relevant information you can work with, disseminating the results people who need it the most. This 'adapttation gap' has been seen by everyone in the Bank, it's a way of saying, you've done a lot but it wasn't enough. The guy in Benin undestands the problems of Cote d'Ivoire, and the guy in Paris probably doesn't.

We have a lot of work to do that has nothing to do with how much we're going to lend to Benin next year.

Kim Henderson
United Nations Development Program
Experience of Democratic Governance Practice

UNDP is the United Nations' global development network, advocating for change and connecting nations to knowledge and research. Knowledge management is important to help the organization, the community, and individuals. It has decided therefore that it needs a systematic approach to knowledge management. This developed out of the use of email to send out policy advice. What ahppened is that recipients of the emails began to answer each others' questions before the policy advisors could. It was then extended to connect staff in the field with the cntral office. It subsequently because institutionalized as part of the UNDP business plan.

What is knowledge management for UNDP? It is a collection of activities, processes and policies to enable organizations to apply knowledge. It has focused more on collection that connection. Our approach, though, is a simple approach, a people-based approach - what do people learn?

Practice architecture: en key elements related to management, including agenda setting, knowledge management, etc., which generally come back to learning. It is set up as a horizontal structure, to break down hierarchies, organizing people into voluntary, flexible communities to share experiences.

A community of practice for us is a group of people with a common professional interest who bond tpgether. It's flexible, it's global, it covers all of our operational elements. It encourages a flat framework, has a multitude of entry points, and is designed to support the practice architecture.

There's no common definition of communities of practice (particularly when people talk about communities of interest - the distinction is that a CoP revolves around a profession, while a CoI revolves around an interest, not a profession). Some of these CoPs are internally focussed, and these are generally referred to as CoPs, while the externally focused ones are CoIs. The main point is that there doesn't seem to be a common definition.

CoPs are organized around the six development areas that we work on (governance, poverty, crisis prevention, energy, AIDS, management practice networks). One of the main measurements we've had over the last 4 years is memberships - has climbed to 12,000 subscriptions from 5500 subscribers. It is evaluated as the service most appreciated by those out in the field.

Essential ingredients include query-based system, which is like the help-desk, though staff tend to answer their own questions; consolidated replies; and leadership - we need to combine top-down and bottom-up approaches;

Products include:
- a monthly digest - a summary of queries and responses on the network, announcements (this becomes a place for the 'administration messages) to go, instead of flooding the network, jobs, resources, and new members.
- consolidated reply - who launched the query, what it was, who responded, and the summary of responses (the filtering and quality control aspect, also an aspect of knowledge building), and list of related resources
- Democratic governance network workspace - decentralized system that allows everybody in the community to add information (different people have oversight of different areas of the workspace).
- Practice newsletter

Day Two

Carl Gutwin
University of Saskatchewan
Supporting Distributed Communities Through Groupware

Case Studies

Case 1: Home care teams: this is where people provide health services to people in the community.

Patients in the community have come out of the hospital and have gone to their homes; they want to stay there because it's cheaper and more comfortable. Home care provides services to these people, including occupational therapy, physical therapy, nurses, home health aides, nutritionists and case managers.

For each client in the community, we have an ad hoc team around that person, made up of people who are providing care for that person. People are assigned to clients as the clients come into the system. So every person will have a different team, and these teams overlap.

The service providers work in different offices, and are community based (that is, mobile). They are professionals and are allowed to do what it is they want to do; they have a high degree of autonomy. Everyone is moving around; each service provider will visit five to ten people in a day, so they're hard to get a hold of. They don't have mobile phones. So face to face meetings are difficult and relatively rare.

This may lead one to think that they don't need to collaborate at all - but we found out that there's a strong notion of collaboration and what other people are doing with that client, because they are concerned to know how another person's care will impact on their own care with that client. There is also the issue of scheduling; they may find they've arrived at the patient's home at the same time. This may mean wasted trips, but they have scheduling flexibility, and sometimes they arrange to meet at the client's home.

This is a system designed for loose coupling: to preserve autonomy and flexibility, to consolidate information buffers, etc. We found that loose coupling isn't widely studied in the literature; people look more at asynchronous groups or the kind of collaboration where people are working together at the same time. This is probably similar to your needs: you have people working autonomously, and they're not working as a single, coordinated team.

The most important thing was not not compromise this sense of autonomy; for example, we could have scheduled everyone from a central office, but this wouldn't have worked because people need to be able to take longer than planned, or to take the opportunity to meet. You never know ehether the client may be asleep, or whether the family may be there. So this autonomy is important.

The information buffer is just a repository of information for a given client. We found that each person was keeping their own client chart. It was impossible for each person to get access to the information in each other's buffer. We didn't want to force everyone to use some kind of centralized system; we just wanted to give people access to each other's documents. So we used a shared document repository. People still used their own charts, but access was available to other people's charts (subject to privacy considerations, etc). Each person could contribute a document to the chart, and the person could choose to share it. A record would be made of any given access. And a timeline of any activities within a chart would be shown. We found that it was important to track the activity as it was to place the content in in the first place, because it showed what other people were looking at, what they were interested in, and identified hot issues.

Supporting loose coordination mostly involved flexibility of scheduling but also being able to coordinate their activities in the client's home. People often left traces of activity in the client's home - a new piece of equipment, new medication, and so on. This seemed to us to be a good opportunity to act as a clearing house. One way is to put an electronic device into the home, where people track what they do. The home care people do this, they put a binder in the home, called the Communication Binder. But there are issues with public information in a place like the client's home. For example, the family would look at the binder, which is fine, but they would change things or take things out.

Finally, these services providers are rarely available for real-time communication. Mobile phones would not address the need; you can't take a phone call when you're providing a service. So forms of asynchronous communication were needed, to avoid telephone tag, where you don't need to worry about being out of touch. So we used a simple cellular IP connection, that they could shut off, and when they were onthe server would forward them any new data that happened while they were off. Service providers check in between clients to see what happened recently with the next person.

In designing these systems it was the service providers only that we talked to. The actual design was a mixture of suggestions, problems that we saw and solved, and things that we came up with that we thought would be useful. We got design feedback by doing a lot of iterations in our design; "So this is how the system would work..." Generally people are better at reacting to designs than doing the designs themselves.

We gave them all laptops. We all know that the electronic health record is coming. We moved it up a little. We wanted to be ready. We wanted to be able to say what we wanted when the guys came knocking to build this centralized monolithic system and who probably won't consult the practitioners.

We did field trials alongside paper record keeping. All kinds of issues came up - typing skills, making internet connections. When this is rolled out, it will be a huge training exercise. The mobility of the group was the big challenge. We considered placing the computer in cleint's homes, but providers often work on a file outside the home. The home is for treatment, not for paperwork; they get ready to do what they do and then they go in. And the work in the home is not very computer friendly either - it would be bad to record interactions in the bathroom or the kitchen table. For in-home work with home health aides, we provided them with a PDA-style thing.

Rather than coming up with an algorithm that would do the scheduling, we provided a system that would allow them to see each other's scheduling. That allowed them to avoid conflicts, or engineer it, as it were. It's tricky sometimnes when people make a change at the same time, but that's rare enough that we could live with it.

Case 2: Distributed software development: developing software in the open source community.

Three open source development teams building free software that people volunteer to build. It's very widely distributed, people in different countries, and very few face to face meetings or even phone calls. All of the developers are free to contribute to any part of the project that they want to; it's all volunteer, there's no hierarchal structure. But software has strong interdependence; if I make a change it will affect other people's work.

NetBSD is a unix-like operating system. It has 170 devlopers in 12 countries. These are not users - this is the upper echelon. You have to be invited to develop and sponsored by someone. There are time zone, labguage, communication issues.

Subversion: this is a revision control system; it tracks changes that are made to the software. It allows you to roll back changes. It had 32 developers in 3 countries. So it's smaller, but the same issues arose.

So, how on earth could these people coordinate their activities to avoid duplication, create errors, avoid working at cross purposes, know who's working on a particular part of the project, what their plans are, who to talk to, and so on?

Given that these people are the cream of the crop of programmers, with day jobs in big companies, you would expect that they use some sort of fancy system. But they solved the problem - and solved it very well - with email and text chat. About 75 percent email. This is about as easy as it gets for a technological solution. How could this possibly be enough?

Think about mailing lists: you send a message, and it gets sent to the entire group.These are the most important mechanism for communication on these projects. People were expected to read the archives before joining the project - this is the apprenticeship; they may have been reading the list for months or years before becoming developers. The lists were also read extensively by developers; 100 percent of the people we talked to read them; they say, essentially you have to do it. There is a high participation level; it begins to take on a more conversational flavour. You get follow-ups and follow-ups, not just announcements. And you have lurkers - people whoare following along but who don't contribute. Lurkers are sometimes viewed as bad things, and papers have been written about how to get people to delurk, but it's not the case that lurkers are a bad thing.

The lists support awareness as follows:

  1. Stating your activity - people were very careful to talk about what they were going to do, before you did anything you told the list first and got comments.
  2. Finding the right people to talk to - you watch the list to see what's going on, and soon you see what sort of things different people are doing, and that tells you who to talk to. Or if you don't know who to talk to, you can just ask the list. The list acts as a proxy for every single developer, because youknow that they're reading it. It's kind of like yelling in a room, "Who knows about blah..." and the right people come along.
  3. Overhearing -

(Stealing cycles from humans - when there are billions of people out there, you can do almost anything -- its a scale thing eg. the ESP game. (Google this)

There's a lot to be said for these simple tools - there's no need for videoconferencing, there's no need for shared desktoips. Get the culture in place - the culture of participation is much more important than the tools you use. You need the critical mass, and there must be a committment to keeping it public - "If it's not on the list it didn't happen."

Ben Daniel
University of Saskatchewan
Human and Technology in the Design and Sustainability of DCoPs

Yesterday there was debate about what a DCoP is. I identified two streams: some people think it's not necessary to define them. But for me, it is important to kn ow what I'm talking about, or if I am in the corporate sector, to tell my clients what I'm trying to sell to them.

I talk about the chemistry of DCoPs, because it's a misture of different elements - governance, computing, development, economics, education, and more. If you take these together you get a big cup, which is the DCoP.

So what are the semantics or syntax of this? If we are going to build tools, we should have ways to represent the problems we are trying to solve. So we have 'D' - which is to 'reverse' (as in decriminalize'), to remove or remove from. There is a distinction between 'D' communities of practice, and others. A DCoP is geographically dispersed that share common practices and interests in a particular area of concern. What makes these communities distributed is that their activities are mediated by 'I' and 'CT' (Information and Communications Technologies). These are more than communities of practice; they are also more than communities of interest, because they are working on specific professional problems, not just interests. A DCoP is a type of virtual learning environment.

The features of DCoPs include:

  • Shared interests - organized around tools, interests or domains
  • Common grounding - people may have a common interest, but it almost takes forever for them to find these points of commonality
  • Autonomy in goal setting - members set an agenda based on the needs of the members at a particular time
    People say you can buil this things. But you can, in a sense - you're trying to bring people together. This to me seems to be a kind of design. But atthe end of the day, every member has to decide what's good for them. If you try to impose something on them, they may decide that it's not for them. You have to design it in such a way that each person has a piece of the pie they can take with them.
  • Voluntary participation
  • Awareness - social protocols and goals - if someone getsnasty, you have to have a way of dealing with that. And there has to be an awareness of expertise - knowing who knows what.
  • Communication - not just mapping - Iam interested in the learning, the content
  • Interaction -0is mediated by face-to-face and is inrched by ICT

Why are people interested in building DCoPs? For one thing people are interested in informal and causal (casual?) learning.There's an element of tacit knowledge if you like. And the entry barriers are low - you don't need a million dollars to get in one. They maximize information and knowledge yield - there might be a lot of knowledge, and it is up to the individual to extract the knowledge and contextualize it. It is an information hub. But it minimizes the cognitive load caused by information overload.

I remember a talk, where someone say, you can't ask someone to learn it for you. However, you can rely on someone to get some information for you, because you don't have time, and take it and put it to your needs.

There are two main technologies for CoPs. One of them is 'push' - automated processes of searching, locating and all of that. Tools can do all of this for you. Joining a listserv for example - the tools send you the mesages. They may not all be relevant to you but that's how the system works. Data and information is sent on a periodic basis - getting an update message from Microsoft, for example. But sometimes you get those nasty messages that won't leave you.

The other technology is 'pull', which enables people to seek the information they need. This allows for a high degree of personalization - you can seek information from colleagues, you can tap into containers, and so on.

There are some issues with push technologies: privacy, for example, relevance(think spam), too much information, security, performance, and user or system vulnerability (think email virus). With pull, a major issue the ability of the user or community to search, locate and retrieve information - sometimes they don't even know how to do web searchers. It also limits the community to what they know or what they want to know at a time - this is one of my main issues. Communities are very important, and societies have accepted for centuries ways of building communities, ways to link different communities together. Society should not only be building communities, but building bridges. That's why we're building DCoP. We're trying to link pilicy-making communities, research communities, and all that.

( Diverse means of association - cross-cutting. Different ways of associating. My thought - community profiles (box diagrams showing overlap with society)

Challenges for DCoPs - they are difficult to engage organically - it is hard to get them to emerge. Why? These people probably don't talk to each other - you have to try to bring them together. It's difficult to build support due to diversity. It is difficult to establish shared understanding and trust among members. It is hard to merge individual interests to form community interests. And so on and so forth.

In building DCoPs you generally push, and let the members pull. But where and how you push is something to discuss.

Kevin Danner
University of Saskatchewan
Building a Knowledge Management and Communication System

There's a lot of smart people with a lot of gold in their heads, and I want to mine some of that gold. You'll notice that my slides are empty. I am hoping you'll do my research for me.

But I do have a poject, and it is the Quality Improvement Network, composed of representatives of various health care associations in the province, and a health quality council that is funding it. 27 individuals in total. I was approached by the council with the idea of creating a communication system (before they went out and spent hundreds of thousands of dollars). Now that's the type of community - it may be a community of practice, it's certainly distributed, but I'm not hung up on lables.

So - you're a participant nin this network. What do you want it to do for you. Now your a system architect or programmer. What's available? Now you're a communications officer. What techniques or media can we use. Now your the change agent of the agency trying to market the system to the users. Now as a participant, what barriers did you face? As a network facilitator, what can we do to overcome these barriers. And finally you can be Stephen and undertake a critical evaluation.

Joe Garcea
University of Saskatchewan
Introduction

I hate these evaluation forms. You get people to volunteer to do something, and then they realize they're being evaluated. Pretty soon you will be handing out evaluation forms for your dinner guests.

Ansit Sarkar
University of Saskatchewan
Interfacing Research, Policy Analysis and Practice: Building a Governance Knowledge Network

The Intrnational Centrefor Governance and Development was founded in 2001 as an interdisciplinary unit - political studies, sociiology, economics, law - and through the GKN network we have drawn on people who are interested in ICT, education and communication. It is interesting when we look at what drives communities of practice - we really don't talk about ourselves very much. Our interest is to talk with each other around the campus but also beyond the campus. Knowledge development and knowledge sharing are seen as our key missions.

The University has been involved in international development for a long time but have moved from project to project. Sincde the university has over the last decade been involved in a number of governance projects, we thought we could briung some of that knowledge together. That is the reason why the university gave us approval. And it was John Lopsinger who said if we can find a way to connect som of your knowledge with what's happening across Canada, that would be useful. So we did a feasibility study that brought us into contacty with other groups such as the Parliamentary Centre.

About the same time SSHRC was in the process of converting itself from a granting council to a knowledge centre, and getting people to move to that agenda, called Image, Sound and *** Technologies Initiative. It observed that academics are not making the best use of these technologies. So as we were moving towarfd the GKN project, we took it in that direction, and organzed this workshop.

The GKN initiative is a virtual Canadian community of knowledge and practice on governance and development. We are interested in knowledge sharing and knowledge creation. We want to act as a platform for sharing lessons, etc., and as a hub.

We want to generate not only knowledge sharing, but to also help in generating collaboration. So one of our goals to put into the platform is tools and processes to do this. And we want to link beyond Canada and North America, and into th developing world, and to encourage strategic partnerships with other international research agencies.

Why worry about this? Research is more likely to contribute to evidence based policy making, and to reduce suffering and save lives, if:

  • it fits within political and institutional limits
  • researchers and policy makers share particular types of networks, and
  • outputs are based on local involvementm and credible evidence
This doesn't mean that reserachers set policy makers or that policy makers define research, but rather that there is a relation of interdependency established.

We are looking at both tacit and explicit knowledge in our knowledgenetwork, and so we are looking at tools to capture both. We not only want to bring the knowledge together, we also want to connect people to people.

Why this? The current knowledge on government and development is fragmented - you have to go on a hit and miss basis. There is also a lack of sufficient awareness of people who are working on specific issues. We have, here are the experts, but who considers them experts and why they are considered so, that's not understood. And any network that we build would have to consciously bring together not only the community of practitioners, but also incorporate the community of interest. We have to ask, do we want everybody to be contributers? We heard this morning that there is value in people who are just listeners or just watchers, who could make a contribution at the right moment.

The stakeholders involved recommended the establishment of a knowledge network concisely focused on thematic lines, a network built upon dialogues, not passive information dissemination. We want to develop a network that links government, the private sector, and the university - there is increasingly an awareness that universities need to be brought into it in a more active way. GKN could also become a catalyst in designing educational programs in the area.

What we expect to accomplish is the evolution of a communkty of practice aggregating policy briefings, field notes, and research papers, along with biannual meetings to share the state-of-the art information and knowledge on governance and development - critical information and just-in-time information.

So when initially we put that into a format we looked at three major institutions - legal institutions, etc. Private consultants do a lot of work, but they don't really get captured. So one of our challenges when we create up the protocol in such a way that we capture their information without takinjg away their fishing rods - without depriving them of a livlihood.The proposed framework, then - a work in progress - is based on a newsletter, a web portal, and linkages with international networks (based on this session we may change this). Whatever technology that we use, we want to make sure there is an opportunity for shared learning amoung people are proficient and not proficient in technology. And while we may start small, we should be scalable - it should be seen as a way of building exclusionary zones. Other people have talked about trust and being able to count on quality, and somebody has to ensure that quality. We also want to accomodate diverse cultyures - where culture is the way we work, the way we organize our activities. There are a lot of cultures, two different government departments will have different cultures.

We have the idea of joint value added potential. This is not just for the academics. There are some who are only interested in the aggregation of knowledge, while there are others who are interested in the creation of knowledge. We want to create a mechanism for allthe types of knowledge dissemination technologies.

Note: GKN knowledge objects repository. Two objects so far. http://triumph.usask.ca/gkn

Pari Johnston
Association of Universities and Colleges of Canada (AUCC)
Linking Knowledge from International Development Projects

AUCC is a national non-government organization representing 93 member institutions represented by university presidents. We are primary a public policy advocacy organization. We advoate on behalf of our members, and have important functions in communications and information sharing. We have three major branches: national affairs, the probram (business) development branch, and an international affairs branch. Then there is fourth branch responsible for corporate services.

I am from the international affairs branch. We undertake policy development and advocacy on key international issues. The handout identifies key international policy priorities. We do national reserach - for example, our report, Progress and Promise. We also have a communications function - we have a publication UniWorld to tell what universities are doing internationally; it comes out twice a year. We have a knowledge sharing and clearing house role - we manage a couple of useful databases. We also maintain listservs, university presidents are our members, but also senior officials responsible for international initiatives.

Our key international policy priorities include: foundations for maintaining and enhancing Canada's place in the world,for example, through student exchanges; to maximize benefits for Canada from the global knowledge-based society, for example by leveraging Canada's research capacities, or by getting young students out for international experience early in their career; and increasing developing countries' capacity to benefit from a global knwoledge based society, as there is a need for Canada to do even more in this area, such as building the capacity of higher educationa institutions in these countries.

One of the key challenges is that there is a lot of capacity and expertise in universities that could be better leveraged to support Canada's international activities. We often hear from government, "How can we better tap into your institutions?" We think we can better communicate and convey what's going on in our members.

Our members are extremely active internationally. We as part of our database know there are almost 3000 international exchange agreements and about 2400 international development projects. There's a fairly even spread among the regions in the developing world and a broad range of sectors.

Our value-add is in serving as a platform in bringing toegether the university administrators and the policy makers. There is, for example, an annual policy discussion with CIDA. Most recently we had meetings with university presidents and policy makers from a range of departments around the foreign policy review process going on right now. We arelooking to make this more effective. We are a real repository of k nowledge about Canada's international engagement; we're updating them right nor, gathering information from our members, looking to produce better reports. We also can and do filter, package and resubit information about what our members are doing in a way that is more useful to policy-makers. This is not only a resource for policy-makers, but also a resource for researchers themselves.

Research communications is a growing area of importance to us. We are being asked to convey the story - not just the data, but the story - of how reserach projects are progressing. What has the impact been of the government's significant investment in reserach in this country? The issue of how to better tell the story is something that is going to be a growing preoccupation for us. In this context we most recently brought together directors of communication of universities.

So much knowledge is stuck in project reports, and it doesn't get disseminated, so one of the questions is how to take this information and distribute it. Many of the projects are more applied, not research, though it's the projects with a reserach piece that have more of an impact. (There is a broad methodology to measure this - CIDA uses a results based framework - they're looking for outcomes as well as impacts). We also recently set up a 'peer assist' project linking Canadian and southern partners, and out of this we set up an electronic 'Impact Plus' system, which hasn't been working as well, useful but not a substitute for face-to-face interactions. It takes time to build trust, it is not clear academicsa are happy with the format, and should perhaps be focused on more specific topics or issues.

Future opportunities include the international policy review going on right now - this is a real opportunity to articulate our strategic vision to capture the contribution of Canadian universities. Also, there is the potential development of the Canada Corps, where we could articulate the strategic contribution of Canadian universities to strengthening key public institutions.

My comment and the question I asked after the session: how do these two 'networks' (centres, actually) exchange information with each other? There ensured a discussion of standards and information sharing through mechanisms such as RSS.

Terry Kuny
XIST, Inc.
The CoP and the Librarian:A Love Story

"The future belongs to neither the conduit or content players, but to those who control filtering..."

I'm not going to talk about libraries per se - the idea of the library as a knowledge repository is an idea that's changing as we speak. But we can think of libraries as wellmanaged organized knowledge ecologies, and some of these underlying ideas are important.

You would be hard pressed to find any discipline as networked as liubrarians - some of the first bitnet exchanges were among librariens, and the interlkibrary loan is older even than the League of Nations. We almost have too many networks, too many small froups.

I do not believe knowledge is power. I do not believe we are in a knowledge eco nomy. We are in an attention economy. And I don't think disintermediation is the rule; I think that reintermediation is the rule. I also have a bias toward usability and open access. I prefer pragmatic approaches and pulling things down from the 10,000 foot level.

My community of practice: my in-box. Also, digilib (mailing list). Also JISC mail. Anyone in the UK in higher education can establish a mailing list - it is astonishingly successful and very low cost. You want to start a community of practice? Start a mailing list. But in your own institution there are too many strings attached. JISC cuts through the red tape. Also instant messaging.Also, weblogs and RSS that replace some of the newsletter activities. Also the IETF - they have countless working groups. If you want an example of a highly successful infrastructure, this is it. Also, wikipedia - this is one of the most interesting collaborative exercise there is. Also match.com.

DCoP has a long history - old wine in new bottles. What makes it a community? The barriers to entry are low, the barriers to exit are low, and I can say things and do things that won't encumber me in an organization.

A lot of CoPs exist - in fact, they are in competition for a scarce resoruce: our attention. All communities are, by definition, distributed - people are distributed. By distributed in the last few days we are substituting for the rhetoric of online or electronic or virtual. The reason we associate withthese things are varied, and we can't force them. There is a lot more association - sitting in on things, tracking, etc - than participation. These things are often not dialogues. I don't think they're goal directed - it doesn't mean they can't be, but mostly they aren't. All the knowledge in CoPs is instantiated in documents - it's really not anything until it's in a document. So knowledge management is document management. (SD - this is the librarian speaking). And: if you require training to use it, the service is already dead.

There's still a big role for 'infomediaries' in this economy - moderators, researchers, distillers, analysts, promoters. Librarians play an important role of identifying and analyzing information needs. Libraries are often seen as the nice quiet passive neutral infrastructure. So there is a principled aspect of librarianship. We provide a neutral infrastructure for broad access (open access, transparency). We don't care what your information need is: we just try to provide resources. Librarians do information management, and increasingly so as we see more and more repositories - and we will see a whole infrastructure of networked information repositories, and that will change scholarship. Librarians are all about metadata (matedata = 'cataloguing for men')and classification. And if you ask about how you want to classify things I will tell you to look at your broader community.

The challenge in sustainability is how you create CoPs that last over time (especially when yo have no resources). How do you enable the cultivation of new communities of practice. You *can* create a CoP, but it's work. There's a kee here - there's trustworthiness, there's stability, there's persistence - but mostly, there's work. Usually it will be one or more individuals who will motivate it - I have never seen a CoP that didn't have a chamption. The audience constrains choice: leading you to appropriate technology, motivators, value, governance. I cannot create a motivation other than providing value - I can't give them money, I don't have money to give them. Now let's face it, ifCIDA says it's going to set up a CoP on Africa, everybody will be there, because it's CIDA, they have money. And while openness and democracy are nice, but sometimes we want autocracy not democracy.

Now let's talk about interoperability. This is the question we had before the break. This is the big question. It's not just a technology question p- it's a social question, forming connections among stakeholders who didn't even know they are stakeholders. What brings us together is not just commanalities, but also what we have that is unique. To share documents, you need standards for data descriptions. When you pull a document from somewhere else, you want it to sit with your other documents and make sense. There *are* shared platforms that we can build and emable - like email, for example. We should copy these sorts of things. There's a lot of harmonization of policies and procedures that takes place. For example, how to streamline the grant making and grant seeking process - so we don't have to key in the same info over and over. Key enablers here are OAI, Dublin Core metadata, and (maybe) a 'Development Markup Language".

Build reluctantly. A lot of technologies already exist out there. There's an old probverb: never accept a gift that eats. Well technology is such a gift. And note well: applying technology to a dysfunctional or noncommunicative organization will not make it a more functional organization.

Email is a key technology. Communicate by email, store stuff on a website. There are things like new messaging technology. Weblogs and wikis and RSS syndication. Communications archives are important. I don't care whether you call them databases, repositories or whatever - it's all website. Everything is a document, everything is in a dtaabase. Thou shalt not abuse the time of others. The key is facilitation and moderation. The tools need to be improved, and there is a real need to learn from the experience of others. Small changes can make dramatic improvements - do usability testing. There is so much work to be done in evaluation and assessment of electronic services. It's really a problematic area.

The challenge is to promote a culture of sharing. We need flexibility of management. Open access encourages sharing, and the opportunity to work on systems tailored to your needs. But not every organization can take advantage of that - a small organization can't, but as an aggregate or collective you can. An open source grants and contributions system, for example. Design at the lowest common demoninator; support linguistic diversity.

Cooperate. The computer science technical reports database - the departments all manage their own collections, and they agree on protocols for sharing. It's going to come down to a set of social agreements, starting small, working out how to do it, and it grew right away. Or see CARL, the UBC public knowledge project, etc. Promote open access. Thereis great software: oss4lib. You're all managing links - nthere's all kinds of links gateways that support Dublin Core metadata. Use that.

Build it and they will come does not work. A;ll CoPs need care and feeding. For example, for my community, I sometimes send out news tidbits. CoPs may need advocacy. Scale matters. My list gets $10K per year for a trivial sponsorship, a very passive low rent thing. Scale creates increased opportunities. Marketing tells others that CoPs exist. You need to discourage the Balkanization of professional communities where possible.

Transparency and opennes does provide a good ground for participation. Different values and motivations are at work. What's important is to preserve the archives. Decision-making online is really hard to do - that's why a lot of the activity is directed toward face-to-face - when you need closure, face-to-face is more effective.

Terry Gibson
University of Saskatchewan

Who do you enagge - who is your audience, who should be at the table? You want a wide variety of partners - a wide variety of partners. Links with international communitie should be pursued. The broader public is a useful partner. But there's computer access issues in the hood. How does the guy that we pick up every week and get sober inform the process? Also: media, also, the business community, because they are important decision-makers in the workplace. I need to know who I'm talking with.

When to engage? Early on and be continuous? Maybe - there's some discussion around this. You've gone to the hood and enaged four times, and they're sick of you. They've seen no fruit - do you keep engage? Engage appropriately, at the right time. Evaluate the impact of the research and brief government officials before the passage of bills that impact public health. Get informed from that stakeholder group so that what you are writing is relevant.

How? No single dissemination method. Electronic products, sure. But that's all changing. Use working groups rather than conferences to encourage dialogue. Produce a variety of outputs - pamphlets for community groups, technical papers, etc. Establish formal links and partnerships. Employ a reserach transfer specialist (think about it - I don't know how to do that!). Create easy-to-use computer models.

I was a strategic training fellow for this class. It was great - based on six substantive reserach areas. (list) The emphasis was to create a framework to guide reserach, reserach methods and analytical tools, and (importantly) building a research database and getting the research into the policy decision-maker's hands. Similar schools share a similar theme: knowledge mustg be shared to realize significant changes. Share the knowledge.

David Mechanic: who will lead the future for population health? Such institutionalization is necessary - it has to rest somewhere - it has to have a champion. I want these DCoPs to inform thatg policy. I will be attending a summer school to learn new strategies for adapting to such constraints while maintaining research rigour. They will talk about implementing the reserach plan and telling the story, identifying the opportunity window, getting the research into the hands of those who can implement it.

It's all the same: let's talk to each other, let's collaborate. It's all they talk about. We're standing at an important moment, and these DCoPs can play an important role here. It is impoortant that the academe recognize the work of those in the field. Cherish, and value that work. Practitioners need researchers to accept their work as evidence. It is this work that is the impetus for new reserach and galvanize new findings.

It is important to have the ability to identify and engage the right mix of people at the right time, that they recognize and value the input of the others. CoPs embrace a broad perspective, and they need to recognize and overcome different priorities. There needs to be a champion or leader to ensure its viability (from grass roots or above - that's been the ubject of debate). What's important is synthesizing the information and transfering it in digestable form to those who make decisions and policy.

Commentary: Robert Dodd; who governs? The answer - it's an interlocking set of elites. What I learned is that it's going to be an interlocking set of CoPs.

Comment: you relly need knowledge brokers. This is all really hard, learning is really hard, learning about another perspective is really hard. So how can all these groups talk to each other and really communicate.

We're learning that that's the essential role. Building bridges. Look at the federal government. You want to see stovepipes! But it's not technology - though you can use technology to crac open the organization. You make it a technology program, but then it turns out not to be about technology.

Significant - extracting the really significant lessons of practice based groups. It all gets lost in reports. If it could somehow be mined, given somehow a credentialized stamp.

Day Three

Trish Paton
Canadian Association for Studies in International Development and Saskatechewan Council for International Coopertaion
Establishing New Research Networks

CASID is an academic association for international development studies. What we bring to this context is that international studies by definition requires global capacity in some form. The organizational perspective is transtioning from something that used to be interdisciplinary to something that is becoming a discipline. Within CASID our linkages have been primarily around graduate students. The issues of collaboration within the NGO community include an increasingly limited access to federal funds and an environment which remains at base a competition for funds.

We haven't over time found that creating the network was that hard - the network basically created itself - but it depends on academic culture. Some departments have an interest in sharing ideas, it's part of the academic tradition. But sharing externally is a different problem; finite resources limit the ability and willingness to share. If there is a single person in a department, sharing resources and making international visits falls lower in the list of priorities.

When we look at networks, we see first that there needs to be some shared understanding of whatthe network is intended to do. Sustaining a network is a recognized utility of purpose, that soime function is being performed. A network, like an email list, that everybody belongs to but nobody contributes to will not be sustained. You need financial resources, but if you have human input this requirement is minimal. Technical resource, skills and time to learn them and use them are also major factors. Finally, longer-termaccess, archived material, is important; if someone goes to check your citations and one of those links has disappeared, then you need to repair your citation.

Networks or communities of practice fall under two main themes: passive construction, you have a community, but you don't have a community that's creating anything itself; and active construction, a more facilitated construction, where your central unit acts as a facilitator. A lot of networks or communities of practice are mixed bag, both active and passive. CASID's listserv is active, but our website is fairly passive. Within the NGO community, the listserv is used to send materials, but gthe websites are mostly collections. And even the listservs are primarily active, in that they're limited to the organization in question.

We were asked to discuss the challenges in creating and sustaining the network - I would tlak of them as challenges in creating the network and developing or growing the network. Resources to continue a network and expand capacity, because they all have websites, require a certain amount of money to store material. That may vary depending on what you have. Existing networks that come to demand different things - your central organization needs to figure out a way of doing these things.

Another challenge is the mix of interested parties at the local, national and international level. You might have local groups or you might have international bodies like OXFAM. That mix means that you aleways have a mix of interests, needs and biases, not to mention a mix of languages and cultures. How you balance those out is a challenge, because there is a desire to do everything, to meet everyone's needs. You need to start somewhere; as Terry said yesterday, pick something small and do it well. You can't do everything.

As well, the competition between traditional means of distributing knowledge, such as publishing a book or an article, versus the more distributed means of disseminating knowledge that are available to us, is another challenge. You can do all of the traditional peer review and so on and publish an article online as they come in; you don't need to eait for eight articles. But there aretraditional values. As long as academics are stuck in that publish or perish world, that's going to be a barrier. And there are variations of value placed on different types of knowledge production. We talk about doing more work that feeds into policy, but one of the challenges is how that's valued, because in many instances it's not. And the issue is where you get involved in the policy process; desk officials are very busy, so you don't know where to get in, beyond the grant application writing process.

The last challenge that we've had is linking an interested network into an active membership, particularly within an academic community where the organizations need a certain bit of core funding, to do the secretarial work, to send mailouts, and so on. On the non-profit side, this is an even bigger question. It's potentially very shakey ground.

Resources, time, interest, technical requirements, utility of information, available personnel, accessibility: these are critical issues of sustainability.

A question that came up is that of new reserach agendas, coming up somewhat out of SSHRC's transformation consultations. There is a lot of question around who controls a research agenda and who drives the research agenda. In social sciences it has often been the individual reseracher, and we shouldn't fool ourselves into thinking the sciences are different. In research agendas there is the intent that they be problem oriented, fixing problems, and wen the agenda is changed the presumption is that the problems are fixed. But who decides when and how the old problems drop of. Is the current breadth of reserach well understood and well distributed? I don't think it is. I think we need to understand this before we start talking about new reserach agendas.

New research participants is the other side of this question. Who are these new participants? Obviously the graduate students. But practitioners could be as well; so could policy makers. There are ways of bringing them in, and they have different skills, but you need to look at what utility they need from your community of practice. They will come into the network with agendas of their own. We need to ask what contributions we are expecting of them. Are graduate students expected to simply carry on your own reserach agenda, or to set their own.

Thinking globally, acting locally. It's a question of linking. But we can't just assume it happens. What is the purpose of the network, what is the point of sharing. Personal collaboration versus program collaboration. Knowing who is left out or left behind - even if we can't fix this, recognizing that this is there.

In the future we will need to look more at the evolution of unplanned networks. We need to look at adjusting expectations, discovering needs, building foundations. We look at email lists first, as the lowest common denominator - but even here you need to secure them. The sustainability of the sponsoring or home institution is important - it is hard to combine sustainability with limited resources and high expectations. Minimal standards of output require a sustainable base, a location where you know you can come back to in three or five years. And we need to look at permanent access to relevant outputs. That's coming from the academic community, but it also applies to the grassroots, who would like to know what has been done before, what worked and what didn't, in a secure environment where people are not going to be judging them about standards and reporting requirements.

If we don't make sure that we can sustain communities that already have an active base, creating new ones is an exercise that will just create something that will disappear. What do you do when you create a community on four-year funding. We talk about creating libraries, but not about creating them where libraries have traditionally been. Nobody talks about shutting down the university library. But we talk aboutcreating libraries where, four or five years from now, we'll say, it's not there anymore.

Gordon McCalla
University of Saskatechewan
Communities, Individuals and New Technologies: An E-Learning Perspective

I see a big relationship between individuals and communities, and it is at the level of the individual where the rubber meets the road. I believe relevqance trumps precision; you should be working on issues that are real issues. In scientific papers, climbing Mt. Everest is progress is in the goal of reaching the Moon. But you won't get there that way.

The key issues:

  • The fragmentation of knwoledge, culture, learning andteaching, and technology
  • individualization
  • context
  • change (and purposeful change)
  • granularity - there's all sorts of levels t which we know things - think about the three minute briefing a cabinet minister has before he makes a decision - he's not going to get into deep issues

The information revolution is getting underway. People think about technology as though it's done, then there's going to be convergence and it's all over. But it's only taking off now; even before the crash people said we would have a crash and then it would really take off in another five or ten years. All those predictions that people made, and failed, it's actually taking off.

I think that IT is localizing, not globalizing. There's so much out there we have to filter it out. We talk about spam, the whole internet is spam. You cannot absorb all that information. So I have this notion of everybody being in their own electronic village - your friends, your contacts, your organizations. Each person has their own. We have the village with the overlapping bigger world. I'm in a village, and a member of the AI community, and each other member has their own village, and through that community I learn new things.

Each of us has one of these things. Each one of us has their own unique village. Some people have villages that are similar to ours, they are the ones we meet at events like these. But the thing is, you really can't understand that easily information about communities you are not a part of. If I was trying to learn about mechanical engineering, I am going to need to be tutored or apprenticed. It's like Joe was saying yesterday, we're going to need diplomats.

That's the kind of world I think we've got. How do we deal with communication in such a world. Ben Daniel asks, how do you foster deep understanding among communities. Ben comes from Sudan; that's a deep issue in Sudan.

I believe that learning and teaching is fragmented in this way; you get a lot of information out of the context it comes in. Software, too: there's no such thing as a pice of software anymore. People used to think that you could 'prove software correct'. But there isn't independent software. There isn't even a boundary to your software any more (think, compilers, operating systems, etc.)

So here's the implications for the design of systems for e-learning or collaborative communities. First is support for the individual. I am looking into personalization. For example, your own personalized portfolios - for example, trying to standardize descriptions of people. You can contain information about, say, learning style. All of the information would be i there. The software would then adapt. Language. More subtle things. There is a lot of interest in the user modelling community around motivation. Motivation to be is deep understanding - that's motivating. It really doesn't have to do with emotions in the traditional sense.

Communities are extremely important. We all live in communities; there are many communities we are part of. We need tools to translate, negotiate, forge interactions between communities. And we need forms of community modelling.

Context is all important - I don't think there is meaning outside of context. Purposes, goals, tasks, why you're doing a thing, the people who are doing it, the reosurces being brought to bear - these are elements of pragmatics. Active modelling. If Google came back with the perfect website in three hours it wouldn't be very good.I think IT is forcing localization because we can't manage all the information that's out there. But if I can filter out all the spam, how do I get in touch with you? I think there's something to the idea of agents (which are like viruses, but "agents of good". Some piece of software that's loaded up with your user model that can act on your behalf, negotiate with other agents. Stephen's beeon online the whole conference, what could we do with information, what could we do with it for good?

My current reserach projects include I-Help, the ecological approach, annd LORNET. The ecological approach - what patterns can we find from mining user interaction.

The I-Help system is aimed at providing peer help to solve problems. The first pilot was in the Canadian prison system. It's important that they do things right, they have a very rigid system, which makes it a good place to test a piece of software. There are two aspects: open, public forums, and private one-to-one chat. Like Karl Goodman said yesterday, the more people we have online the better. A chat system with one person, or even a few, would help. But if we had tens of thousands of people it would be a lot easier to find someone to talk to.

You've got a personal agent, and you ask the agent to get you help, and the agent finds a resource. Or the personal agent does a Google search and finds an answer on the web. Or the agent is asked for help, and the agent decides that the best way is to find somebody to help, so it contacts a matchmaker agent, and the matchmaker finds three other agents, and negotiations commence between these other agents (an agent-agent negotiation), and finally the person himself is asked, and asked, are you willing to help the other guy, and the help session is set up.

Audience comment We've always had a problem describing expertise. Maybe you could use CVs. Response: CVs are not enough. To say that I know the concept of recusion - how well do I know it? The way I know it might be more shallow than someone else's.

We should think about rankings - like the ESP system talked about yesterday. Helping people is a way to learn things.

Me: this is the sort of thing that shouldn't be set up in isolation - being a part of this system should bea part of your activities in a more comprehensive space, the same spce you use for your blog, for example.

Te econological approah - can we capture end-use data. I'd like metadat to be user models. Keep in there who used the thing. Each person who interacts with a learning object, what their evaluation was, keep track that. Recommender systems.

LORNET - list of partners.

Comclusion: ICT is fragmnting and localizing. Really new technologies are possible and we're working them. Virtual humans are still humans and virtual communities are still communities (we need coffee right now, for example, very desperately).

Robert Gardner
Parliamentary Centre of Canada
Broadening the Research Agenda Through Virtual Networks

The experience the Parliamentary Centre brings is practitioners, in a parliamentary system. Building a networks is really the foundation for capacity-building, in a parlimentary system. The Centre's manadate is to strengthen legislatures (legislators?); you have to have strong legislatures (legislators?) to have strong democratic governments. So it provides aid and assistance to legislatures - on reseearch, on computers, and connecting with legislators from other countries.

I am interested in policy networks with Africa, to look at things like gender equity, and so on. There are logistical details here. But the idea is that these legislators build up expertise, and they take it back to their legislatures, and hopefully out of that better policy results, and so you end up with better gender equity, and so on. So that is the model, and we want to do this on a sustainable basis.

Now that's a community of practice, isn't it. Our path is different - we created these networks as a means of capacity building; it was the policy networks going virtual, but with severe technology constraints.

As a lens to look at this: SWOT - strengths, weaknesses, opportunities and threats.

Strengths - what I mean by that is starting points. It starts with tremendous strengths in Africal parliaments - the members are extremely enthusiastic, and want to solve problems, and their parliaments are also. They have a thirst for knowledge. They are open to new technologies.

But there is certainly a need for greater capacity - technologically they are weak. But using the networks is a useful tool for capacity building. The biggest weakness is limited resources. The digital divide is well understood. There's one ISP in Ethiopia, and it's often down. Connections are very slow. And then you run into money problems. Most members don't have their own offices, let alone computers or high speed connections. So if they are going to invest in that sort of thing, they need to find donors to do it.

You have to understand your users and their situations. Most of them will be using internet cafes. So they're not out of touch but they're not completely in touch.

Opportunities: so what's the potential of DCoPs in this context?

  1. Linking up members. Interesting: I reas really happy to hear that email is sufficient. What you want to do is make sure that somewhere in parliament there is one computer with a good printer and lots of funding for paper. I would put it in the library, but you can't impose that; technology is prestige in a lot of parliaments. In some it's the secretariat. In some it's clerks. So you can link up members for a relatively small investment. So that's a network that we established for a strategic purpose. We have also applied the same idea to how you link up other institutions in parliaments - legislative libraries, for example.
  2. Linking up Parliamentary Centre staff and activities. Much of our work is done in teams, but members are very often abroad, so we need sort of a movable virtual network. You have to have some information resources for them somewhere; they need access to the centre. You almost have to turn the home outfit into a virtual centre to cupport activities in the field. So I see the information management needs of the Centre as being a sort of movable virtual community of practice.
  3. Critical mass: all the people working in the different organizations - UNDP and so on. It's almost a netowk. Sort of like virtual matchmaking.

The Parliamentary Centre has these resources all over the world, and members probably don't talk with each other enough. I really like that idea of debriefing. The PC wants people to reflect more on what's working and what's not. But we talked about that - who wants to admit their mistakes. In career terms, who is going to say, "Well that mission was just a mistake, I didn't plan well, and so on." It's a cliche, but we want to set up best practices andso on. How can we be talking and dialoguing. And maybe there's another distributed community of practice.

Those are the opportunities. I'll skip over the threats.

Recommendations for action (note that as an outside consultant I am not necessarily speaking for the Centre here).

First, basically take up the challenge of the resource-poor countries, and think of ways the technology can be applied in poor and fragile countries. The idea of email is a good one to start with. It's a challenge that should be taken up systematically. It would be a pilot to look at these kinds of networks, and saying, what would be the tehnological support in these countries for building these communities, how would they work? Maybe set up pilot projects in some African nations, some good places to start. Somebody should provide appropriate funding to set up pilot projects.

Second, look at capacity building programs. Would the effectiveness of these programs be increased with DCoiPs? Probably. But do most NGOs have the capacity to use these, do they know enough about these. Would investing in DCoPs in NGOs be a good use of funds? Well, probably. None of the NGOs have time to play that knowledge brokering role that we've been talking about. Is that a role that somebody could play somewhere. A really good national library could provide a specialized research support for NGOs. Or maybe within and between NGOs. Maybe not enough NGOs know about that, about the potential of these tools.

Third, let's think about some sort of communication strategy to talk about the potential of these things.

Finally, to link up capacity-building practitioners. The DCoP, even though it's technology advanced, is already useful. So we could have sophisticated networks and databases to talk to each other if the politics of it allow. How do you build those links? The Parliamentary side isn't so hard; there are relatively few people and they know each other.

It's back to the knowledge creation agenda. Boiling down experience into some kind of principle, some kind of lesson. As capacity building people reflect on their successes, as researchers see broader patterns and build more tools.

SD comment: I want to challenge the idea of focusing on the leaders. First, because this merely increases the difference in capacity and opportunity between the leaders and those they govern - making the rich richer, if you will. Second, because the leaders have a substantial disincentive to adopt this technology (and there was much discussion about their nreluctance to do so) because of the disintermediating efect; when people are empowered, leaders become less important and less influentil.



Stephen Downes Stephen Downes, Casselman, Canada
stephen@downes.ca

Copyright 2024
Last Updated: Mar 29, 2024 07:24 a.m.

Canadian Flag Creative Commons License.

Force:yes