Learning from experiences: Its limits? Its future relevance?

Original Message

From: Corinne Sprecher, posted on 2015/03/06

Hello KM4Dev Community! I’ll try once more, as it seems the message did not make it into some/many mailboxes last time. I trust there are quite some well informed and visionary people on this list and I would thus like to explore with you the limits of learning from experience and implications for KM of development agencies in the future.

Let’s put it this way: KM practices have put a lot of emphasis on learning from experiences, e.g. in commonly known methods such as experience capitalization, lessons learnt and good practices. At the same time we have become aware of the limits of learning from past experiences, just to mention some:

• Significance of the context: Can experiences be transferred to other contexts?
• Success – scale – fail syndrome: Replication and scaling up of experiences made by others and solutions of others often lead to failure.
• Lack of experience and urge for new solutions: Complex future challenges in development cooperation and in partner countries sometimes cannot effectively build on existing knowledge of development cooperation.

As Einstein has put it: “Problems cannot be solved with the same mind-set that created them.”

Do you feel limits of learning from experience in your work? What is your approach to dealing with (or side stepping) these limits? What relevance do you think will learning from experience have in KM in the future?

Actually, this query is part of a work I do for the Swiss Agency for Development and Cooperation (SDC), gathering some ‘food for thought’ for the future orientation of KM and organisational learning at SDC.

Looking forward to reading your answers and engaging in a discussion!

Kind regards, Corinne

Contributors

All replies in full are available in the discussion page. Contributions received with thanks from:

Eva Schiffer
Ewen Le Borgne
Jaap Pels
Valerie Brown
Peter J. Bury
Ueli Scheuermeier
Camilo Villa
Pete Cranston

Related Discussions

• Sharing Lessons and Worst Practices (30 January 2009)

• How to collect and present Lessons Learnt (2 December 2011)

• Lessons Learned - The Loch Ness monster of KM (18 July 2013)

Summary

To start I will state some limits of learning from experiences of others that were mentioned in the discussion: We experience limits… - when moving forward on complex issues. - when trying to short cut (social) learning processes. - when we separate “knowledge” from “practice” / “doing”. (the art is to find “conversations that wait to happen” and light them). - when we get trapped in an “old-fashioned”, distorted comprehension/model of learning as teaching, instruction, transfer… - when we approach a new context assuming we know the formula on how to “fix” it. - because we do not declare loudly and clearly enough what it is that we really can say we know. - when we focus only on recording/ knowledge products and do not integrate reflexion and learning processes about the products. - when trying to (or fearing of losing) control of the learning.

Below a summary of the discussion along the following themes: - Dealing with Complexity - Models of Learning, Social Learning and the Illusion of Managing Knowledge - Communities of Practice and (Campfire) Exchange - Working in Networked Ways - Learning and Feedback loops

Dealing with complexity Eva Schiffer argues that there are strong limits as to how we can learn from other people’s experiences when it comes to moving forward on complex issues. These require those involved to let go of old beliefs, change behaviours, go through difficult transition. What stands often in the way is a misunderstanding what technical and what adaptive challenges are. (See: adaptive leadership by Heifetz) A technical challenge means that the solution is known by an expert, the challenge is to find that expert/solution and implement it. An adaptive challenge means that the solution is not known and needs to be co-created by the people who have the problem. Worst use of case studies or lessons learnt is to have an adaptive challenge and throw technical solutions at it. Reference: Heifetz (1994) Leadership without easy answers.

Ueli Scheuermeier adds that as soon as things become complex, instructive teaching and learning fail. An innovation/experience has to be deconstructed into its various aspects and then reconstructed differently to have success in another setting. This requires a creative and explorative exchange in action between people who together are working to figure something out. He puts forward that Communities of Practice (CoP) are such groups that allow this kind of de- and re-construction processes. Riff Fullan underlines the necessity for reinterpretation and adaptation of what has been learnt in one context when taking it to a new context and he stresses that the new context could actually be quite close to the original one, both geographically and socially. He thus proposes to take an approach that does not assume you already know the formula. Instead you need to build a common understanding and a common position and co-create solutions.

As a concrete example, Sam Lanfranco, contributes two lessons from a research on the roll out of the Internet across several Argentine provinces: (1) One lesson was that the key animating unit (agency/person) needs to engender trust across all participating stakeholders, and this is based only a bit on past record, or current expertise. It depends crucially on the dynamic within the process and across all stakeholder groups. (2) Stakeholder engagement is crucial not just in decision making (planning, implementation, monitoring and evaluation) but in the process of generating “policy windows” as a prelude to specific policy dialogues and subsequent collaborative decision making. In short: trust in leadership, and participation in generating the policy windows where policy issues are to be addressed.

For all who want to dive deeper into the theoty of complexity, Jaap Pels, points us to the work of Max Boisot among others on the ATLAS experiment at CERN around the question of how to organise for an emergent, creative process. At the core of his work is a conceptual framework called I-Space Framework ([1]; it influenced the development of the Cynefin Framework by Dave Snowden). The framework refers the level of codification and abstraction of knowledge (experimental knowledge, narrative knowledge, abstract symbolic knowledge) to its diffusibility (‘how it flows’). The figure below shows the effect of ICT: Boisot refers to the diffusion effect of ICT on the one hand, meaning that more people can be reached in the same time thanks to technology. On the other hand the bandwidth effect also become evident: With more bandwidth also less codified knowledge can be diffused wider. This allows quite large “clan” structures to emerge even distributed far geographically.

See also: Remebering Max Boisot: [2] Minutes of a Presentation by Max Boisot, London, 23 July 2010: [3] Video of a presentation on the Atlas experiment : [4]

Models of learning, social learning and the illusion of managing knowledge

One part of the discussion evolved around mental models and the understanding of learning, social learning and knowledge management we are confronted with in our work. Etienne Wenger has worded it like this: Formal education has given people distorted models of learning. We need to rescue learning from this and restore its fundamental role in human life.

For Camilo Villa, learning is a combination of: (imagined in a triangle): Knowledge (what you already know) + information (data you receive/share) + experience. Learning in daily life happens by combining the three aspects (whereas in education there is often a focus on knowledge and information). Experience thus continuous to be a must.

KM should evolve from the failed illusion of pretending to manage knowledge to a more humble approach, aiming to create learning experiences and harvest the lessons derived from experience. We can somehow control the settings and the atmosphere surrounding an event but we cannot control the knowledge that would circulate or evolve. We can translate the learning process into knowledge products but we cannot capture knowledge itself.

Arthur Shelley quotes John Holt: “Learning is not the product of teaching. Learning is the product of the activity of learners.” Learning and knowledge creation can’t be outsourced! But we learn faster from prior knowledge. No need to remake all mistakes made by others. This requires critical thinking to ensure that prior learning and knowledge is: - Robust (evidence based) - Relevant (even though being adapted to your new context)

Learning from experience has always been the most effective way to develop knowledge in an applied way. I don't believe our role is to manage knowledge, it is to facilitate its creation and flow for greater future adaption and application (so more knowledge leadership is what helps). By creating open environments and trusted relationships we facilitate flow (application & adaption) of both knowledge and experiences that create our future (and by locking up knowledge and limiting experiences we limit our future).

As I say to my learners - my job is not to fill your heads, it is to open your minds.

He also points out his article on learning how to apply knowledge: [5]

In the words of Eva Schiffer: Social learning processes can’t be cut short. People involved need to co-create solutions and “do the hard work” themselves. Stories of others can be inspiration, showing us that it can be done. Not learning from books but learning by doing. Examples of others can help you explain what you are struggling with and/or inspire.

The discussion about the misnomer of KM goes in line with a survey of APCQ with a number of KM professionals on: What will make the biggest near-term impact on KM? ([See: http://www.apqc.org/knowledge-base/download/340786]). Jaap Pels pointed out particularly one of the conclusions of the paper: “KM practitioners will have to change too... The KM practitioner as a quality improvement, risk management or operational change management driver is where the biggest change will come. … KM needs to move toward ‘a primary focus on organisational learning’. … ‘the term ‘knowledge management’ today is used as a synonym for knowledge sharing, without any context to the larger requirements and many phases—including innovation—of organizational learning. … ‘knowledge managers need to pay attention to the emergent application of ideas from cognitive science and behavioral economics.’”

In the words of Ueli Scheuermeier: What does the art of facilitating such collaborative learning looks like in an environment that goes nuts on IP? Maybe KM is on the way out because it increasingly becomes a misnomer. Maybe its LM (learning management?) of which KM is just a tool.

Camilo Villa adds to the discussion about mental models of learning that most people are not even aware about HOW they learn.

This points us to a need for models of learning, such as the Framework for Social Learning and its ability to transform practice proposed by Beverly Trayner and Etienne Wenger. It is a broader perspective on what effective learning entails. See: [6]

Bev Trayner clarifies that the framework for social learning not necessarily implies f2f learning events. She outlines an example of a wiki and how the framework helps, to conceptualize it. What aspirations do we have about the value creation of the use of this wiki? What indicators can measure the different kinds of value. What conditions need to be in place that will help us achieve those aspirations? Aspirations, indicators and conditions can all be negotiated with different stakeholders and provide a framework for evaluating how we have done, or what we have learned in one year’s time. She points out that according to the article on “collaborating for impact in large development organizations” by Nancy, Rachel and Aldo in the latest KM4Dev journal ([7]) the organizations find it hard to answer the question “What do you want to evaluate and what are your indicators of success?” This is what this framework is about!

Valerie Brown adds the classical experimental adult learning cycle to the discussion: values > facts > ideas > actions and points out that every step needs recognition of complexity and an openness to change. There is no silver bullet, no one tool that fits all.

Jaap Pels also puts forward a different way of looking at development, learning and organisations. In the sense of: The metaphor you use will guide the solutions! Think of development (and thus learning) as people having conversations with each other. Think of SDC as a collection of conversations by people. Thus for the future orientation of KM and organisational learning (OL) at SDC: - What primary process is staff talking about (why SDC is on earth, raison d'être) and what is learning about here? - What admin (procedures, finance etc etc etc etc/ secondary process) is involved and what is learning here? - Do organisations learn? And if you think of SDC as a collection of conversations; is OL relevant or how does it shape? - Is SDC knowledge not the total 'SDC staff / partner / clients capacity' / a network? - Is KM about what happens in SDC's people's network or about moving around explicit stuff (e.g. information) How about the thought that learning can only stem from experience and SDC - any organisation - is there to accommodate such.

Communities of Practice and (Campfire) Exchange Riff Fullan comments: learning from experiences of others – and learngin from own experience: This is not an either-or-scenario. What we aim for is to get different experiences, different ‘knowledges’, come together in a dialogue. In this case the own experiences are also always part of the mix. This way a mutual learning scenario is created.

Ueli Scheuermeier: Knowledge in Communities of Practice is often operational and not just analytical. “Doledge” (see [8]) as Ueli names it. It means all that is required to actually go out there and make a difference in the real world, as opposed to knowing about the world. Peer exchange learning allows people (COP members) to deconstruct what works in one place and reconstruct it in another place or setting or even theme. This can be powerful, and fast, and actually bypass hierarchy and reporting loops. The role of HQ becomes thus one of providing an environment that allows peer-exchange learning among its frontline people, observing and documenting what is going on, and drawing conclusions form that on what it may mean for managing the whole organisation. In his view the whole idea of classical “reporting” for any purposes other than administration is in question. The most fascinating exchanges that I was involved in or that I was observing were indeed “conversations waiting to happen”. Someone comes along and gets it going. Questions: How to find these? How to get them going? How to keep them going? What skill-set is needed?

Pete Cranston points out that conversations in CoPs do not always have to arise spontaneously, they can be triggered and managed. Good Example: Sanitation CoP run by SNV (Contact: Petra Rautavuoma, WASH sector leader)

Pete Cranston dug in past conversations on this dgroups and found discussions on doledge from 2004, 2005, 2008 and 2010. In 2005 Steve Denning for example pointed out: “Stories when written down become artefacts. They are stories frozen or congealed in time, and as such have a tendency to die, unless they also happen to be great works of art.” Pete points out the importance of story tellers and today’s community of story tellers: the star bloggers, youtube and vine artists! He points to the ILRI web content, for example ([9]). The videos, blogs, shared presentations, streams of tweets, pictures, project event wikis and social network chatting models of how stories of change can be captured and shared so their stories meet our criteria that people are triggered to talk and discuss. (Good example of a modern form of the Alexandrian library?) On the other hand the challenge remains that we know also what works best in sharing and learning and that is conversations, preferably face-to-face, often best if facilitated so that all those involved can lose themselves in the process. This is however expensive and the content almost impossible to capture (see also this blog post Blogpost on the difficulty to capture conversations meaningfully even for people being present: [10])

Pete thus proposes the following view on the dilemma between the “Alexandrian Library” and the “Campfire Exchanges”: Is there any other way to meet our complex world than becoming better learners, socially and individually? And aren’t conversation and exchange processes the stepping stones to people becoming better learners, more able to analyse and discuss meaningfully with others, and above all to learn to trust their judgements on the basis of that exchange and shared experience? The essence of social learning for him is good, diverse content and rich conversations combined to generate an environment in which people learn how to learn, reflect together, and make judgements that they trust. Isn’t that enough? And isn’t that embedded into the best organisational practice?

Christina Merl summarises her approach to complex, networked learning as follows: - to be as transparent as possible. - to understand and consider the perspectives of stakeholders while serving the needs of the network members. - to build in systemic feedback and reflection (learning loops) As long as there are people out there who want to make a difference, learning will find its way. One challenge she is facing, is the co-ownership in a network. A situation where both, sponsors and members are “happy” and ready to contribute as they both see a benefit. She observes that sponsors tend to want to take control of the process (learning, outcome, change), which discourages practitioners from contributing and sharing. Thus, practitioners go away while more and more people are attracted who like to discuss things on meta level. For practitioners this is not relevant, a waste of time for practitioners. They want to exchange on practical experience, high quality peer advice. Others enjoy discussion on meta-level but nothing happens afterwards. She summarizes: Somehow I think that, once the traditional (business-based) evaluation and control effort sneaks in, a lot of dynamics and motivation get lost.

Working in networked ways

For Riff Fullan another way to encourage learning from experiences is encouraging more networked ways of working. One element of which is “working out loud” (John Stepper see [11]). It means providing a window into your work. This gives others the opportunity gain insight in what you do, how you do it, etc and learn from it; and on the other side it provides the possibility for getting and giving each other feedback on a regular basis (multidirectional feedback). He sees multidirectional feedback and working out loud as part of working in networked ways, and possible alternative ways of learning to CoPs, which tend to be more rare and more structured.

Learning and Feedback Loops Bev Trayner took up a question about the roles of reports (that often do not get read) in learning loops in this way: Apart from anything else a learning loop is a continuous process and a report is not. It is important to pay attention to to how and where we create these continuous learning loops: Time in the agenda, or space on a platform, for stories of how members tried to apply something and whether or not it worked is one example of a learning loop. Why don’t we throw that question out to everyone - how do you incorporate learning loops (that are not reports) in the projects you are engaged in? Nancy White adds another dimension to the discussion on written learning products by saying the value of written products is often highest for those who write them as a process of getting clear.

Pete Cranston comes back to the role of the Alexandrian library for learning. In his words: We won’t – can’t – throw away the Alexandrian library since we have to find ways to record what has happened, and what we think we have learned (which is as often in novels, or stories, or videos). He illustrates the need for learning through an example of an approach that did not work but was still applied repeatedly; also in the same organisation. In his view the problem is that we do not declare loudly and clearly enough what it is that we really can say we know. We are often too tentative, too vague, delivering high level bullet point recommendations. To claim that one knows something is quite bold.

The criteria for whether something useful is produced (e.g. out of a failure) is that it makes people think. Meaningful outputs that might enable people to learn across contexts are those that require people to talk, question and reflect on the basis of what they read/hear/see in the documentation.

The question is how organisations are able to put in place processes and develop a culture in which there is a balance between meaningful recording of experience (reporting, documentation) and a continuous process of reflection and exchange (learning) about those products.

For example lessons learnt databases: what social processes do organisations that have such have in place about constructing and sharing the lessons, which make them live a bit more?

He adds some final statements on what he knows he can say about feedback and learning loops: 1. Individual difference in terms of learning and behaviour are important determinants of how people work. This is not much acknowledge in organisations. Activists get rewarded for what they do, reflectors much less. 2. Most people in organisations won’t use social media regularly and consistently for work 3. Who does use social media is random and a function of busyness and type of person see 1. Extroverts versus Introverts, the later need a controlled, safe and probably closed environment. 4. People do read, listen to and learn from social media. 5. It takes communication and KM people to make things happen. 6. It is and always has been about conversations. Thus facilitators are a necessary and cost-effective investment too, as above.

Contributors to the discussion: Valerie Brown, Peter J. Bury, Pete Cranston, Riff Fullan, Sam Lanfranco, Ewen Le Borgne, Christina Merl, Jaap Pels, Ueli Scheuermeier, Eva Schiffer, Arthur Shelley, Robin Van Kippersluis, Camilo Villa, Beverly Wenger-Trayner, Etienne Wenger-Trayner, Nancy White

Thanks a lot!