Thinking together: What makes Communities of Practice work?

Knowledge and knowing in Communities of Practice

In CoPs, learning is portrayed as a social formation of a person rather than as only the acquisition of knowledge. Learning entails change in one’s identity, as well as the (re-)negotiation of meaning of experience. In the original formulation of CoPs the main focus is on the person becoming more competent in the context of idiosyncratic practice (Lave and Wenger, 1991). The formulation of CoPs was founded within a postmodern framework that tends to be skeptical about the notion of knowledge (as a term), associating it with appointed (or self-declared) experts who ‘monopolize’ the possession and creation of knowledge as their source of power. This explains why knowledge is silent in CoPs, being approximated with the concepts of learning, meaning and identity.

Practice is considered as ‘a set frameworks, ideas, tools, information, styles, language, stories, and documents’ (Wenger et al., 2002: 29). According to Wenger (1998), CoP members’ negotiation of meanings in practice leads to the development of three structural elements of CoPs: mutual engagement (how and what people do together as part of practice), joint enterprise (a set of problems and topics that they care about), and shared repertoire (the concepts and artifacts that they create). In CoPs, ‘belonging is enacted through the mutual engagement, sharing of repertoires, and negotiation of the joint enterprise(s)’ (Iverson, 2011: 43), and for an individual it may take different forms across different communities, ranging from full participation (‘leading the practice’ by the core group) to more peripheral or occasional participation (Handley et al., 2006; Wenger et al., 2002). Thus, being a member of a CoP is not necessarily something that people are aware of. However, they do still experience a sense of togetherness when, often owing to facing similar real-life problems, and not necessarily because of liking each other, they organize themselves around negotiating a practice that they all share and identify with (Wenger, 1998).

Furthermore, Iverson and McPhee (2008) applied mutual engagement, joint enterprise and shared repertoire in their ethnographic study of volunteering groups, which was helpful in identifying the specific characteristics and differences between CoPs based on how people interact ‘in practice’ rather than based on the labels that may be externally attributed to ‘possible CoPs’. What is particularly relevant to our discussion is that their research provided empirical evidence that CoPs cannot be ‘set up’ as formal teams, and that to better understand CoPs it is important to pay attention to the nuances of the lived practice.

Thus, the work of Lave and Wenger, as well as other early contributions to the CoP concept (Brown and Duguid, 1991; Orr, 1996), paved the way for the current popularity of the studies of knowing-in-practice (Nicolini, 2011; Orlikowski, 2002; Rennstam and Ashcraft, 2014), which was labelled as ‘the quiet revolution’ (Easterby-Smith et al., 2000). In the spirit of this approach, to put it simply, knowledge is potentiality to act, while knowing is using what one knows in practice. Following this perspective, knowledge ‘sticks to the practice’ in the sense that the potential to act is developed in the social context, but it also ‘leaks through the practice’ when practitioners from different contexts learn from each other as they try to address similar real-life problems (Brown and Duguid, 2001, 2002).

Personal knowledge and indwelling

The foundation of reported research on personal knowledge and indwelling is that the understanding of knowledge and knowing in practice can be refined further by drawing more strongly on the contributions of Michael Polanyi who introduced a sophisticated conceptual model of human knowledge. Polanyi’s conception of personal knowledge offers a coherent view on knowledge and knowing that, importantly, is compatible with the essence of the CoP concept with its roots in identity. While Polanyi advanced considerably what is known about knowledge and knowing in the contemporary literature, there still remains much opportunity for building on his work. As Tsoukas and Vladimirou (2001: 975) write: ‘. . . no self-respecting researchers have so far failed to acknowledge their debt to Polanyi . . . [even though] Polanyi’s work, for the most part, has not been really engaged with’.

In order to unpack the role of indwelling in CoPs, it is important to understand Polanyi’s conception of personal knowledge. Central to Polanyi’s view of personal knowledge is the idea of the tacit component that is a necessary ingredient of all knowledge. ‘Personal’ implies that knowledge, in its richest form, can only exist within individuals and that it is necessarily grounded in the tacit dimension that people cannot easily say, as in Polanyi’s (1966b: 4) popular assertion that ‘we can know more than we can tell’. In other words, the tacit dimension can be thought of as the bottom of an iceberg that stands for the major part of what people know and that underpins everything that people know, and hence ‘a wholly explicit knowledge is unthinkable’ (Polanyi, 1966a: 7).

Thus, the tacit dimension to knowledge warrants that the personal coefficient is present in all knowing. Knowledge is developed through indwelling, which is an aspect of the knowing processes that accounts for learning (Polanyi, 1962a, 1966b). The process of indwelling captures the relationship of a knower’s body with the external world that they learn about as the experience of everyday life. From this perspective, a knower’s body includes rather than excludes the mind, and therefore indwelling applies to the development of both physical (e.g. sports) and intellectual knowledge (e.g. mathematics) – often at the same time. Peoples’ bodies, and thus their knowledge, is an instrument in relation to which they attribute meanings to the objects around them: ‘it is by making an intelligent use of our body that we feel it to be our body, and not a thing outside’ (Polanyi, 1966b: 16). Hence, people rely on their bodies, and so on their personal knowledge, while they attend to a focal point of attention in any given moment, as when surgeons dwell in their medical knowledge to perform a surgery using surgical tools, or pianists dwell in their musical knowledge to deliver a concert on a piano (see also Dörfler and Ackermann, 2012; Tsoukas, 2005).

So to a greater extent people dwell in a knowledge area, the more their bodies fuse with that knowledge area, thus their knowledge area becomes part of their extended identity. In general, such understanding of indwelling provides a considerable explanatory power; where, for example, it is possible to make sense of how Formula 1 drivers can legitimately claim that they feel the car as their body or mathematicians feel united with their equations. Specifically in the area of CoPs this understanding of indwelling offers more substance to the note of investing one’s identity in practice (Polanyi, 1962a, 1962b, 1966a, 1966b). Indwelling itself can also be shared but this requires putting trust in another person, as stated by Polanyi (1966b: 61):

In order to share this indwelling, the pupil must presume that a teaching which appears meaningless to start with has in fact a meaning which can be discovered by hitting on the same kind of indwelling as the teacher is practicing.

Although Polanyi mentions that indwelling can be shared, he does not elaborate on this aspect of the concept, as his focus is on personal knowledge, and shared indwelling is trans-personal. However, in CoPs the trans-personal dimension is essential, and thus we bring the idea of shared indwelling into the CoP concept. The notion of shared indwelling illustrates that people with different personal knowledge, but who manage to find ways to meaningfully attend to the same problems, can indirectly share their tacit knowledge by extending their identities into the same knowledge area. Indirectly sharing tacit knowledge in this sense means that each individual engaged in the trans-personal process of shared indwelling will (re)develop their tacit knowledge based on the experience of mutual performance in the shared lived practice. As they attend from their bodies to the problem, their shared indwelling becomes interlocked in the fleeting moment during which the extended identities of the participants overlap. Therefore, interlocked indwelling, as a trans-personal knowing process, can help understand forms of learning partnerships better.

We use ‘interlocked indwelling’ as a transitory concept, the role of which is to help us understand the essential knowing process at the heart of the CoP concept that we call thinking together.

Developing Communities of Practice by ‘thinking together’

Having discussed the notions of tacit component and shared indwelling, we now use these ideas to develop the concept of thinking together. We expect that the emphasis on the process of thinking together may help to gain a better understanding of the nature of CoPs and their fundamental learning processes, which are of high relevance to anyone interested in operationalizing the CoP concept. By developing a better understanding of thinking together we hope to provide practitioners with a useful point of focus for fostering such communities in organizational settings. In the subsequent argument, we develop the concept of thinking together in three steps.

Research design

The empirical study of CoPs was conducted in the NHS Scotland. We present cases from two different areas of NHS Scotland, namely: dementia and sepsis. The first case study describes a struggling CoP ‘to be’ while the second describes a thriving one; this contrast made it easier to observe salient characteristics, and it emphasizes the points we make. The empirical study had a qualitative character and it comprised of 29 semi-structured interviews or loose conversations with an average length of one hour each, and they took place in various hospitals across Scotland. The managers in the NHS Education for Scotland helped arrange the interviews with practitioners who expressed interest in the topics relevant to the study, and so a mix of purposive and snowball sampling was used (Biernacki and Waldorf, 1981; Teddlie and Yu, 2007). The practitioners were from CoPs at different stages of maturity and they were all healthcare practitioners (rather than patients, caretakers, etc.). The topics discussed covered social learning, their experience of CoPs, and the learning culture at their immediate workplace, with additional discussion on how each of these translated into better performance. While many participants had been aware of the concept of CoP, technical terms (including CoPs, thinking together, knowledge sharing) were presented and clarified during each interview when appropriate.

Following March et al. (1991) we attempted to ‘learn richly’ from this sample of practitioners by paying attention to the specific context of the particular CoPs, looking at the multiple aspects of the interview material, and thinking reflexively about alternative interpretations. The gathered data were rich and messy, and hence a method of analysis was used that could help to structure the data, make sense of it, and communicate the research results in a meaningful way, while not losing too much of its complexity. Our way of achieving this was through applying a causal mapping method in the analysis of data.

Causal mapping

Causal mapping was used because it was able to respond to the demands of idiographic data (the interviewing deliberately encouraging open responses). Causal mapping is a formal technique where a person’s thinking about a problem is modelled using directed graphs. The structure of causal maps emerges from an analysis of the interview material by identifying possible causal relationships of concepts represented by short phrases (quotes) that are linked by unidirectional arrows that represent expressed causality (Eden, 1992; Laukkanen, 1994). Various approaches to causal mapping have been refined over the years, including both quantitative and qualitative ‘styles’, but each approach is governed by a set of guidelines that need to be followed for the resulting maps to be amenable to formal analysis (Bryson et al., 2004). Causal mapping is well suited for structuring, coding and making sense of rich, idiographic qualitative data from studies concerned with the explorations of social practice, as it was the case in this research. Causal maps of this type represents action-oriented statements connected by causal links signifying beliefs of the interviewee about how their world works. In this study, the maps were developed, represented and analyzed using a dedicated causal mapping software (Decision Explorer ¹ ). All the maps, constructed for each interviewee, were based on the audio recordings of the individual interview, and those separate maps were subsequently merged where the meanings of statements appeared similar. This process allowed us to immerse ourselves in the recorded conversation, and to pay attention to non-verbal cues (such as the tone of voice). Thus, the process of mapping was as important as the final map because it ‘forced’ careful listening by the researchers.

The final merged map comprised of 1869 statements connected in a network of causal links. Decision Explorer (software that allows both visualization of parts of the map and analysis) was next used to identify possibly interesting patterns in the network by using a range of analytical functions: domain analysis (direct in/out links for each statement), centrality analysis (multi-layered domain), and identifying presumed vicious and virtuous cycles. These analyses identified patterns that were copied into NVivo where the fragments of interview transcripts were uploaded. Consequently, two models of the data emerged that mirrored each other’s structure: a model in Decision Explorer and a model in NVivo. Using both models it was possible to jump quickly between the fragments of causal maps, the analysis in Decision Explorer, and the corresponding parts of the empirical material represented in NVivo.

In the next section, we discuss the findings from the two empirical case studies.

Bringing dementia professionals out from isolation

The first case took place in the topical domain of dementia. In the UK, an important role in helping patients with dementia is performed by Allied Health Professionals (AHPs), a group comprising a number of professions specializing in supporting people, including dementia patients, through different types of therapy. AHPs include, among others, occupational therapists, speech and language therapists, art therapists and dieticians. These different professions did not naturally have mutual access to each other’s knowledge because of working in different locations and working with patients at different stages of their disease. A group of AHP leaders wanted to make a difference: they decided to bring dementia professionals out of isolation.

The AHP leaders believed that it would be beneficial to expose the dementia practitioners to each other’s practices and so reveal their otherwise inaccessible tacit knowledge. The proposal was intended to prevent ‘reinventing the wheel’ and allow for arriving at ways of doing things that seemed to work best for everyone, and contribute to seeing dementia from a more holistic perspective – as a journey comprising of different stages that all needed to be understood and looked after. In other words, the AHP leaders aimed for engaging practitioners in a shared practice; a description that seems to be a perfect fit with the CoP concept. The AHP leaders decided to ‘set up’ a CoP.

As a first step, the AHP leaders prepared a charter outlining what they expected from the CoP, what the benefits were likely to be, and also the required code of conduct for the future CoP members. Subsequently a discussion forum labelled ‘Community of Practice’ was designed, hosted by the health services’ library, fully open to the public with the expectation that in time it might reach a wider audience. As one of the AHP leaders later commented, this initial step to ‘set up’ the CoP had seemed relatively straightforward. Two of the AHP leaders became dedicated administrators of the discussion forum, their role involved uploading and organizing the content and monitoring the user activity. Moreover, to increase the recognition of CoP, the AHP leaders were promoting the discussion forum in informal conversations and in their email signatures. They also started to use the discussion forum to publish a quarterly newsletter about dementia that was based on the stories received from practitioners across Scotland about their day-to-day work. With time the newsletter became a success in the sense that the AHP leaders were receiving positive comments and emails not only from within Scotland but also from other parts of the UK and even from other countries. The newsletter did serve some of the AHP leaders’ original goals as it was promoting the work of dementia professionals and it was exposing them to each other’s practices. The AHP leaders were receiving more contributions from enthusiastic practitioners than they could possibly include in a single edition of the newsletter. Meanwhile, the sole purpose of the CoP discussion forum seemed to be to serve as one of the delivery channels for the newsletter, but there was very little conversation happening on its pages. As noted by one of the AHP leaders: ‘We try to encourage discussion, and that has not gone well. We’ve had people put questions out, and no answers. And we don’t know whether or not people are then replying outside of the discussion pages. They might be’ (Dementia Consultant).

When interviewed, the practitioners who had signed up to the forum typically explained that they had not been posting comments in the forum because of lack of time:

I think that it’s about managing your time. And actually allocating time. You know, so now I’m thinking: I need to allocate myself some time every week to go on to that CoP and just say that is the hour that I’m going to go on and I’m going to do that. (Alzheimer Scotland Dementia Nurse Consultant)

However, our analysis of the interview data (using the two software packages) suggested that the reason why people would not use the CoP site was that it did not provide them with immediate value to justify their time investment (the construct most central in the causal map model). In addition, they identified that there had not been conversations already taking place that they could join or observe and the range of topics had been defined in too general terms making it difficult to relate to more specific problems that could be of particular interest. Users would have been prepared to find the time to use the discussion forum if it had attracted them with something they perceived as immediate value, such as: engaging discussions, new working relationships, ability to share their views, solutions to their problems, opportunities to see what others are doing, and some tools, documents or techniques that they could use in their work. This observation is illustrated by the following quote, which is indicative of similar views expressed by most of the interviewees:

I think sometimes [in] that face-to-face kind of communication that you can have with your team . . . you [can] say: ‘I’ve encountered this problem today, what shall I do about it?’ . . . It’s immediate, and you get your response immediately. [Whereas] sometimes within the CoP you might post up a query, a dilemma, and there’s no actual guarantee that anybody will respond to it. (Dementia Liaison Team)

Furthermore, our analysis of the causal map identified an interesting vicious cycle (Figure 1): people can only submit their resources through administrators, which leads to the consultants publishing the newsletter, which leads to people feeling encouraged to engage in the shared practice via the newsletter and not through direct conversation on the forum, which leads to the CoP website being a place for resources rather than a place for conversation, which then self-sustains the cycle.

OPEN IN VIEWER

Figure 1.

Dementia case.

This vicious cycle suggests that while the newsletter seems to have addressed some of the consultants’ goals, the emphasis on the newsletter was also a possible distraction from fostering direct conversations.

Gradually, it became apparent from the series of interviews that the AHP leaders confused the discussion forum labelled ‘Community of Practice’ with the actual CoP. While CoPs can and often do make use of various online tools, it is very important to draw a sharp line between the CoP and the tools it uses – the dementia story illustrates this well. In addition, although the discussion forum could have enabled interactivity that would at least qualify for marginal participation, this interactivity never took off as the forum became simply a delivery channel for the newsletter. The marginal success of the newsletter was even making less visible the fact that there was no CoP, that, using Kuhn and Jackson’s (2008) terminology, it was knowledge development that users needed, but what was happening on the discussion forum was merely knowledge deployment in the form of requesting and providing information. There was a community of practitioners that cared about the same real-life problems, and they engaged to some extent, but did not mutually engage. There were no opportunities for interlocked indwelling on things that the practitioners genuinely cared about, and so thinking together could not take place. The point of creating a discussion forum was merely a step towards cultivating a CoP, but definitely not the moment of actually establishing one.

After their initial participation in this research, the AHP leaders thought that thinking together was a useful way to talk in concrete terms about what it takes to foster a CoP. Feeding back the research results (outcome of the analysis of the maps and NVivo) enabled the AHP leaders to easily understand why there was not as much conversation happening on the CoP discussion forum as they had initially expected. While they were happy about what they had achieved with their newsletter, and as they wanted to continue working on it, they also decided to start a small informal learning group among themselves that they hoped might evolve into a broader community. Additionally, the NHS Education for Scotland, who were closely following our research with the dementia CoP, incorporated our research results in internal documents, invited us to give talks at their events, and we participated in various meetings around the CoP topic.

Educating a hospital about sepsis

Our second case took place in the topical domain of sepsis. We concentrated on a team specializing in diagnosing and treating sepsis called the Critical Care Outreach Team (henceforth Outreach Team). The team and the hospital where they are based have been recognized both nationally and internationally for the quality of their work:

Analysis of the results has seen Borders General Hospital Intensive Care Unit record some of the lowest patient figures for out-of-hours admissions, length of stay, need for ventilation and need for renal replacement therapy in the country. On top of this the number of cardiac arrest calls at the hospital saw a remarkable reduction from 465 in 2000 [when the Critical Care Outreach Team was established] to 48 in 2013. (Healthcare Improvement Scotland, 2014)

At the foundation of the Outreach Team there was a need for improving the diagnosis and treatment of sepsis. It had been believed by its leader, a Clinical Nurse Specialist, that it was necessary to spread the active responsibility for diagnosing sepsis beyond intensive care (where very sick patients are treated). The reason for that need was that sepsis could occur anywhere in the hospital and therefore it was very important that as many practitioners as possible were confident about recognizing the symptoms early.

The Outreach Team comprises of five senior nurses who specialize in sepsis and who all have experience in intensive care. Not only is the team responsible for quickly responding to cases of sepsis in the hospital, but they also educate the staff in the wards about diagnosis and first response, provide them with supporting tools and systems, and help to improve their communication about sepsis. The importance of this education and communication was also evident from the analysis of the causal map. Moreover, analysis highlighted the facilitation of education about sepsis ‘on the job’ as the most significant (central) theme that enabled the whole hospital to develop its organizational ability to perform well in this area of strategic importance.

The range of regular actions of the team included: demonstrating to practitioners how to deal with sepsis ‘in practice’; mentoring junior doctors and junior nurses who are allowed to spend time with the team; organizing training courses about sepsis in the hospital, which are delivered by an interdisciplinary teaching team; designing objects that support interdisciplinary communication about sepsis such as small cards with key definitions, descriptions of symptoms and required actions, which are distributed among practitioners; and, convening interdisciplinary sepsis-related meetings where patient cases are discussed. As the leader of the Outreach Team commented:

We are a bridge between the intensive care and the ward areas. Historically, the intensive care was quite a secretive place. It was an inner sanctum that patients came to and then the nurses didn’t see them again until they come back out again. And there wasn’t any sort of joint working. And back where we started there was a nice term came out that ‘we should change and it should be critical care without walls’. Not physical walls, but metaphorical walls. And that was our starting point. There was lots and lots of different things. We got nurses who were in intensive care to go out and spend time in the wards to see what it looks like. And we got nurses from the wards to come spend time in intensive care. And that served a lot of useful things. People get to know each other. (The Outreach Team’s leader)

And then we saw an opportunity for another sort of learning: that if student doctors, student nurses, staff nurses came and spent time with us and see what we do, that would increase their learning. And to this day that’s growing and growing. So several student nurses as part of their training now they’re asked come and spend their time with us. And the student doctors as well. (The Outreach Team’s leader)

By the time we started our research there, the listed actions had become ingredients of the hospital’s sepsis-based practice, with a community of different types of practitioners organized around it. Practitioners from across the hospital identified themselves with sepsis because it could happen to their patients in the most unexpected moments. As a result, they genuinely cared about various real-life problems surrounding sepsis and they were willing to invest their time in learning more about it. Our interviews showed that owing to the Outreach Team’s work, people started ‘to come on board with sepsis’. In effect, the Outreach Team began to be seen as a leading group of a productive CoP (which had never been ‘set up’) with high impact, and with more peripheral members joining from various departments. As practitioners invested their identity in thinking together about what it meant to treat sepsis they not only acquired the useful facts and definitions but became competent in translating their learning into practice. The source of competence was the tacit knowledge that was being shared regularly among the CoP members. This sharing occurred partly in the ‘staff exchange’ between the wards and the intensive care and partly through the mutual engagement of the CoP members more generally.

I guess a lot of our success [of the Outreach Team] has been through education. A lot of nurses that see [the Outreach Team’s leader] in the ward, they learn something. [He] is a great teacher. (The Outreach Team’s member)

Our team is just part of a whole culture that’s changed. And maybe we have been a little bit of a catalyst in that change, or maybe instrumental. (The Outreach Team’s leader)

Although the leaders of the sepsis community had not been aware of the CoP concept, after we introduced them to this concept, and our use of the revelations from the causal map analysis, they agreed that it made sense to view themselves as a CoP. It is notable that even without any prior knowledge of the concept, the leaders cultivated a thriving CoP. The opportunities for interlocked indwelling on the same problems were given, and as genuinely interested practitioners they naturally engaged in building learning partnerships by thinking together. Furthermore, the members of the Outreach Team naturally emerged as core members through their mutual engagement. Owing to the core members’ outreaching activities, caring practitioners from various areas of the hospital started to engage in more or less intensive forms of participation, thus establishing the more peripheral layers of the emerging CoP. With regards to Kuhn and Jackson’s framework, the analysis of the interviews clearly showed that knowledge deployment was taking place in the form of mutual instruction and improvisation in the face of highly urgent, non-routine and problematic circumstances.