This is a write up (not review) of Architectures of Knowledge: Firms, Capabilities, and Communities by Amin and Cohendet (2004). It’s a recommendation from my supervisor for pre-viva reading but it also resonates with the Social Council project so I will be commenting on that in these notes as well. Health warning on this post – it’s a long one……
The book is a detailed examination of the nature of knowledge within organisations and the role of communities of practice and networks in its creation. The authors discuss the way in which communities of practice contribute to innovation and organisational success but at the same time reminds me of why I tend to reject organisational models which are based on market based thinking – though interestingly the authors end up in a similar place on this, pointing out the limitations of defining knowledge purely as a by-product of transactional information processing. They discuss knowledge as a practice rather than as a possession.
Its also something of a blast from the past for me as my masters thesis was based on an early community of practice at London Business School which I helped set up over 15 years ago. At the time the idea of social spaces online was a very new one and individuals were unlikely to be part of more than one of these communities – what strikes me as a read this book and reintroduced myself to some of the literature is the potential of more mainstream social behaviours to accelerate these communities within organisations if they are supported in the right way – and the fact that the boundaries of any community with a digital aspect are much more porous now.
Back to the book.
What is knowledge anyway??
The authors start by discussing the nature of knowledge and describe earlier research which either assumed innovation (defined as the main outcome of knowledge production) and knowledge are created by the individual (Argyris and Schon 1978) and ‘articulated and amplified’ by the organization, or that it is the result of a continuous dialogue between tacit and codified knowledge Nonaka and Takeuchi 1995). They take a different view:
Without denying the importance and salience of these two lines of research, this book argues, along with scholars who have pioneered work on learning in communities, that the time is right for research to explore the relationship between two other dimensions of knowledge in order to explain the innovative performance of firms: between knowledge that is `possessed’ (in different parts of the firm such as a functional unit, a department, a group of experts in a given domain); and knowledge that is `practised’ (processes of `knowing’), generally within communities of like-minded employees in a firm. (Amin & Cohendet 2004: location 56
In drawing out the centrality of innovation in a competitive context they highlight the fact that
as the generation of new knowledge becomes ever more pressing, the prime challenge of the firm ceases to be that of ensuring the coordination of existing bodies of knowledge. Instead, it becomes that of delicately matching the architecture of work with an always unstable architecture of knowledge that draws on the continuously changing capacity of interpretation among actors. (Amin & Cohendet 2004: location 74)
This phrase captures the transition of an organisation as being defined simply as a collection of transactions and processes to something which produces and ‘holds’ knowledge. It is this transition that I want to highlight and discuss with the Social Council team.
The authors phrase the same issue a different way later in the chapter:
Fransman (1994) has interpreted this critical change as a shift from firms conceived as pure `processors of information’ to firms conceived as `processors of knowledge’ (Amin & Cohendet 2004: location 105
They spend some time drawing out a separation of knowledge and practice and suggest that community provides a vital element in the process of knowledge formation rather than the more usual economists emphasis on codified and hierarchically sanctioned knowledge:
We identify community as the all-important site of knowledge formation; the site where hybrid knowledge inputs meaningfully interact. We share the claim by Brown and Duguid (1991: 53) that `it is the organization’s communities, at all levels, who are in contact with the environment and involved in interpretative sense making, congruence finding and adapting. It is from any site of such interactions that new insights can be co-produced.’ Accordingly, we assume that the process of generating, accumulating, and distributing knowledge-both in sites of informal interaction and informally constituted units such as R & D labs-is achieved through the functioning of informal groups of people, or autonomous `communities, acting under conditions of voluntary exchange and respect of the social norms that are defined within each group. Communities can be considered as key building blocks of the organization and management of corporate innovation and creativity. (Amin & Cohendet 2004: location 201
They repeatedly emphasis the need to combine practical and strategic knowledge and they do not separate the theoretical from practice or the scientists from the artisans:
Our aim in this chapter is to edge the knowledge-based perspective in a new direction, towards an understanding of knowledge generation and acquisition in firms as a weakly cognitive practice, or, more accurately, as the product of habits of everyday interaction in which thinking and acting are combined in inseparable unity, as are different types of knowledge, tacit and codified, mental, manual, and technological. (Amin & Cohendet 2004: location 923
By putting knowledge at the centre of the innovation, and as a result growth, process the authors are rejecting the view of knowledge as a byproduct that can be codified as they are identifying it as a creative and dynamic system within an organization that requires interaction and exchange rather than management.
Thus, according to Machlup, information is fragmented and transitory, whereas knowledge is structured, coherent, and of enduring significance. (Amin & Cohendet 2004: location 340
Knowledge also requires reflection and active learning by the communities:
Further, information is acquired by being gathered, whereas knowledge can be acquired by thinking and doing. Any kind of experience, accidental impression, or observation, and even an `inner’ experience, can initiate cognitive processes leading to changes in a person’s knowledge. Such differences underline the need to develop a distinction between knowledge and information that is not restricted to a simple stock or flux distinction, and also a framework that explicitly acknowledges the cognitive capabilities of individuals. (Amin & Cohendet 2004: location 340
This active view of knowledge rejects the idea that it can simply be codified in order to be passed on as tacit knowledge:
The assumption that codified knowledge can substitute tacit knowledge has been widely criticized, on the grounds that codified knowledge requires tacit knowledge in order to be useful. (Amin & Cohendet 2004: location 365
In terms of the balance between the doing of work and the creation of knowledge these two processes need to be treated as complementary to each other as is often the case within communities of practice. The authors point out that while codification can be at odds with the transmission of tacit knowledge it is still important to support the ‘elaboration of a common language and a collective representation of the community’. They suggest that:
In using the terms `strong’ and `weak’ rationality in this book, we are fully aware that there is a risk of thinking purely in terms of different `degrees’ of (the same) rationality. One alternative is the Habermasian distinction between strategic/instrumental and communicative rationality,19 although even this dualism does not fully recognize the powers of non-cognitive processes such as bodily knowledge. (Amin & Cohendet 2004: location 545)
There is some reference to networks as opposed to communities but this is not a book that takes a SNA view of the world despite having some resonance throughout with its ideas:
Through networks, agents can organize an efficient circulation of codified knowledge through a structure that renders compatible different segments of specific tacit forms of knowledge. Agents agree to increase their specialization in a given tacit form of knowledge, because they are confident that the other agents will increase their specialization in complementary forms. This arrangement reduces the risks of overspecialization, and relies intensively on building mutual trust in the production of knowledge. The degree of trust also influences the choice between specialization and cooperation in the production of knowledge: (Amin & Cohendet 2004: location 708
An how do we manage it?
Less relevant for the viva but of interest for the Social Council is a section on the role of management in the process of knowledge production as they suggest one outcome from taking a community approach to knowledge production is a reduction on the reliance on senior over middle managers to coordinate the process of change and to create a singular vision for the organization – this is very much inline with our approach on Social Council. It also reflects the need for managers to be active in the knowledge creation process
Among the theorists of organizational learning who are clearly sceptical about the leading role of managers in organizational learning, a recent influential voice is that of Senge. Senge, in defining the `learning organization’ as `a place where people are continually discovering how they create their reality. And how they can change it’ (1990: 12-13), sees only an indirect role for managers in the design of the organizations. For him, managers should limit their intervention to: (1) stimulating the adoption of `system thinking’; (2) encouraging personal mastery of their own lives; (3) bringing prevailing `mental models’ to the surface and challenging them; (4) building a share `vision’; and (5) facilitating `team learning’. (Amin & Cohendet 2004: location 837
This is not to say that they are rejecting a need for management in this process and they describe a need to create incentivisation and also to balance the needs of the knowledge production with those of the transactional requirements of the organization. In addressing incentivisation they are to some extent challenging a competence based approach to management if the real value a person might add could be in connecting other people’s knowledge and ideas.
There is a further management challenge if we accept or even embrace a central role for communities in generating knowledge and as a result innovation as these are dynamic and social systems that can be coaxed and encouraged but not mandated. Indeed, they touch on Dewey’s view that knowledge is a form of participation in the world and should be manifest with a physical change in the world and not simply with communication. For Piaget fans out there they also touch on the idea that `intelligence is internalized action and speech, and that both knowledge and meaning are context dependent’ (Nooteboom 2000a: 2). (Amin & Cohendet 2004: location 965
Communities of practice as a vehicle for creating knowledge:
Anyone who is active in a community of practice will be clear on their ability to create knowledge – what is significant in the argument that the authors are putting forward is the way in which they are seeking to codify this process in order to express it as a tangible organisational asset in the context of an organisation within the knowledge economy. Its worth considering at some point the relationship of public service to this meta-analysis of the knowledge economy – or perhaps to just put it to bed in favour of a network society analysis. In discussing community as a context for knowledge creation they acknowledge the role of social interactions in this:
Individual and collective learning is also the product of social practices that are not simply the sum of a multitude of individual neural nets. Communities in their specific social settings characterized by conventions of meaning and communication and the cultures of action and interpretation that are the product of social organization and interaction act as learning environments in their own right. While their dynamics too, as we argue below, generate knowledge as an embodied, iterative, relational, and practical act, this is not the product of the `conscious unconscious’ of neural computation, but of the `conscious unconscious’ of the embedded social conventions and social iterations in each setting. (Amin & Cohendet 2004: location 995
It is here that they start to explore the role of culture in the knowledge creation process and the need for careful reinforcement or creation of social norms is needed to facilitate this kind of approach.
Collins’s argument can be pushed further, towards a recognition of the influence of social practices in intentional actor networks on learning and knowledge formation. This has been one of the pioneering insights of the literature on the sociology of scientific knowledge-based on detailed studies of scientific and other knowledge communities-that has grown in influence since the 1980s. Philosophically, this literature shares with pragmatism the assumption that everyday activity is the powerful source of socialization, which, in turn, is the stimulus and carrier of cognitive activity. `Social practice proponents argue that knowledge-in-practice, constituted in the settings of practice, is the locus of the most powerful knowledgeability of people in the lived-in world’ (Lave 1988: 15). (Amin & Cohendet 2004: location 1004
They suggest five reasons for this
- Knowledge is not simply communicated between actors (human and machinic), but is generated through communication-speech acts, conversations, bodily gestures, glances, expressions, data exchanges, machine-to-machine interactions, are the relational iterations through which we know, understand, and learn. (Amin & Cohendet 2004: location 1011
- Sociality, from casual conversations to orchestrated social occasions such as conferences and formal dinners, counts as an important knowledge practice. It cements the trust and mutuality for tacit knowledge to be circulated, it can reinforce group feelings and identities necessary for shared knowledge conventions, it provides the serendipity for new knowledge encounters, and it allows ideas and routines to be tracked and modified. Importantly, its ambient vitality can actually stimulate and shape thinking, even when it consciously jars. (Amin & Cohendet 2004: location 1015
- Novelty is the product of connections, as Callon (1999a: 2) explains: Innovation is by definition an emergent phenomenon based on gradually putting into place interactions that link agents, knowledges, and goods that were previously unconnected, and that are slowly put in a relationship of interdependence: the network, in its formal dimension, is a powerful tool for making these connections, and for describing the forms that they take. (Amin & Cohendet 2004: location 1024
- It is usually through visualization of the elements of myriad interactions that characterize a given social ensemble, involving intricate links between humans and non-humans (for example, machines, chemicals, paper, tools), that knowledge paths are cognized and given effective power and influence. Without the complex interactive ontology of social phenomena, the need for cognitive signs would disappear, but then so would the basis of empirical knowledge. (Amin & Cohendet 2004: location 1035
- Fifthly, thus, it is the ordering and alignment of the ensemble of relations that make up a community of interest that generate knowledge in an intelligible and usable form-tacit and codified. It is the manipulation of social relations in service of a given project that counts. (Amin & Cohendet 2004: location 1048
Another distinction is the made between individual and organizational learning – the latter relying on social interactions to be effective:
According to Popper and Lipshitz (1998), organizational learning is not an extension of individual learning, because organizations and their members lack the typical means for undertaking cognition as a mental act. They suggest, instead, that organizational learning is the product of purposeful collective structures and cultures, rather than the work of particular cognitive dispositions at group or individual level. They emphasize the role of `Organizational Learning Mechanisms, defined as `institutionalized structural and procedural arrangements that allow organizations systematically to collect, analyse, store, disseminate, and use information relevant to the performance of the organization and its members’ (p. 168). (Amin & Cohendet 2004: location 1103
The authors make a clear link between their treatment of knowledge, the centrality of community and a move away from hierarchical organizational structures.
Distributed competences, rule-free and flat organization, social capital, employee autonomy, information sharing, connectivity, results orientation, flexibility and adaptability, continuous learning, and visionary leadership have become the new watchwords of knowledge-based success. Much of this new rhetoric is hype in the service of legitimating a new brand of business and concealing its power inequalities, but it does signal an entirely new way of conceptualizing knowledge production. (Amin & Cohendet 2004: location 1065
They also discuss the different and diverse forms of community which might exist within an organisation and that these might encompass different learning styles as well as different subjects, either within traditional work divisions or across the organization. They describe these communities as being “bound by relations of common interest, purpose, or passion” (Amin & Cohendet 2004: location 1128). Within this discussion they make the distinction between epistemic (dedicated to the deliberate creation of a specific area knowledge for example in an R&D function) and communities of practice which reflect to less formal exchange of information intended to improve practice and experience (the term is credited to Lave and Wenger (1991)). These communities of practice often meet informally and
“Wenger (1998) and Brown and Duguid (1991, 1998) state that self-organization is an essential characteristic of communities of practice.” (Amin & Cohendet 2004: location 1156
In common with the authors I would argue that this separation of knowledge acquisition models is not always helpful and that the singularity of the epistemic community can lead it to ignore hybrid or diverse ideas from the community of practice, which in turn can lack the methodological rigor often present in the former. It is more realistic to consider the two as complementary activities:
We will argue, in contrast, that, while the separations between deliberate and non-deliberate forms of knowledge and that between tacit and codified knowledge are useful analytically as a means of dissecting a complex knowledge process, it would be an error to assume that these separations hold in reality. While epistemic communities may be established explicitly as knowledge communities, the sociology of their knowledge practices is not radically different from that of communities of practice. (Amin & Cohendet 2004: location 1175
Within the context of the knowledge economy the convergence of these forms of knowledge community is increasing central to the success of an organization and the authors argue that knowledge based organisations need a culture which supports collaboration, narration and improvisation. They suggest (citing Callon (1999a) that experimental networks are a key component in the creation of new knowledge.
Network stability, in terms of composition, duration, replication, and length, is a key property. Emergent networks, in contrast, yield identities, interests, and competences as the outcome of `temporary and experimental translations [ traductions, in the original French]’ (p. 29). In emergent networks, actors with different knowledges come together into an uncertain venture, with ill-specified objectives (only experimentalism is the goal). In order to work, equivalence (linguistic and cultural) needs to be established between interests that are distant, incommensurate, unstable, and uncertain. `A sustained effort to interest, enrol, and form alliances is needed before the usefulness is accepted by agents other than A, agents who in the process of translation often change their identity’ (p. 32). In emergent networks, `actors are condemned to interaction’ (p. 32) and the organization of difference through translation and enrolment plays a crucial role in supporting the production of experimental knowledge. (Amin & Cohendet 2004: location 1249
At the end of this section the authors are arguing that the socialization of knowledge and learning is of greater significance, ie impact, than formal systems of knowledge management.
Does space matter?
The authors go on to discuss to role of ‘space’ and expands the definition of space as tethered by proximity or terriotory to a wider definition which encompasses digital space and networks.
(The chapter) does not assume that knowledge falls into bundles organized along neat geographical scales and contours (for example, that tacit knowledge requires spatial proximity while codified knowledge is ubiquitous or that knowledge externalities are spatially agglomerated). Instead, it defines spaces of knowledge and learning in terms of the traces of corporate organization and communication-that is, as organized spaces of varying length, shape, and duration, in which knowing is dependent upon the ways in which actants-human and transhuman-are mobilized and aligned in pursuit of particular corporate goals. The spaces are defined by the contours and forcings of actant effort and by organizational architecture. (Amin & Cohendet 2004: location 1313
Some emphasis is put in the advantages of spatial proximity to support internal and external innovation but the authors do not look specifically at the effects of social media and user generated spaces in this discussion. The authors specifically cite the role of proximately in the process of tacit knowledge sharing around practical skills without referencing the use of digital tools to spread these kinds of skills but they also acknowledge the role ‘distanciated ties and the organizational architectures and infrastructures that support them’ as being “highly significant knowledge spaces, involving forms of learning and a unity of tacit and codified knowledge that cannot be described as inferior or radically different from the putative powers of face-to-face presence and spatial proximity.” (Amin & Cohendet 2004: location 1418
They introduce the rather pleasing concept of ‘ba’:
Nonaka and Konno (1998) have suggested that the Japanese philosophical concept ba (roughly, `place’ in English) helps to highlight that the `shared space for emerging relationships’ is the `foundation in knowledge creation’ (p. 40). (Amin & Cohendet 2004: location 1429
‘Ba’ is discussed as having different facets of `originating ba’, `interacting ba’, `cyber ba’ and `exercising ba’ (Amin & Cohendet 2004: location 1434. This concept is used to outline how ‘being there’ is no longer constrained by physical presence.
The next section in the book is interesting with respect to the design of knowledge spaces but is not directly relevant to the viva so I am just dropping it in here for future reference:
The spatial strategies come in many forms that are hardly recognized in theorizations of the spatiality of corporate knowledge. One of these is the strategy of displacement. There are sites set up for new epistemic communities or projects that cannot be boxed into existing spatial arrangements, but that are crucial for generating path-breaking innovations. For example, Schoenberger (1999), who supports the new localism thesis that the multilocational firm can no longer ignore local corporate cultures as sites of learning, is quick to note that, when the firm `realizes it needs to change’, it may `consciously set out to create a new kind of place within the firm, including `organizational and geographical separation from the centre’ (p. 216). Project teams and task forces are a typical example, dislocated from usual places of work or established R&D centres, and bringing together project-centred specialists from different locations within the firm. Lee (2002), for instance, shows that some of the Korean conglomerates that have had to restructure and adapt radically after the East Asian crisis, are increasingly relying on such sites, where teams constructed from different locations in Korea live together for weeks or months to develop new ideas. (Amin & Cohendet 2004: location 1519
The corporeal space of displacement as a stimulus for creativity can be light and performative, which adds another interesting twist to the relational versus physical interpretation of knowledge spaces. Hatch (1999), for example, writes of `empty spaces’ in organizations, to distinguish spaces for creative action that are not regulated by rules and norms and act as working and thinking spaces that are not easily located spatially or structurally. Hatch argues that organizational creativity can be viewed in terms of the achievements made possible by ambiguity and emotion (and also tempo) exploited during the empty spaces of improvisation in jazz. Innovation in jazz occurs when the openness in a structure of a tune `permits any of the musicians to take the tune in a variety of directions’ (p. 85) and use of the ambiguity opened up by empty spaces in the tune. (Amin & Cohendet 2004: location 1532
Layout has become a visual symbol of how serious companies are about encouraging creativity, and is increasingly used in business chat and in the media to rank high performance and innovative companies. One trend, for example, is the replacement of offices by `hot desks’ that busy employees constantly on the move may book for a finite period. While this move has been driven in part by a desire to cut out wastage in the use of office space, it also aims to normalize circulation, exploit the possibilities of new desk partners and conversations, and inculcate a sense of attachment to projects and broader corporate goals rather than to a particular office and its trappings of security and ownership. Another trend has been the creation of common spaces such as atria, gyms, cafe offices, green spaces, streets, and squares-all within a building-in order to encourage serendipitous contact and conversations away from the heat of the moment, as a means of circulating information, sparking new ideas, and developing new socialities. This trend plays on the classical expectations of free social interaction, relaxation, and civility in widely shared public spaces (see Box 5.3 for some examples). (Amin & Cohendet 2004: location 1620
Knowledge management needs different management
The next chapter opens up the discussion of the relationship between management and knowledge management practice and points out the need to align the two. You cannot expect to fully benefit from the connections and potential of a community of practice without in some way acknowledging the importance of the behaviours that these encourage/require. This is not only with respect to community creation or stimulus but also the way in which participating staff are managed:
For Wenger, McDermott, and Snyder, community design should seek to evoke aliveness, based on actions that `bring out the community’s own internal direction, character, and energy, to find ways of helping an institution that is by definition natural, spontaneous, and self-directed, to `realize itself’ (p. 51). They recommend, thus, that communities should be helped to evolve in a natural and self-reflexive way, perhaps through the offer of open systems such as web sites, community coordinators, or problem-solving meetings; they advocate open boundaries, so that knowledge from the outside can help community members to see new possibilities; they suggest that communities should allow for differentials of participation among members, so that novelty and change are generated through varied inputs from core, active, and peripheral members of the community; they recommend that ways are found to help members realize the value of community-based knowing and action (for example, through efforts to make explicit or evaluate the value added from teleconferences or informal gatherings); they suggest that familiar collective events and gatherings are interwoven with exciting events, so that the community can generate novelty through both exploitation and exploration; and they invite managers to help communities to find a rhythm of engagement (based on a cycle of events, from informal lunches to web-site meetings) that is neither too fast nor too sluggish. All these designs are aimed at recognition and publicity, rather than the engineering of community. (Amin & Cohendet 2004: location 1767
By harnessing the energy and willing participation of community members there is a keen business benefit – but this must be balanced with a lack of control or enforced direction from management as this additional effort is only tapped if it is given freely. This is central to the management dilemma of any organization seeking to harness this kind of effect and in my view can be best managed by ensuring that managers are active participants in these communities from than taking an observer or supervision role if for not other reason that being part of the emergent language and culture that a strong community can drive. It is also important to remember the limitations of communities with respect to a tendency towards homogeneity and the risk of parochialism. To be successful with respect to learning communities of practice must have an external as well as internal focus. There is a conflict between community/network and hierarchical structures that needs to be balanced if purpose/invocation is also to be balanced.
Incentivisation is critical and it has been argued that, if the producers of knowledge cannot appropriate the benefits of new knowledge, then they have no incentive to produce it. (Amin & Cohendet 2004: location 2129
The authors cite the need for ‘soft infrastructure’ such as away days, clubs, social events etc etc as a way to help balance these two strands but we can also consider language sharing and individual connectors as a way of mitigating conflict or maintaining balance. The authors also contrast communicative and hierarchical cultures as forms of governance. This is a whole area of thinking to be considered with respect to the Social Council.
To summarize, the governance of the firm as a community of communities seeks to benefit from the diversity of interactions between communities (the innovative `sparks from interacting communities, as described by Brown and Duguid). It seeks to bridge the hard architecture of learning, which exists in many visible forms, and the soft architecture of learning. (Amin & Cohendet 2004: location 2012
The final analysis is of the danger of treating communities of practice as a tool which can be managed by a hierarchy instead of a social and organic effect of like minded individuals seeking to improve their own knowledge.
Our analysis in this book suggests that policy attempts to steer the internal dynamics of firms and networks will not work, as they are too remote from the precise, grounded, and evolutionary processes behind knowledge formation in individual contexts. But, this is not to say that certain general principles of public policy action cannot be identified. For example, a product-specific, cluster-specific, technology-specific, or know-how specific approach to science and technology policy risks missing more than it captures, and may be less preferable than a programme of sustained generic support-through generous and long-term investment in universities, technical colleges, public research institutes, basic science and technology programmes, arts, media and cultural industries, and centres of experimental and future knowledge. Such a programme would help to secure not only a varied ecology of knowledge but also a foundation for emergent, new, and unanticipated discovery. (Amin & Cohendet 2004: location 2084
The authors conclude by how far policy making is from the identification of communities of practice as being critical to the creation of knowledge and as a result innovation. I would argue that this is still true (though recent FP7 funding did make this change in emphasis). They also reference the need to build trust between economic units in order to create wider knowledge sharing spaces and its interesting to note that this should, theoretically at least, be easier to achieve within public services. The authors make a number of recommendations of public policy changes which could positively influence the process of knowledge creation but these seem weak in the face of their suggestion that these communities are most influenced by organisational culture. They also recognise the danger of government intervention in these kinds of social constructs and this is critical if we are to accept their view that knowledge must be considered not only as a possession but also as embodied practice.
Its useful to have a reminder of the complexity of community and the fact that it is not a term that can or should just be thrown in with a vague reference to civic (more on this in the next literature review post on a Peter Dahlgren book). The authors also provide a different way of viewing the clash which can occur within an organization which comprises networked and hierarchical aspects. Within the book the arguments for more networked behaviours is clear in the event of innovation being an explicit organizational goal – and this is almost a given at this point for any organization which relies on technology to mediate its relationship with the public. What is left open is the role of organization leaders in this process and this is the thought which I want to take to the Social Council team for discussion.
Leave a Reply