posted 1 Jun 1998 in Volume 1 Issue 6
J. Seely Brown and Paul Duguid focus on three issues relating to the organisation of knowledge. They consider some of the contradictory features of knowledge; how it can be intangible, ‘sticky’ and ‘leaky’ at the same time concluding that we should be looking to combine the material with the virtual world to succeed with organising knowledge.
Discussions of organizing knowledge concern themselves primarily with one of three issues:
1. How to produce or generate something almost intangible.1
2. How to manage and coordinate something that, in von Hippel’s term, is ‘sticky’.2
3. How to enter the market with something that is, on the one hand, hard to commodify, yet, on the other hand, often ‘leaky’.3
Together, these three appear incoherent. It is hard to imagine anything that is simultaneously intangible, sticky and leaky. From the perspective we present here, however, these differences reflect less incoherence than different stages in the production, development and use of knowledge.4 They also reflect distinct perspectives on those different stages. Issue one reflects the problem of locating useful knowledge production - a problem of particular interest to the folks in R&D; two reflects the problems of strategic management and IT in dealing with inchoate knowledge; while three reflects the organizational and economic challenges of controlling knowledge as it becomes progressively more robust. Recognizing all three together - and the feedback loops and tensions they produce - are among the central demands of organizing knowledge.
Some knowledge is inextricably communal in its production. Knowledge-based arguments about the firm lean heavily on Polanyi’s notions of ‘personal knowledge’, and in particular, ‘tacit knowledge’.5 This sort of knowledge, similar to ‘know how’ rather than to ‘know what’,6 emerges not through detached reflection so much as through action and participation. The dynamic process of knowing has both epistemological and organizational significance.7Implicit, tacit and dynamic, this sort of knowledge resists explication, commodification and exchange - either in or out of the market. It cannot simply be parcelled up and passed from those who know to those who learn. From an individual’s perspective, the process of acquiring knowledge involves not ingesting or receiving objective knowledge, but becoming a member of a community of practice and coming simultaneously to share and participate in its collective knowledge.8 In terms of learning, moreover, practice is not simply a method for inducting new members into community membership. It is also the way communities learn about the world - the way they produce new knowledge. Given humanity’s social character and the way people divide labor among themselves, the bulk of new human knowledge arises through social practice as people work together on shared, collective and coordinated, but divided, tasks. Here, then, we depart somewhat from Polanyi or Nonaka and Takeuchi9 in suggesting that not all knowledge has its origin in individuals. Important knowledge and knowing are collectively produced and held.
This social process of knowledge production involves collective sense-making and judging as people, by engaging in densely shared practice, develop collective background assumptions and methods of investigation and validation. Developing practice changes a community’s world view, which in turn provides new perspectives on practice. The process is a dialectical one, with the new practice being shaped by the community’s assumptions and vice versa. In a parallel fashion, the community shapes the practice and, reciprocally, the practice shapes the community. The practice, the perspective and the community are, in certain ways, mutually defining, inseparable and dynamic.
Consequently, the complex new behaviors of particular communities can, on the one hand, be almost impossible to replicate outside the community and, from the outside, easily appear abstruse, absurd or trivial. Von Hippel suggests the problem is that ‘one does not know in advance which subset of… information will be relevant to anticipating relevant features’.10 To which we would add, that such behaviors can be so profoundly implicit that they do not even emerge as information, relevant or irrelevant. They are simply practice. In developing knowledge, the routes a community of practice follows, the standards it sets, and the warrants it uses are inextricable from practice. When the knowledge is inchoate, it is often inexplicable without engaging in the practice.
Overall, as communities develop new practices, knowledge develops - but not as an aggregate of individual contributions, like adding bricks to a wall. It is more like the mixing of color, where the outcome requires several contributions, but the process and product dissolve individual contributions indivisibly. Organizational synthesis of knowledge, then, may not begin with a step from individuals to groups or communities. Rather, it may begin with communities. Synthesis involves the articulation into a larger whole of the knowledge of distinct communities - going from palette to composition (to continue our painting metaphor).
Organizational learning in practice
The ways that organizations go about learning vary dramatically. Many look on the challenge as one of gathering information about a detached set of circumstances. They conduct market surveys or they monitor incoming customer service calls. They analyze the results, build them into databases and pass these out to the people in production, sales and service, for instance, who then modify their routines to reflect the customer needs. This centralized, detached information-gathering and top-down dissemination, while clear and simple in plan, places too much faith in individuals and information as the focus of learning. Social practice, with its awkward dynamics, is almost entirely missing.
Yet organizations are never detached from their environment, and that environment is rarely static. Indeed, a great deal of organizational activity or ‘enacting’11 aims directly at engaging in and stirring up its environment. The same organizations that attempt to monitor the customer base are usually involved in simultaneously influencing it through new products, better service, marketing, advertising and sales.
Moving from a top-down, centralized to a bottom-up, diffused approach turns things around in several ways. It may limit the scope of inquiry, but it can increase the depth dramatically. Instead of starting by detaching information from the context that makes it valuable, this approach tries to develop knowledge out of engagement. It also relies on the people who are actually engaged in the practice to develop the knowledge.
Xerox’s ‘Eureka’ project illustrates a process of organizational knowledge production in situ. It not only relies on situated knowledge of customer activity, but it deploys the very people who must work in those situations to develop that knowledge.
Previously, Xerox had taken an ex ante approach to customer service. It provided service technicians with detailed instructions of how to deal with machine behavior. Unfortunately, the machines did not obey company procedures and, in the end, the only way for technicians to operate was to disobey them too. The technicians developed their own, non-canonical practices to handle the emergent and unanticipated in the machine and the needs of the customer's community.12
The Eureka project has helped to consolidate this turn around from top down to bottom up.13 It has provided a dynamic central database of problems and service tips to capture new problems as they have emerged. Such databases are not uncommon. But often they flood with information of varied provenance and value. Once they do, the use value of the database as whole drops precipitously. By contrast, with Eureka, technicians themselves build and validate what goes into their database. As a consequence of this filtering, content reflects the leading edge of the collective understanding of a broad community of technicians all engaged in shared practice, with a shared background and shared belief in what is useful and what is not. The database contains the gems of understanding without the dross amongst which those gems are originally found. The value of Eureka - and it has proved immensely valuable - lies in the way it reflects not ‘information’ in general, which can be infinitely varied, but the local, specific knowledge produced by participants.
The virtual should extend and transcend the limits of the social while at the same time honoring its salient features of division and boundedness. In the light of a social theory of knowledge production, not only do organizations and their role need reappraising, but so do technologies. In designing organizations, people consider not simply individuals and information, but also communities, communication and knowledge. Technology design must face the same demands, though today it often avoids them. Once communities, communication and knowledge come into consideration, for example, the implicit claim, often made, that technologies can replace social institutions, looks extremely dubious. More often than not, such arguments are merely a sleight of hand, a quick confusion of social systems with technological systems.
It is confusion over social systems and technological systems that has allowed many firms to conclude that an intranet will meet the challenge of synthesizing knowledge. For genuine synthesis, technology will be vital. But design must engage directly with the social institutions it hopes to address. Too often it assumes these can simply be bypassed.
Here, the issue of reach and reciprocity comes back again. Social factor, we argue, demand reciprocity, though this inevitably serves to check the implicit paradigm of limitless reach implicit in a good deal of technology design. The increasing reach of technologies is unquestionably of colossal importance. It is equally important, however, that striving for reach does not disrupt the need for social reciprocity.
Reciprocity is difficult to handle. It involves not just simple tit-for-tat, but rather the means for all participants in technologically linked interactions to participate fully. It remains a common lament on the Web that most design there only simulates interaction.
Some emerging technologies suggest richer possibilities for reciprocity are under development. New forms of multicasting over the net offer denser forms of interaction (without needing the infinite bandwidth that seems to be just around the corner and for which we are always told to wait). But many of the demands here are, of course, social not technological. Reciprocity, as we have described it, is similar to what Lave and Wenger refer to as legitimate peripheral participation.14 Intriguingly, new communications technologies have gone a long way to provide peripherality. This allows us to lurk on the side of interactions in which we are not taking part and of communities of which we are not members.15 Better technologies will also support increasingly fuller participation. None the less, it remains for social groups, not technologies, to determine legitimacy.
The rewards of reciprocity, we believe, are high. But it also has demands. It limits the control that can be exerted through communication. And it also limits the scope of participation. Full reciprocity is only possible among relatively small groups. Reach, by contrast, sometimes seems infinite. Ignoring reciprocity only makes technologically mediated participation more difficult. But technology that can recognize and to some extent parse how relations within communities (where need for reciprocity is high) differ from those between, may actually help to extend reach between communities without disrupting reciprocity within. We suspect that coming to understand the challenges of the between relation should be a significant issue for new design.
One important issue here involves the local informality found within communities is distinct from a certain level of explicitness and formality demanded between. In the past, digital technology has focused heavily on the explicit, ignoring how, on the one hand, much that is implicit simply cannot be rendered explicit and, on the other, when what can be is rendered is explicit that transformation can profoundly affect the social relations in play.
For instance, in many situations, asking for explicit permission changes social dynamics quite dramatically - and receiving a direct rejection can change them even further. Consequently, people negotiate many permissions tacitly. A great deal of trust grows up around the ability to work without explicit permissions. Contrast this with what are now called ‘trusted systems’. These are technological systems that in fact eliminate the need for social trust. They simply prevent people from behaving in ways other than those explicitly negotiated ahead of time. Everything must be agreed (and paid for, usually) ex ante. For high security demands, such technologies will be increasingly important - people are glad they can trust bank machines. But if new technologies ask people to negotiate their social interrelationships like their banking relations, they will leave little room for the informal and the tacit. By contrast, technology that can respond to the implicit in human relations may both help foster trust and remove a great deal of the burden of using technology.
One goal for bringing these divided issues - formal/informal; explicit/implicit; organization/ecology; technological/social - together may be through thinking not of replacing the social, material world with a technological, virtual one, but to consider bringing the two together. In this way, the virtual should extend and transcend the limits of the social while at the same time honoring its salient features of division and boundedness. Vague though this prescription may seem, it holds for us intimations of what we mean by pursuing both reach and reciprocity in organizing knowledge.
This article first appeared in ‘Web--weaving: Intranets, Extranets and Strategic Alliances’, edited by the late Peter Lloyd and Paula Boyle. ISBN 0 7506 3866 4