Computer technology has given us easy access to all kinds of information, but too much information can sometimes be as limiting as not enough if you can’t use it to solve a problem or gain an advantage. Confusion arises over what knowledge management is, and what it involves. Some people view it as just an upmarket label for information management, and therefore something our profession should naturally embrace. Others see knowledge management as a useful term to signal the more complex work involved in organizing access to networked information resources, and thus equate it with subject gateways. Cynics dismiss knowledge management as the latest management fad: yet another effort by management consultants and IT vendors to sell their “solutions” to desperate businesspeople who ought to know better. These are all fair comment up to a point, not least because there is still quite a gap between knowledge management theory and knowledge management practice. The need to turn information into knowledge has spawned a genre of knowledge-management tools and employees whose job it is to take an overwhelming mass of data and make it tangible, accessible, and useful.

Organizations can be viewed as systems of interpretation and constructions of reality. In order to survive, they must find ways to interpret events so as to stabilize their environments and try to make them more predictable; organizations must also find ways to interpret events so as to be one with the environment—an environment that they choose. When managers “enact” the environment, as Weick (1979: 164) put it, they “construct, rearrange, single out, and demolish many ‘objective’ features of their surroundings … they unrandomize variables, insert vestiges of orderliness, and literally create their own constraints.” Through this process of sensemaking and reality construction, people in an organization give meaning to the events and actions of the organization. This occurs at many levels: the individual and their many roles, the team, the organization, the ecosystem and the community. Sensemaking is the manner by which each entity or layer goes about describing, understanding, cognizing, and enacting its world. Strategic failure can be related to sensemaking conflicts where the various individual layers do not envision the same world and thus enact very different realities while following the same communicated strategy. Tactical failure can be related to miscommunications: the wrong information, at the wrong time, to the wrong person, said in the wrong way, with the wrong emphasis, or being wrongly absent.

Coherence, the concept of “holding together” and of self-recognition of the boundary of self, acts to tie the levels of organizations together much like the role of the unified electro-magnetic-weak-strong nuclear force in physics. When the various organizational levels enact differing understandings, coherence is endangered and uncertainty introduced. Identifying and influencing such coherence regarding the organization, work teams, and projects have become critical tasks of those we label as “managers” (Hambrick, 1997; Liedtka & Rosenblum, 1996; Bacharach et al., 1996; Teece et al., 1994; Nath & Sundharshan, 1994; Allen, 1994). As Kabanoff et al. (1995: 1075) noted:

All organizations need to solve a fundamental problem—how to maintain internal cohesion while producing economic outputs … To meet the requirements for cohesion and performance, organizations adopt different patterns of structures, processes, and values.

And keeping track of those structures, processes, and values is the task of knowledge management. Unfortunately, too many managers confuse their awareness of this task with the desire for “big brother.”


Knowledge management per se is often seen by one group of consultants and academics as an oxymoron (Skyrme, 1997); at many companies there is no knowledge and it cannot be managed. By contrast, other consultants (and many knowledge management software and systems firms) insist that corporate knowledge is a collection of identifiable objects, which can be managed top-down in the form of hierarchy, directory trees, and taxonomies (e.g., Wiederhold, 1997). And, in further contrast, yet another group of academics insists that corporate knowledge is mainly tacit, existing in the mental schemas and conversations among the members of a company’s economic web (c.f. Ngwenyama & Lee, 1997). The result of this cacophony of approaches to knowledge management is often confusion and little progress (Duffy, 2000). Systems are purchased and implemented, only to be ignored by the majority of intended users (Davenport, 1995). Knowledge management is “the next fad” or “yesterday’s fad,” depending solely on the experiences of the commentator (Davenport, 1999).

Practical knowledge management requires the development of tools whose use is intuitive and easy. Being practical means recognizing that users have daily tasks to be accomplished, none of which currently include educating, improving, or correcting the knowledge management system. If more than ten minutes’ training is required to turn a user from novice to beneficiary, the system is not practical. If the system demands continual interactions from the user for categorization or learning tasks, the system is not practical. Practical knowledge management means letting users forget that the system is there. Most of us use writing instruments that do not require care and feeding (fountain pens, mechanical pencils, and quills being the exceptions). When we need to write something down, we just do. Until knowledge management is packaged in a similar manner—so that users can just do—it will remain an initiative “with much promise.” Yet, the tools exist now for the promise to be transformed into benefit.

The knowledge management market space shares its borders with data-mining and search tools. The differences are subtle but important to understand. Data mining focuses on legacy data that is more transactional and financial in nature. For example, a massive database of sales figures can reveal an overall picture, or you can drill down to specific territories, products, and customers. Search tools ferret out information you know to be present, but they don’t necessarily reveal anything inherent in the data. Knowledge management tools go a step further, by enabling you to collect and organize information, search for what you need, and share your findings with other people. The most popular knowledge management technologies are online databases, document management systems, and groupware. The typical corporate approach is a suite of tools based around groupware and/or an intranet-based web, with discussion-based applications and database management. Hypertext-linked knowledge databases are generally supported by a specialized search engine and online company thesaurus. More sophisticated knowledge management systems use intelligent search agents, case-based reasoning (notably for customer service/help desk applications), and neural networks (for data mining).

There are major functional components that technology must provide in order to support the goals of any company regarding knowledge management:

  • Gather—the system must provide a means to gather information from electronic documents in file systems, intranet pages, groupware folders and databases, external web pages, news feeds, e-mail, and relational databases.

  • Contribute—users and/or administrators should be able easily to add information into the system where appropriate.

  • Organize—information should be organized and categorized in meaningful and sufficiently intuitive ways for people to use easily.

  • Distribute/deliver—knowledge must be distributed and delivered effectively and users should be able to find information wherever it resides, even if they are unsure what they’re looking for.

  • Collaborate—users need to be able to send messages to each other, participate in workflows, route information to other users, contribute to discussions, and schedule meetings with each other.

  • Refine—users should be able to refine or analyze the contents of the knowledge base in different ways.

The very goal of making knowledge “tangible, accessible, and useful” is daunting, as witnessed by these three quotes from the knowledge management literature (Milton et al., 1999):

  • “Organizations contain such a vast amount of knowledge that mapping all of it would be a futile endeavor” (Davenport & Prusak, 1998).

  • “Tacit knowledge is vital to an organization, yet is very difficult and time-consuming to capture and codify” (Nonaka & Takeuchi, 1995).

  • “Ordinary language is the main form of communication, yet is so full of jargon, assumptions and ambiguities that people often fail to understand what others are trying to say” (Allen, 1990; Dunlop, 1998).


In knowledge-intensive businesses, where turnover is high and experts are always scarce, the need for organizational knowledge sharing and reuse is mission critical. Traditional means of knowledge transfer, such as putting the most experienced people on every new project, are impractical. Knowledge workers need to learn quickly from repeatable best practices and from the experience of others across the organization. In short, communities of practice need an infrastructure for collaborative knowledge work that supports many-on-many communication, coordination, knowledge sharing, rapid learning, and organizational memory.

Knowledge management applications attempting to deliver that infrastructure typically fall into the following broad categories:

  • Knowledge databases and repositories—storing information and documents that can be shared and reused, for example client presentations, competitor intelligence, customer data, marketing materials, meeting minutes, policy documents, price lists, product specifications, project proposals, research reports, training packs.

  • Knowledge route maps and directories—pointing to people, document collections, and datasets that can be consulted, for example yellow pages/expert locators containing CVs, competency profiles, research interests.

  • Knowledge networks and discussions—providing opportunities for face-to-face contacts and electronic interaction, for example establishing chat facilities/talk rooms, fostering learning groups, and holding best-practice sessions.

Any knowledge management system is a suite of technologies, which, if properly integrated, provide a single user interface to any number of information resources and business processes. The intent of the knowledge management system is to provide employee access to all of these disparate information sources—in essence, to act as a universal integration mechanism. Content needs to be accessible both through pull—finding a document or a person—and through push—publishing and alerts originating from other than the individual user. At the same time, the knowledge management system is challenged to provide enough flexibility at the individual worker’s desktop so that each can adapt it to reflect their individual professional (and personal) information needs. This is done through categorization, context, search, personalization, and profiling (c.f. Milton et al., 1999).

What categorization brings to the knowledge management system is to create a context for information. Humans don’t operate on isolated items of information, but in domains of understanding that are created by interrelated layers of meaning. Yahoo! designed categories with broad appeal to the broad internet audience. Inside the organization, the corporate knowledge management system has a related, but subtler design challenge: to create a context that reflects and supports the organization’s business. Within each organization, elements such as current business practices; management initiatives; corporate history, structure and culture; available professional resources; and learning requirements build up a context for working with information. For the corporate knowledge management system to succeed, the information available within it must reflect those established patterns and familiar context.

The search component provides a centralized facility for pinpoint access to specific information items throughout the collections available at the knowledge management system or accessible to it. The challenge in integrating search functionality is to confront the widespread frustration and skepticism that have developed from users’ experience with inadequate search mechanisms on the public internet. Four prerequisites of an effective search offering should guide the development of the corporate knowledge management system search facility: comprehensive indexing; metadata access; full text access; and concept-based search.

The personalization facility is a critical ingredient in productivity enhancement and effective individual information management, both professional and personal/collaborative. In the new internet e-business environment, a facility for personalization becomes a necessity, not a luxury. Personalization has become a necessity because the volume of information available in the internet-enabled electronic business environment has outstripped the capacity of the individual to process it. Additionally, knowledge workers are beginning to appreciate a new level of power and effectiveness in the ways in which they can work with the new expanses and combinations (views) of information available to them. Personalization allows individuals to tailor information feeds and displays to bring unprecedented depth of perspective, clarity of focus, and timeliness of data to each new decision. In addition to personalizing at the presentation level, corporate knowledge management system applications should provide tools enabling individuals to personalize their knowledge management system’s content by centralizing, managing, and prioritizing the delivery of information on a job-function or interest basis. This facility allows the user to eliminate whole channels of information that are not currently of interest or germane to the job function supported by their desktop.

Building on this concept is the idea of individualized profiling, which knowledge management systems should also support. Individual profiling acts very much like information filtering in reverse. Individual profiling starts with a profile of user functions and interests, and then heuristically scans the information environment for new documents or other elements that might be of interest. Armed with the individual’s job and interest profile, the knowledge management system can then suggest new items of interest it has located in the process of “crawling” the information sources available on the knowledge management system. While suggestive computing is still in its infancy, the knowledge management system environment is a place in which it can achieve a range and scope that can make this functionality a helpful new mechanism on the knowledge-supporting desktop. By combining the tracking of people and the metadata descriptions of information resources, the knowledge management system can maintain personal profiles automatically and provide intermediation to organizational expertise. The level of profiling is very granular—for each person, the knowledge management system tracks projects, documents, organizational position, the tools they use, their knowledge management system design, etc.


Knowledge management systems take two basic approaches to gathering knowledge: harvesting and contributing. Products using the knowledgeharvesting approach are information resources that an organization’s users can leverage as needed to complete their business tasks and actively search for information from widely disparate data stores. At the back end, these tools provide highly sophisticated and configurable indexing and search engines, with advanced technologies such as web crawlers, neural networks, summarization, concept searching, and hyperlinking. Many of these products also provide the flexibility to allow users to find actively what they want when they want it, and can automatically deliver specific information to specific users based on their preferences, their roles and responsibilities, or their position within the organization.

By contrast, products making use of the contributing approach are designed to allow users to add information and expertise to the system, making it available throughout the organization. Some products take this approach a step further by providing environments similar to groupware platforms, acting as the main system that employees use to work on projects and collaborate with others.

Some products can clearly be categorized into one of the two approaches to knowledge management. For example, vendors with products focused on the knowledge-harvesting approach include those from Autonomy, Excalibur, Magi, PC DOCS/Fulcrum, and Plumtree. Vendors with products focused on the user contribution and knowledge-sharing approach include those from Intraspect, Open Text, and Xerox.

In general, most companies have needs that span both approaches and, consequently, any product recommendations must adequately address those needs. The approach above fails to demand the creation of a corporate taxonomy, a company-wide directory, a common file structure, a Lotus Notes database, or a common desktop implementation. Instead, the need for users to be comfortable with a system and to believe that they are individually gaining value and benefit from its use is recognized and given prominence. There is no recommendation for training, for “change programs,” for extensive interjections of consultants, or for imposed alterations to how current employees go about doing their dayto-day tasks. Selling is not a component of this approach. Rather, it is constructed around the belief that if you give people the opportunity to access and use something of value, they will; and if they don’t, that compelling such usage is counterproductive.

Large knowledge management practices at consulting and software firms will suggest that the above is doomed to failure. There is neither command and control nor a need for large and intrusive systems. The recommended technologies can all be added to existing systems with minimal integration needs. For those whose livelihood is dependent on selling large-scale systems, which demand extensive integration with current technology and potentially extensive substitutions of current practice, the threat is economic, though the objections raised will be phrased as philosophic and academic. Few of the large practice groups describe their key audience as “users.” Rather, their target is management and systems people. And therein lies both the problem and the opportunity.


Companies of the twenty-first century need to provide a compelling place where people find, create, share, discuss, and apply the information they need and capture the value of this knowledge work for the organization. As people work in the environment, sharing documents, holding e-mail conversations, running projects, having a virtual meeting, locating needed information or contacts, they are drawing from and contributing to a collective resource that benefits the entire community. These are high-minded ideals, and they are frequently violated. Yet, if there is to be any value from the “system” part of a knowledge management system, the ideal of “collective” resources and benefits needs to be instilled throughout the organization. Knowledge management cannot work if it is to be based on selfishness. Jim March and Herbert Simon (1993: 2) once wrote that an organization is a “system of coordinated actions among individuals and groups whose preferences, information, interests and knowledge differ.” If we accept this notion, then the central task of knowledge management is to assist managers in “the delicate conversion of conflict into cooperation.” The pragmatic and practical question for knowledge management is how to become a useful tool and not an end unto itself.

In this quest to become a useful tool lies the link between complexity theory and knowledge management. One complexity-oriented view suggests that organizations and their members can be seen as recursive networks of holons—entities that are simultaneously parts and wholes when viewed from different “levels.” Each holon is viewed as a semiautopoietic entity—organizationally closed, closed to efficient causation, yet open to energy, information, and dissolution—with self-awareness as to its own status. The autopoietic goal of each entity is coherence with regard to self-referenced concepts of identity. When the holonic levels enact differing ontologies (the set of beliefs comprising the sum total of their sensemaking), coherence is endangered and uncertainty introduced. Some resolution of the conflict between the need for coherence and the existence of differing ontologies must result. Into this breach steps “knowledge management” and its advocates. Knowledge management, it is claimed, is the useful tool capable of coherently resolving the conflicts.

The key is to stop viewing knowledge as a thing—an item that can be possessed, contained, bartered, or sold—and instead to view it as a process. Knowing is evidenced by action. Activity theories in general argue that knowledge is constantly evolving. Rather than studying knowledge as something individuals or organizations supposedly have, activity theory studies knowing as something that they do and analyzes the dynamics of the systems through which knowing is accomplished. Recast in this way, knowing is: (a) manifest in systems of language, technology, collaboration, and control (i.e., it is mediated); (b) located in time and space and specific to particular contexts (i.e., it is situated); (c) constructed and constantly developing (i.e., it is provisional); and (d) purposive and object-oriented (i.e., it is pragmatic). This view leads to a working hypothesis (from Lissack and Roos, 1999): knowledge is whatever it takes to create a willingness to act. Knowledge management is then the task of creating an appropriate environment—an environment, which allows such willingness (to act) to occur.

Complexity-oriented knowledge management becomes a task of continuously attending to environments and to the way in which people interact in those environments. Knowledge management becomes the domain of the corporate anthropologist and the architect rather than the IT specialist and the “team” from the big-six accounting firm. Valueadded in knowledge management comes from attending to the little things that enhance the environment: the corporate environment, the family environment, the community environment. Rarely does valueadded come in large-scale implementations, multimillion-dollar “change” programs, “quality intitiatives,” or the creation of corporate taxonomies.

We build our identity through daily interactions. Collective action is regulated through the constraints of the environment and the structure and the “culture” of the actors. Similarly, culture and identity are built by action and interaction with others and the environment. Change in the organization cannot be undertaken under mere technical considerations, as though ceteris paribus were true. We learn much more than we can articulate from our enactment of our environment, and in the process we create tacit, incorporated knowledge that will influence our interpretation of the context. Miller (1993:119) notes: “Experience … shapes the lenslike cognitive structures through which managers see the world.” The rules for using knowledge greatly depend on the institutional setting in which the usage occurs. The setting provides the appropriate vocabulary for saying something meaningful—for understanding and establishing rules, as well as finding, creating, and using terms. In effect, the words we use and the knowledge we impart just by talking and acting are part and parcel of defining the environment. Coherent words are those that allow the mental images they conjure up to “match” the tasks with which their users are associated. Discordant words, by contrast, merely provoke unease and ambiguity. Similarly, coherent knowledge management environments are those that allow usage to occur without thought. The very act of needing to think about the use of the system is the beginning of the definition of its incoherence. The proof as they say comes in the pudding—no tool has value without use, and practical knowledge management is about tools.

This approach is anathema to the big consulting firms who are heavily invested in the systems approach and who translate that into demanding that their clients heavily invest in new “systems.” This approach does not demand that mindsets and mental schemas regarding knowledge, communications, and knowledge management be changed before the activities begin. In direct contrast, it argues that by creating an environment and a set of reinforcing activities, the attitude, mindsets, and schemas will change in due course. This may be heresy to a prominent group of academics and consultants who have much invested in the “changed minds change actions” approach—but it is much easier to cause changed activities than it is to change minds. It is much cheaper, too.