next up previous

Processes of Knowledge Preservation: Away from a Technology Dominated Approach *

Kai Romhardt (lic.oec.HSG)
University of Geneva/geneva knowledge group
Hadlaubstrasse 110, 8006 Zürich, Switzerland
Tel./Fax.: ++41-1-361 45 63

*This article is a building block of a knowledge management concept the author developed with his partners at the University of Geneva. Compare Probst/Romhardt (1997), Romhardt/Probst (1997) or Probst/Raub/Romhardt (1997).

1. Processes of Knowledge Preservation: Away from a Technology Dominated Approach

Voices of Managers

"In our R&D center we have a small group of absolute product experts. The most skilled and most respected retired some days ago. We are sure that a good part of our product expertise will be lost but we do not know how to save his experiences for the future."

(R&D manager of a food-company)

"Some months ago I realized what it means to work for a company that knows how to protect valuable knowledge. I participated in a presentation of a younger colleague and saw him flipping charts that I had produced some time ago. The charts had transformed into company knowledge and the speaker had no idea of their origin."

(Partner of a consulting company)

"In our company a multitude of project groups on different levels are trying to create an electronic memory for their area of responsibility. But an integrated solution for the whole organization is missing, which will cause interface problems in the future. I fear that if we have reached the end, there will again be only a fraction of our experience at the fingertips of our employees".

(IT manager of a big service company)

The special importance of an organizational memory has been stressed by many management thinkers in the last years, but in most management concepts the conscious handling of the organization's past plays secondary role. In general, we can describe memory as a system of knowledge and capabilities that preserves and stores perceptions, actions and experiences over time and secures the possibility of recall for the future.[1]. The organizational memory is the crucial point of reference for new experiences: without memory, learning is impossible.

Underestimation of experience

The preservation of knowledge is an important building block within the concept of knowledge management. The value of the organizational memory is today particularly underestimated in the process of reorganization. The following statements of managers seem to be typical: "We have to become leaner." "We have to build up a younger staff." These arguments often prepare the way for radical outsourcing or downsizing activities. "Ballast" is discharged, but do we talk about well thought out or negligent separation from our own (out-dated?) experience? 2

Irretrievable losses

The dismissal of employees unwilling to change may break a blockade but always at the price of losing personal know-how and expertise. Many companies made the bitter experience that through lean management and the unavoidable discharges and outsourcing activities valuable know-how left the company and had to be bought back by hiring expensive external consultants. Certain company-specific skills, e.g. the architecture of old industry plants, were lost forever. The loss of certain critical information may reduce the performance of entire company departments.[3].

The memory of the company

Let's take the case of Andy Miller. Andy has been working for the last thirty years in the sales department of a big American trading company. Everybody knew him as the `good soul' of the hundred employees in the department. Most of his time was spent in informal talks and in this way he talked daily to nearly every salesman. A change in management induced a detailed analysis of the individual sales performance for which an external consultant was hired. Andy Miller, who never had sold very much and was aged over fifty, got his dismissal at the end of the quarter. The personal comment of the consultant: "Miller could seldom be found in his place and spends most of his time in non-sales-related chats." After Andy's dismissal problems never experienced before started. The coordination of sub-unions did not function as well as before. Responsibilities that seemed to be clearly defined became unclear and the number of customer complaints started to rise. The general mood changed. There were various complaints. Anniversaries and birthdays were forgotten, new employees felt poorly supported and broke unwritten rules of the company. Day by day it became clear that with Andy's dismissal the "memory of the company" had also been dismissed. He was someone who had had built up detailed knowledge about people and processes within the organization and he had shared that with others during his `unproductive' chats.

This case underlines that without intentionally keeping experience, unexpected losses may occur. Man is a creature deeply rooted in his evolutionary history and he creates his own identity by referring to his past experiences and using his unique ability to learn. Indeed, many companies complain that they have lost a part of their corporate memory in the process of reorganization. This collective amnesia is often the result of the non-reflected destruction of informal networks which steer important but seldom observed processes. Consultants have already named this disease collective Alzheimer syndrome, a phenomena which can be found particularly in shrinking companies.

Unlearning versus protecting

Management theory discusses the tension between the destruction and protection of old skills, capabilities, and information mainly under the topic of unlearning. Hedberg defines unlearning as the process in which learners discard their old knowledge[4]. This definition requires a rigid separation from past knowledge that burdens the carrier. This separation is necessary for a new start. Organizational unlearning has to start if the actual patterns of interpretation and reaction or if the organizational theory in use no longer fit the changing challenges of the organizational environment. The problem is in the selection process between knowledge that is no longer needed and knowledge that is absolutely necessary for the future. In this logic, unlearning means to question one's own routines and to let go of things we are used to.

Experience as a starting point for improvement

Should all of our existing customer data be deleted because our marketing did not function well in the past and has to be reorganized? Absolutely not. Should successful teams be disbanded, because they worked on the wrong questions? Absolutely not. Should we dismiss all our employees who have reached a certain age because they can be categorized as too inflexible for future changes and because their early retirement will be subsidized by public institutions? Absolutely not. Purposeful protection of critical data and information is of enormous significance for every organization. The evolution of the organizational knowledge base is only possible in reference to existing knowledge. Individual psychologists think that old experiences can not be overwritten by new knowledge and therefore cannot be deleted. Instead, old rules are marked 'obsolete' and do not become operational under changed circumstances. But, they still exist as an optional path of action that widens the scope of action of the organization in a turbulent environment[5].

Processes of knowledge preservation

Organizations that would intentionally manage their experiences in order to have them on call for the future have to master three basic processes of knowledge management. They have to select out of the large number of organizational events, persons or experts and processes only those that are worth being preserved. They should be able to store their experience in a suitable form and, as the last step, ensure the actualization of the organizational memory.

Chart 1: The main processes of knowledge preservation

1.1 Selection of valuable knowledge and information

Principle: selection

Every organization gains new experience day by day that may be useful in the future and therefore should be protected. Project reports, meeting minutes, letters or presentations emerge from a variety of places. Every day customers come up with complaints and problems but also with ideas and praise. It is impossible to keep track of all these organizational events. Let's take the example of a salesman who - the same as his colleagues - often presents his products to business partners. This salesman has produced a sales presentation which visualizes the product advantages much better than present sale tools. That his colleagues do not know anything about his presentation may be a problem of insufficient knowledge identification processes within the company or it may have its roots in inadequate communication. Perhaps no one ever offered any sharing-incentives to our presentation professional. The perspective of knowledge protection asks what is going to happen if the salesman leaves the organization tomorrow? Who will find his central documents or presentations on his better or worse organized hard disk? Are the most important processes and contact persons documented? In many cases, the unexpected departure of an employee leaves a painful hole because of insufficient documentation during the time of employment. As documentation means work and money, an investment that rarely pays in the short run and rarely brings a reward to the writer, we need selection rules. It makes no sense to document everything. We cannot and should not keep everything. The challenge lies in the selection between protection-worthy and not-protection-worthy knowledge entities. We have to transfer valuable data, information and skills into organizational systems in which they can be used by the whole company. An ARTHUR ANDERSEN system is a good example:


Systematic knowledge protection and selection by ARTHUR ANDERSEN ONLINE

The internal information system of ARTHUR ANDERSEN is the backbone of this knowledge-sensitive organization. Electronic discussion groups exist in all areas that are important for their consulting know-how. The quality of contributions to these news-groups is extremely volatile and many quotations are out-dated very quickly and become worthless.

Chart 2: From a Divergent System towards a Convergent System (source: Arthur Andersen)

The challenge lies in analyzing the single information bit of this divergent system. ARTHUR ANDERSEN defined clear responsibilities for this analysis and selection process. Every organizational center of competence (e.g. total quality management) has a manager or a team that is in charge of this process. Supported by an interactive knowledge creation process, single contributions and documents are condensed into central documents. For example, twelve experience reports concerning the introduction of Lotus-Notes are summarized in one master-document, which sums up the central lessons learned. This synthesis enables the reduction of the information load and ensures that only relevant topics or reports are in the system. The user does not have to read all the single reports and saves time. Typical products of these converging processes are best practices, best firms, presentations, process definitions, studies, articles and impact-analyses. These are stored in a structured way by ARTHUR ANDERSEN ONLINE and are accessible for all members of the organization. The knowledge managers are responsible for the clean up in their center of competence. Without the deletion of irrelevant or out-dated information, the world-wide databases would grow at a rate of three percent daily and would soon become dysfunctional.

Parallels with the human brain

Organizational selection and storage activities are comparable with processes within the human brain. To make an impression on our long-term memory an information has to pass the filters of ultra-short-term and short-term memory[6]. These gatekeepers of long-term memory divide relevant from irrelevant perceptions and in this way protect the brain from permanent overstimulation. Unfortunately, the conscious part of our psyche can influence this process only in a very limited way. The consequences are various learning tricks and techniques to outwit the gatekeepers and to filter valuable bits out of the passing stream of information.

Organizational routines of knowledge protection

On the organization level comparable problems exist. Not all selection mechanisms are planned in a systematic way. Organizational routines - such as the filing of a certain document type - take care of the automatic execution and conservation of certain processes. Nearly every office has its file cemetery or dusty archives that represent the wrong way to deal with knowledge and information preservation. In these areas, routines are anchored very strongly and employees who are in charge of the concerned systems will not change their behavior without being pushed.

Knowledge documents

Therefore, organizations will never be able to manage all the processes of knowledge selection. That also would not make sense. But for core areas of the organizational knowledge base (e.g. knowledge about customers) efforts of purposeful selection and documentation should be taken. The materialization of this knowledge (the part that can be made explicit) in knowledge documents as knowledge maps or lessons learned separates experience from the individual and secures its possession by the organization for the future.[7]. Crucial for this process is the concentration of knowledge on certain core points and strictly relating it to observed problems. Only knowledge that will be usable for a third party in the future are valued to be protected and saved for the future. The rest only steals time and trust in the quality of the documentation system. Sometimes less is more. Established documentation systems should be seen through this perspective and be tested on their right to exist. Nevertheless we have to take into consideration that only a small part of our future information needs can be estimated, which demands a not-too-rigid selection approach.

Documentation of success stories

Another chance to gain knowledge about an organization's past is to illustrate leading ideas as leading principles, models, stories or other symbolic forms. These storage media are very helpful because of their potential to gain quick access to thinking patterns within the organization.[8]. For example a Swiss retailer engaged a consultant to document the process of a very successful introduction of a new product line. By interviewing all the important actors within the strategic process and the formulation of specific key factors of success, the internal success story was reconstructed. The findings of this study were summarized as a case study and can today be used for training. The case shows that a positive and successful event in a company's history can be used for motivation and knowledge transfer.

Identify key-employees

New technologies like workflow-management or document-management-systems open a new dimension for the protection (and surrender) of organizational knowledge. Despite these technological options, it's still people who make good or bad selections. Employees like Andy Miller cannot be replaced by machines or computer systems. Their experience is the key for a useful organization of the company's past. These key employees have to be identified and their continued presence in the company assured. This is the most secure preventative against collective amnesia.

10.2 Storage of knowledge

After separating the protection-worthy part of organizational knowledge from the less important entities, we have to store it in the organizational knowledge base in a suitable form. We distinguish three forms of storage: individual, collective and electronic storage of organizational knowledge. As these three storage processes follow different logic we will discuss each separately.

10.2.1 Individual storage or "Who still knows about...?"

Unfixed assets

Whether caused by lay-off and termination, retirement or death, organizations permanently lose valuable knowledge kept by these individuals. The Economist named these outstanding employees with a ironic wink, unfixed assets. This is the exact problem of a company's relationship to its knowledge workers. Knowledge that is anchored only in the head of the employees is of volatile quality. Once losses occur, they can only be replaced with high investment and incalculable side-effects. The easiest way to save intellectual capital is to create an atmosphere that does not stimulate thoughts of changing companies.

Incentive systems and barriers to exit

If top-performers like their social environment, they will be less open to lucrative offers from competitors. But, top-performers have quite different motivational structures. Knowledge about these motivational structures that may offer the right incentives can be gained by careful listening. If we think that an excellent working atmosphere added to an average income, is sufficient for the long-term commitment of an employee, we will most probably lose some of our best experts. Exit barriers[9] may be created by social or material incentive systems. In order to be effective, they have to take the personal needs of the employee into consideration.

Flexible holding-mechanisms

In many cases it will not be possible to create sufficient exit barriers. Many qualified top-performers who have their own ideas, cannot stand big companies permanently and risk the jump into independent work. The establishment of flexible cooperation with these alumni is a rewarding option to preserve the access to their know-how after the termination of their contract. Alternative forms of cooperation are operations as trainers, consultants and selective cooperation in difficult talks with old customers and more. The guiding principle is the creation of win-win-situations. Consulting firms with annual fluctuation rates near thirty percent use the permanent exodus of skilled employees for the establishment of a strong network. They have `their people' in nearly all important companies and they secure exclusive access to important information. The clever management of relations with retirees is a neglected field. One might get the impression that the 'elders' have no value at all. Their exit of the organization is often abrupt and communication stops. An investment in a relationship with retirees enables access to information and customer contacts that would, without them, be lost to the organization forever.


Storage of experience from `gray managers'

Before founding ABB CONSULTING SA, the Swiss management of ABB was facing a dilemma. On one hand, early retirement was a necessity to make room for younger promising talents. On the other hand, ABB did not want to lose their highly experienced and well connected managers. In addition, older managers wanted to influence when and how they retired in a more flexible way. The founding of ABB-CONSULTING SA facilitated this retreat in installments. This company employs today only former top-managers of ABB Switzerland that want to start a new career as a consultant at the age of sixty. These gray consultants act mainly in daughter companies of ABB and profit from their world-wide network as well as their industrial experience.

The consultants of ABB CONSULTING work in various fields of consultancy such as short-term management, public relations or cooperative projects with state or other official parties. The knowledge of the old foxes can be applied to many areas. Their experience provides ghost-writing for top-management, temporary leadership of a production plant they know from their active corporate life, or as an experienced process coach in complex projects. While regular and sometimes forced retirement normally cuts the knowledge transfer between employer and employee, ABB found a flexible solution that works for both sides. The company has extended access to the know-how of their former employees while the gray consultants get personal and financial recognition of their expertise. By the way, the fees for expensive external outplacement consultants can be saved.

Systematic transfer of capabilities

Another option to preserve critical capabilities is the continuous training of the 'successor'. The successor should be introduced step by step into his future tasks to guarantee knowledge transfer in critical areas. In many European companies current position holders try to defend their knowledge monopoly till the last day in order to strengthen their power position. The Japanese have different management principles.


One principle is called sempai-kohai and expresses the close relationship between a male pair that is built of an older, teaching sempai and a younger kohai who has to be instructed.[10]. Every `rookie' is assigned to an older mentor (who is sometimes only a few years older) and who transfers the necessary tricks to the younger generation. The relationship between sempai and kohai is strengthened by shared leisure activities in a systematic way so that trust can be built for the free-flow of information of all kinds. The transfer of implicit knowledge is also supported by this personal relationship. In 1993 when the Japanese steel industry discharged 25 percent of its 150,000 working force, many observers feared a radical decline of the average qualification. But the decline in quality did not happen, as sempai-kohai ensured the quick transfer of knowledge within the shrunken work force.

Purposeful explicitation

However, the damage caused by the loss of experts can be reduced with simple tools. One of the easiest ways is the implementation of structured exit-interviews that must be conducted by trained experts. In these interviews, the critical knowledge of the person (special documents, contacts, project engagements) is made explicit and documented. These talks should be coordinated with the successor. Led in an open, positive atmosphere, these interviews may capture much valuable information and learning references about the organization. One gets to know why the expert decided to leave the company and may get ideas for the adaptation of actual exit barriers. SANDIA NATIONAL LABORATORIES in Albuquerque, New Mexico, documents these interviews on video.[11]. This is their method of capturing the wisdom and know-how of resigning scientists and of keeping their expertise in the memory of their organization.

10.2.2 Storage in the collective memory

The human memory is transient and not static. Psychologists claim that we change or rewrite our own past with every act of remembering it. The problem with remembering is that `wrong' memories feel exactly like `right' memories.[12]. In order to avoid a complete confusion between self-constructed reality[13] and his `real' past, man needs feedback of third parties who confirm his own picture or force him to adapt it to their perceived version. That's how groups or other collective entities become the regulators of their own experience. Collectives store shared experiences differently than individuals. For example, psychologists can identify the borders between differently involved regions of the 30-years-war by conducting "depth-interviews". The horror of these times is burnt deep into the collective memory and after three hundred years still influences the daily behavior of the descendants.[14]. What is the link to knowledge management and knowledge protection? It is the proof that historical experience is deeply rooted in organizations, although it is often missed by a quick observer.

collective wins over individuals

Collective storage can be seen as historical ballast, but it may also be very productive. This can be illustrated by a laboratory experiment in which single persons and a closed group were trained in the construction of transistor radios.[15]. One week after the training, the people were reunited and asked to remember and execute the construction steps they learned. Individually trained people were put into small groups while the `group-trained' people stayed together and tried to remember collectively. The results were analyzed and showed that the group-trained could recall more details and constructed better radios. Detailed video analysis showed that during the initial training a series of social and cognitive relations were established which the researchers called transactive memory system[16]. This collective memory was superior to that of individuals.

Extension of the memory

Another phenomena of collective storage was gained through the observation of pairs [17]. People use certain other people with whom they interact intensely as extended memories. This is a way to increase their own memory capacity. They develop a feeling for which details of a shared social situation will be remembered very good (e.g. names). This division of memory has the funny effect that two partners can only remember a social situation exactly if they are together.

Documentation of important processes

Many team success stories cannot be explained by these observations. Group processes or collective problem solving processes have their own self-dynamic and are difficult to understand for external observers, or even the members of the group. Despite these difficulties, much can be done to capture important processes for the future and to gain starting-points for improvement. Employees of the Japanese-American joint-venture FUJI-XEROX documented every important step and detail of the cooperation process.[18]. Through this, future employees of the company should have the opportunity to learn from and about the past.

Taking minutes

The most traditional form of documenting a meeting are the minutes. But good minutes and the people who can produce them are rare, and, this activity has often been seen as a tiresome duty. The result of this low regard are too short or too long, redundant, badly-written, poorly structured, full of gaps, late-arriving annoyances in paper form. But for organizations in which most employees work in fast-changing project teams, minutes become a central source of collective storage of the actual project status. Japanese companies train their moderators specifically in documentation techniques. They want to guarantee that experience and decisions do not fade away or are lost. By studying the existing minutes, new group members can learn about the actual status of group discussions.

Importance of shared language

More powerful than writing is the spoken word. It offers the biggest option to build and anchor collective experience. The spoken word is closer to us than the written word. Over time, companies create their own vocabulary, which newcomers have to learn. This vocabulary is much more than a collection of the usual efficiency-creating abbreviations. Normal words like quality, change or safety are defined in a firm-specific way and reflect the company's past. These terms are often linked to strong emotions. An industrial company had suffered for a long time under a highly-paid consultant team that had conducted a failed reorganization project called "horizons". Years later, the simple mention of the word "horizons" easily killed any new attempt to bring consultants into the organization. Every organization knows terms like this, terms that are permanently embedded in daily actions. Knowledge management has to try to anchor its central ideas in the organizational language and use this power for its purpose.

Collective formation of concepts

One possibility to anchor and store central ideas is the process of collective formation of concepts[19]. By conscious questioning of central terms at the beginning of a group process `clear terms' can be picked out as a central theme. The collective deconstruction and redefinition of terms reduces the danger of future misunderstandings. Especially buzzwords, like total quality management or process reengineering, often stay misunderstood. A clear vocabulary avoids confusion.

Shared experience

The integration of ideas is a complicated issue in decentralized or heterogeneous companies. Investments in shared language or shared experiences can become enormous. At SWISS TELECOM, 20,000 employees participated in a so called "mind change workshop" as part of the reorganizing process of the whole company. In this workshop, the participants can experience `change' in small groups and may relate this change experience to their personal situation. All employees see the same video clip about insufficient customer orientation and discuss this problem in their working groups. Shared experience emerges that may be transferred to the working place and change daily behavior. The "Change TELECOM" project that is shaping all organizational structures and strategies can be `felt' and anchored in the organizational memory.

10.2.3 Storage in electronic memory

Unlimited storage capacities

IT-revolutions have multiplied electronic storage capabilities. If this development continues we will soon have virtually unlimited electronic storage capacities that cost very little.


Nearly all traditional storage media can be digitalized. Video cassettes can be replaced by CD-ROM's, documents can be scanned and digital cameras are already in the market. The average future computer user will soon have access to all forms of storage media under one surface. The qualitative difference of digitalized storage media comes from their easy editing and reuse, as well as their cheap and fast distribution over networks. With rising digitalization the base of the global digital memory of mankind is growing. The Internet gives more and more users access to these data masses. The Internet and intranets of today are just the beginning of a development that is unpredictable even for experts. If libraries, magazines, sound, movie, and text archives grow together and widely accepted standards for the organization and structuration of the digital raw materials emerge, the Internet will become a meta-archive for everything.

Consequences for the organizational memory

This development has dramatic consequences for companies which act in knowledge sensitive environments. First, they have to anticipate that their competitors have principal access to the worldwide datapool and use it for its purposes. Second, they have to look at their own electronic knowledge base. In knowledge-intensive companies, an important part of their know-how is stored in digitalized documents like presentations, forms, blue-prints or reports. Their systematic filing and re-use becomes a growing competition factor.

Memory loss

There are many reasons why access to the electronic memory may shatter. If for example the knowledge documents of a personal computer were not fed into the system, they are not available for the electronic memory of the company and its employees. If the documents were codified incorrectly or filed in the wrong place, they may never be found again and are lost (perhaps for ever). If a codification can not be interpreted or if a network or single computer is not connected to central databases, the organization will not remember that this document exists.

Only a few companies organize their electronic memory consistently and are able to master these difficulties. Most organizations fight against their own information systems and data structures that hinder the establishment of efficient and user-friendly corporate memories.

Controlled vocabulary

Another problem is the introduction of a controlled vocabulary that codifies documents with a well-defined set of key-words. This corporate language may be built by a selection and definition of central terms within the company and permits a later classification of documents to central action or knowledge fields of the organization. The disadvantage of this solution lies in the high investment for the establishment and maintenance of the vocabulary.


The intelligent connection of documents is the key factor for success in the establishment of a powerful electronic memory. Similar to the neuronal structure of the brain that captures memories in connection structures, a good IT system may lead the way to experience that has nearly disappeared. The chaotic connection structure of Internet and intranets is both an advantage and a problem for the systematic storage of organizational experiences. More and more hypertext-links lead to nowhere because their reference side does not exist anymore or has been changed.

10.3 Actualize and remember

Importance of actualization

The conscious and structured storage of organizational knowledge is not the end of the knowledge preservation process. The organizational memory is operational only if desired information can be remembered in acceptable quality. Beside sufficient selection and data-sensitive storage, we have to pay attention to actualization processes. Companies may lose money if they base investment decisions on out-dated or incorrect knowledge. At ARTHUR ANDERSEN the leaders of competence centers are responsible for the up-date of their data-bases, documents and discussion groups. So the uncontrolled growth of databases can be avoided and users do not have to suffer under old or irrelevant information.

Chart 3: The Death Spiral of an Electronic Knowledge Base[20]

Death spiral

If actualization processes cannot be controlled, knowledge systems may easily get into a death spiral. In managing their memory, companies have to solve mainly trust and access problems. If trust in data quality exists and easy access to the system can be guaranteed, systems will be fed and will be used in ways that increase data quality. If the actual database is already full of mistakes, trust cannot be built and no-one will invest much energy into the system. Data quality decreases further and the system dies. In times of short half-life death may arrive very quickly.

Organizational amnesia

The biggest problem for the organization memory is corporate amnesia. Typical statements of managers are: "We once knew about it, but now we seem to have forgotten it." There are many ways that organizations lose their memories. Employees walk away, excellent teams split, data-bases become infected by a virus or whole functions are outsourced. These actions reduce collective memory. Sometimes access to the memory is simply blocked, sometimes forever. Examples are the permanent or limited overload of experts or the low motivation to share experiences with others. Equivalent memory barriers exist on the electronic and collective level.

Chart 4: Types of Organizational `Unlearning'


This table illustrates that the protection of experience and know-how can be described as a permanent battle against the natural process of forgetting things. Forgetting is a well-known phenomena. A foreign language we once knew but do not speak is lost word by word. Untrained muscles shrink. Learning psychology speaks about conserving repetition, well illustrated by the pain of learning vocabulary. Many trainings fail because fresh knowledge cannot be translated and applied in daily activities. Much knowledge fades away over time and when we need it may be already gone. This is the reason for action training in many companies. They train their employees in their typical working situations and hope in this way to conserve the new know-how for a longer time.

Keeping the production line warm

Another case shows that sometimes high investment is necessary in order to avoid collective amnesia. Over many years, the US government ordered overpriced submarines without any actual demand in order to avoid the loss of skilled workers in this specialized field of construction. These workers were the only ones with the special welding skills and without their knowledge, the US defense department would have faced serious problems in the future. To secure the availability of this know-how for the next twenty years and to establish a pool of future trainers, the US-government decided to keep the production line warm and kept on building submarines.

[1] Following Oberschulte (1996). Oberschulte constructs connections between organizational learning and organizational memory.

2 Compare Davenport (1996: 35).

[3] So-called redimensioning activities that should reduce costs have led to the loss of capabilities in many companies. Compare Mitroff (1995: 27).

[4] Compare Hedberg (1981: 18).

[5] Similar Cohen/Levinthal (1990).

[6] For different forms of memory, compare Vester (1978: 43ff.). For the importance of the actual memory, compare Wessells ( 1994: 107ff.).

[7] The necessity of the materialization of knowledge in knowledge documents is stressed by Schüppel (1996: 256f).

[8] Compare Probst/Büchel (1994: 21).

[9] Compare Bonoma/Slevin, (1978: 205).

[10] Compare Economist (20.04.96: 58).

[11] Compare Economist (20.04.96: 58).

[12] In this process, the brain instinctively mixes experienced and 'heard' stories. This phenomena is named cryptomnesia by psychologists. Compare Kotre (1996).

[13] The different approaches to how reality is constructed is not part of this paper. Interested readers can find ideas in the schools of radical constructivism, knowledge sociology and different psychiatric schools. Compare Watzlawick (1986, 1988); Berger/Luckmann (1994) and Sacks (1995).

[14] These observations are compatible with the Freudian idea of a collective Super-Ego. For modern forms of collective memory research compare Hejl (1991).

[15] Compare Liang/ Moreland/Argote (1992).

[16] Compare Cohen/Bacdayan (1994).

[17] Compare Wegner (1996:189ff).

[18] Compare Economist (20.04.96: 58).

[19] Schüppel (1996) classifies this method as a process to manage implicit knowledge potentials. Compare Schüppel (1996: 264f).

[20] Following Manago/Auriol (1996: 28).

Berger, P.L./Luckmann, T. (1994): Die soziale Konstruktion der Wirklichkeit: Eine Theorie der Wissenssoziologie, Frankfurt (Main): Fischer.

Bonoma, T.V./Slevin, D.P. (1978): Management and Type II Error, in: Business Horizons, 4, S. 61ff.

Cohen, M.D./Bacdayan, P. (1994): Organizational Routines Are Stored as Procedural Memory: Evidence from a Laboratory Study, in: Organization Science, 5: 4, S. 554-568.

Cohen, W.M./Levinthal, D.A. (1990): Absorptive Capacity: A New Perspective on Learning and Innovation, in: Administrative Science Quarterly, 35: 1, S. 128-152.

Davenport, T.H. (1996): Some Principles of Knowledge Management, in Strategy&Business, 1: 2, S. 34-40.

Economist (20.04.96): Fire and forget?, 20.04.1996, S. 57-58.

Hedberg, B. (1981): How Organizations Learn and Unlearn, in Nystrom, P.C./Starbuck, W. (Hrsg.): Handbook of Organizational Design, S. 3-27, New York: Oxford University Press.

Hejl, Peter M. (1991): Wie Gesellschaften Erfahrungen machen oder: Was Gesellschaftstheorie zum Verständnis des Gedächtnisproblems beitragen kann, in: Schmidt, S.J. (Hrsg.): Gedächtnis - Probleme und Perspektiven der interdisziplinären Gedächtnisforschung, 2. Auflage, S. 293-336, Frankfurt (Main): Suhrkamp.

Kotre, J. (1996): Weiße Handschuhe. Wie das Gedächtnis Lebensgeschichten schreibt, München: Carl Hanser.

Liang, D.W./Moreland, R./Argote, L. (1992): Group versus Individual Training and Group Performance: The Mediating Role of Transactive Memory. paper presented at the 1992 meetings of the Operational Research Society of America/The Institute of Management Science in San Francisco and at Carnegy Mellon University.

Manago, M./Auriol, E. (1996): Mining for Or, in: OR/MS Today, Februar: S. 28-32.

Mitroff, I.I. (1995): Warum schaffen wir es nicht? in: io Management Zeitschrift, 3, S. 27-31.

Oberschulte, H. (1996): Organisatorische Intelligenz - ein Vorschlag zur Konzeptdifferenzierung, in: Schreyögg, G./Conrad, P. (Hrsg.): Managementforschung 6: Wissensmanagement, S. 41-81, de Gruyter.

Probst, G.J.B./Büchel, B.S.T. (1994): Organisationales Lernen: Wettbewerbsvorteil der Zukunft, Wiesbaden: Gabler.

Probst, G./Raub, S./Romhardt, K. (1997): Wissen managen: Wie Unternehmen ihre wertvollste Ressource optimal nutzen, Frankfurt (Main)/Wiesbaden: FAZ/Gabler.

Probst, G.J.B./Romhardt, K. (1997): Bausteine des Wissensmanagements - ein praxisorientierter Ansatz, in: Dr.Wieselhuber & Partner Unternehmensberatung, (Hrsg.): Lernende Organisation, S. 129-143, Wiesbaden: Gabler.

Romhardt, K./Probst, G. (1997): Building blocks of knowledge management - A practical approach. Input-Paper for the seminar: "Knowledge Management and the European Union-Towards a European Knowledge Union", 12-14 May 1997 at Utrecht, Kennis Centrum/Dutch Ministry for Education.

Sacks, O. (1995): An Anthropologist on Mars, London: Picador.

Schüppel, J. (1996): Wissensmanagement-Organisatorisches Lernen im Spannungsfeld von Wissens- und Lernbarrieren. Dissertation, Hochschule St.Gallen.

Watzlawick, P. (1986): Wie wirklich ist die Wirklichkeit? 14. Auflage, München. Zürich: Piper.

Watzlawick, P. (1988): Die erfundende Wirklichkeit - Wie wissen wir, was wir zu wissen glauben? Beiträge zum Konstruktivismus, München/Zürich: Piper.

Wegner, D.M. (1996): Transactive Memory, in Mullen, B./Goethals, G.R. (Hrsg.): Theories of Group Behavior, S. 185-208, New York:

Wessells, M.G. (1994): Kognitive Psychologie, München/Basel: UTB.