An investigation of emerging knowledge
distribution means and their characterization
Licentiate Thesis
October, 1999
Niclas Eberhagen
Department of Computer and Systems Sciences
Stockholm University
Electrum 230, S-164 40 Kista
Sweden
Abstract
This work investigates emerging knowledge distribution means through a descriptive study. Despite
the amount of attention that processes and structures for knowledge management has received within
research during the last decade, little attention has been directed towards the actual means used for the
distribution of knowledge by individuals. In this respect it is the aim of the study to contribute with
knowledge regarding knowledge distribution means. The study consists of a survey of emerging
electronically mediated distribution means followed with a characterization and analysis.
For the characterization and analysis a framework for interpretation of the different distribution means
was created based on the constructs of organizational learning and the levels of knowledge system
interpretation. Within the framework characteristics and concepts were identified and then used for the
analysis of the knowledge distribution means. The characterization of the different knowledge
distribution means as such may be used as an instrument for evaluation since it generalizable to other
means of knowledge distribution.
The results of the study show that knowledge distribution is not an isolated event. It takes place in
larger context, such as organizational learning, since it touches upon other activities or phenomena
such as knowledge acquisition, knowledge interpretation, and organizational memory. The concept of
genre of knowledge distribution was found to be a viable concept to base exploration and
development of support for knowledge distribution. The investigated distribution means only partly
support a model for knowledge representation that captures both the problem-solution as well as an
understanding of their relationship. In this respect existing distribution means must be enhanced or
new ones developed if we wish to endorse such a representational model.
Acknowledgements
I wish to give thanks to my supervisor Bengt G. Lundberg for his comments and support during the
work with this thesis. I also wish to give thanks to Gösta Sundberg for his comments on an earlier
draft of this thesis.
To my wife Sofia and daughter Melina I wish to send my warmest thoughts. Without their
encouragement and faith in me this work would not have been possible.
Table of contents
1 Introduction .........................................................................................................1
1.1 Background on the management of knowledge ...............................................1
1.2 Research aims ...................................................................................................4
1.3 Definitions and limits of the thesis ...................................................................4
1.4 Research method ..............................................................................................5
1.5 Overview of content .........................................................................................8
2 Theoretical background on knowledge ...............................................................9
2.1 Knowledge........................................................................................................9
2.2 Knowledge and information .............................................................................10
2.3 Knowledge creation and acquisition ................................................................11
2.4 Managing knowledge .......................................................................................11
2.5 Organizational memory ....................................................................................12
2.6 Organizational learning ....................................................................................13
3 Towards a framework for characterization..........................................................15
3.1 Constructs of organizational learning ...............................................................15
3.2 Interpretive levels .............................................................................................17
3.3 The framework .................................................................................................18
4 Theoretical background for characterization .......................................................20
4.1 Media choice and use .......................................................................................20
4.1.1 Information richness theory...........................................................................21
4.1.2 Social influence model ..................................................................................22
4.1.3 Social presence ..............................................................................................23
4.1.4 Symbolic meaning .........................................................................................23
4.1.5 Critical mass theory .......................................................................................23
4.1.6 Interactivity ...................................................................................................24
4.1.7 Bindingness ...................................................................................................24
4.1.8 Adaptive structuration theory ........................................................................24
4.1.9 Dimensions of communication media ...........................................................25
4.2 Theories relating to structure and context for knowledge distribution .............26
4.2.1 Genre of knowledge distribution ...................................................................26
4.2.2 Communication contexts ...............................................................................27
4.2.3 Feedback in the communication process .......................................................29
4.2.4 Interactivity in the communication process ...................................................29
4.3 Theories on knowledge content and knowledge processes ..............................30
4.3.1 Knowledge creation processes ......................................................................30
4.3.2 Adopting and contributing knowledge ..........................................................32
5 Identified characteristics within the framework ..................................................35
5.1 Concepts and theories within the framework ...................................................35
5.2 Characteristics of knowledge distribution ........................................................37
5.2.1 Characteristics at the technical level .............................................................38
5.2.2 Characteristics at the representational level ..................................................39
5.2.3 Characteristics at the knowledge level ..........................................................40
6 Identified knowledge distribution means ............................................................42
6.1 FAQ ..................................................................................................................42
6.2 Success or failure stories ..................................................................................43
6.3 Step by step ......................................................................................................44
6.4 Reviews ............................................................................................................44
6.5 Subscriptions ....................................................................................................45
6.6 Bulletin boards .................................................................................................45
6.7 Discussion groups.............................................................................................45
6.8 Hypertext documents and traditional manuals .................................................46
7 Characterization and analysis of the distribution means .....................................47
7.1 Characterization at the technical level ..............................................................47
7.2 Characterization at the representational level...................................................48
7.3 Characterization at the knowledge level...........................................................50
7.4 A model for knowledge representation ............................................................52
8 Conclusions, future research and final observations ...........................................54
8.1 Conclusions ......................................................................................................54
8.2 Future research .................................................................................................55
8.3 Final observations.............................................................................................56
References ..............................................................................................................61
1 Introduction
Organizations of today has come to realize and recognize the strategic necessity of treating
knowledge, not as a commodity but as a resource needed to be managed and given support. It is not
that knowledge, as a phenomenon, hasn't been important before. Knowledge and skills of the workers
as a resource has always been important to the organization even if it has not received the same
conscious attention as information.
Hansen, Nohria, and Tierney (1999) state that it wasn't until the 1990s that chief executives started
talking about knowledge management. As the foundation of industrialized economies has shifted from
natural resources to intellectual assets managers were compelled to examine the knowledge underlying
their businesses and how that knowledge is used. At the same time, the rise of networked computers
has made it possible to codify, store, and share knowledge more easily and cheaply than before.
Ever since the widespread usage of Internet as a means for communication was established, computers
have come to be seen more as a tool for distributing, seeking and sharing, information and knowledge
than just as a computing machine. The usage of electronic media, such as computers, for distributing
knowledge has during the last decade been almost explosive, e.g. the growth in usage of the Internet
for communication, thus giving rise to and enabling new knowledge distribution means. Electronic
media have helped overcome time and space constraints when sharing knowledge between individuals
separated by physical boundaries, and they have made it easier to record, store, search, and retrieve it.
The background to the increasing focus in organizations on knowledge and its management may be
found in the work, going back as far as Drucker (1988) and Huber (1984, 1990), relating to the growth
and development of the post-industrial society and its informational demands as well within the
literature concerning organizational memory. Attention towards the need for methods and theories on
knowledge management has been called for at different conferences (ICIS 1998) as well as within
organization science literature (Wiig 1996 and Angus, Patel, and Harty 1998). Growing focus on
knowledge has become apparent within the literature on organizational learning and organizational
memory, e.g. Walsh and Ungson (1991) calling for methods on managing it.
The organizational processes and structures for managing and distributing knowledge have in the last
decade been given a lot of attention. Walsh and Ungson (1991) as well as Ackerman and Halverson
(1998) have explored the structures and content of organizational memories. Huber (1991) has
explored the management of knowledge through the constructs of organizational learning. Stein and
Zwass (1995) as well as Wijnhoven (1998) have explored the processes and management of
organizational memory. Goodman and Darr (1998) have explored the processes behind knowledge
exchange. Despite the above research not much attention has been directed towards the actual means
individuals use for knowledge exchange or distribution in promoting organizational learning. The
technological aspects of the means for knowledge exchange, such as the media, have received a fair
amount of attention, e.g. the review of technologies for knowledge management of Davenport and
Prusak (1998). Swan, Scarbrough, and Preston (1999) point out the same and emphasize that the real
issues to focus on are the people management issues and not just the technological ones. It is the aim
of this work to contribute with knowledge regarding the means used by individuals for distributing
knowledge through a descriptive study.
1.1 Background on the management of knowledge
In the post-industrial society organizations are expected to face a different world compared to that of
today. Huber (1984) characterize the post-industrial society as experiencing greater levels of
knowledge, complexity, and turbulence, and that each of these will be increasing at a considerable
rate. The amount of available information will grow and its absolute growth will increase. The
increase of knowledge will lead to large increases in both specialization and diversity of technology,
economy, and society. The high degree of diversity and specialization will lead to large increase in
social interdependencies and thus escalate the level of complexity and its absolute growth. The
increase in turbulence follows from the rapidity of events and increasing knowledge, causing many
1
technologies to be more effective and shorten the duration of the events, thus permitting more events
per time unit.
Huber (1984) points out that communication technologies, such as e-mail system, voice mail systems,
radio-phones, and computing technologies (mainly used for storing, retrieving, and processing
information to derive new information) will increase in availability to the individuals and increase the
efficiency of communication, making it timeliness. The cost of modern electronic media for
distributing knowledge has become increasingly lower and the media has become exponentially more
available (both in terms of accessibility and usability). The time for distributing knowledge has
become shorter, making it more or less instant. Knowledge, due to the electronic media has become
more easily copied and searchable since indexing capabilities and search tools has become more
efficient and easier to use. That the distribution of knowledge has overcome, due to the properties of
the electronic media, both geographic boundaries as well as barriers of time is undeniable when one
considers the Internet. Modern communication technologies will make available sources of
information, which were before external or unknown to the organization, and keep the information
more up-to-date. Decision making in post-industrial organizations will be more frequent, faster, and
complex thereby making the decision-task loads greater than before. This will put high demands on
the acquisition and distribution of the available information.
According to Huber (1984) organizations will need to, more than before, scan the environment for
information about the existence of problems and opportunities or for information to be used in the
future, and probe the environment for information that is not routinely gathered. Organizations must
also guard themselves against information overload, as the amount of available information will
increase to the individual decision-maker. The increasing global competition within the post-industrial
society between organization calls for rapid product and process innovation. Decision-making,
innovation, and information acquisition and distribution are processes, essential to the knowledge
work and knowledge management, which will gain increased importance within the post-industrial
organization when meeting the demands of post-industrial society and environment. The focus must be
placed on design features that organizations can employ in order to make them more effective. Such
design features may be technological support and procedures within the work practice, as defined by
Alter (1992).
Huber (1984) stresses the importance to focus on the management of the decision-processes and on the
designs for information and knowledge acquisition and distribution since information and knowledge
will be required for the organization to survive.
Drucker (1988) points out that the organizational structure is changing, decreasing in numbers of
organizational levels and becoming more decentralized. This leads to an increase in the information
load and the decision-task load of management, demanding better managing of knowledge work
processes.
Further theoretical background concerning the above-mentioned impacts of information technology on
knowledge and information work can be found with Huber (1990) where he extends his ideas with
respect to the use of advanced information technology. In a number of propositions he proposes that
the information network in organizations will become flatter as computer-supported technologies will
lead to more direct information distribution. Huber (1990) further states that if electronic media is
available it will be used, which will spur the distribution of knowledge. The effect of this is that new
ideas and innovations may be communicated and spread more rapidly through the organization and
enable new ways of seeking and sharing knowledge.
Drucker (1994) states that the knowledge society will inevitably be far more competitive than any
society we have yet known for the simple reason that with knowledge being universally accessible,
there will be no excuses for non-performance. There will be no poor countries with respect to
knowledge. There will only be ignorance. The same will be true for companies, industries, and
organizations of all kinds.
2
Leonard-Barton (1995) points out that organizations that are successful innovators are those that build
and manage knowledge effectively through such activities as developing shared problem-solving
skills, experimenting, integrating knowledge across functional boundaries, and importing expertise
from external sources.
The need for methods and theories for knowledge management has been answered by such authors as
Walsh and Ungson (1991) who have pointed on the need for developing and using organizational
memory means. Their concern lies with the contents of the memory, consisting of decisional stimulus
and the organization's responses, and the management of the memory. They have identified different
repositories wherein the organizational memory resides, such as individuals, culture, transformations,
ecology, external archives. They note that it is only individuals within an organization that can hold
the information about both the decision stimulus and the organization's response, either individually or
inter-subjectively, as in the case of a common culture. Only individuals have the cognitive capability,
by themselves or as a part of a social collectivity, to fully understand why a decision stimulus has
arisen and given a specific response from the organization in the context of an organization's history.
This understanding comes from an analytical assessment of the relationships between the cause and
effect (a decision stimulus and organizational response). The understanding fades with the passing of
time and the passing between individuals. Therefore it becomes necessary to develop and use memory
means that preserve, not only decision stimulus and the organization's response (the problem-solution
pair) but also why it has occurred so that individuals in future situations may draw on these in guiding
good behavior.
Walsh and Ungson (1991) also point out the need for organizational memory means that preserves the
knowledge and history of stimulus and response by using the analogy of how clans function. Members
of a bureaucracy or other similar organization that are engaged in transaction processes often yield
high costs because of the need for equity between the parties involved. Members of a clan are more
socially tied to each other and share a collective memory. Here there is no need for equity in a
transaction process between parties because the individuals believe that they will be treated fairly in
the long run. Organizational memory means that preserve the knowledge and history of previous
transactions may support members in an organization in reducing transactional costs by keeping
records of what has transpired before and why.
Wijnhoven (1998), drawing on Walsh and Ungson (1991), has tried to bridge the gap between
theoretical conceptualizations of organizational memory and information systems by proposing
methods for analysis of organizational memory contents, i.e. knowledge and information, and the
analysis of organizational memory means, i.e. procedures and media. The methods aim for a better
understanding of the design of organizational memory.
Stein and Zwass (1995) argue that the preservation of organizational memory has become increasingly
more important to organizations as it is recognized that experiential knowledge is a key to
competitiveness. They stress that since the development and widespread availability of advanced
information technologies, information systems has become a vital part of the organizational memory.
They have developed a model for an organizational memory information system, called OMIS, which
is rooted in the constructs of organizational effectiveness. The framework offers four subsystems,
resting on the foundations of mnemonic functions such as acquisition, retention, maintenance, search,
and retrieval of information, which support activities leading to organizational effectiveness.
Goodman and Darr (1998) have studied the processes for distribution of knowledge and examined the
role of computer aided systems (CAS) for enhancing organizational learning in distributed
environments. Goodman and Darr (1998) have studied how features of CAS enhance organizational
learning in matching problems and solutions and how organizational context, defined in terms of
characteristics of the problem and solutions such as complexity, can influence the role of CAS in
organizational learning. Their theoretical framework has focused on the decision to contribute and
adopt knowledge in distributed environments, taking a basis in organizational learning. Organizational
3
learning occurs when one unit acquires knowledge from another unit in the same organization or
between different organizations.
Goodman and Darr (1998) have specifically studied the intersection between the features of CAS and
inhibitors to contributing or adopting knowledge in the light of different organizational context
variables. They have examined two cases of information environments for knowledge sharing. The
study focused on how the intersection between CAS features and the decision to adopt and contribute
enhance or inhibit knowledge sharing. A computer aided system for organizational learning is not only
a system capable of bridging space and time such as e-mail, but also encompassing a memory device
with indexing systems and search aids accessible and known to all members. The system should also
have a mechanism by which all organizational members can dynamically share and update solutions.
Goodman and Darr (1998) concluded that the demands to build effective organizational learning
processes in distributed environments are likely to accelerate rather than decline. They therefore
advice both practitioners and researcher to concentrate on the core decisions of adopting and
contributing and the costs inherent in these decisions, design future CAS as to minimize these inherent
costs, consider functional equivalent mechanisms independent of the CAS for reducing these cost, and
recognize the importance of characterizing the problem-solution environment when designing any
CAS. The work of Goodman and Darr (1998) will be further explored in a later chapter.
1.2 Research aims
As was stated earlier the organizational processes and structures for managing and distributing
knowledge have in the last decade been given a lot of attention but little attention has been given
towards the actual means used for the knowledge exchange or distribution. There are examples of
systems that have been developed to support organizational memory and knowledge exchange.
Ackerman and Malone (1990) have developed an organizational memory system called "answer
garden" that allows organizations to develop databases of commonly asked questions that grow
organically as new questions arises and are answered. Another work of interest here is that of Sneiders
(1999), which is very similar to the "answer garden" system. He has developed an evolving FAQ
(frequently asked question) answering system that provides pre-stored answers to users' questions
asked in ordinary language. Both of these systems address the mechanisms employed for distributing
knowledge in the structured form of question-answer. However, there is still a lack of knowledge
concerning what means individuals use for knowledge distribution and what characterize them as
opposed to traditional ways of distributing knowledge. It is the aim of this research to make a survey
of emerging knowledge distribution means and look on their characteristics through a descriptive
study.
Walsh and Ungson (1991) have, as discussed before, explored the content of organizational memory.
They have pointed on the need for not only to store and make available knowledge about the decision
stimulus and the response of the organization, but also the understanding of why the stimulus has
arisen in the context of the organization's history. This implies a model for knowledge representation
that captures both the cause (problem) and the effect (solution), as well as the understanding of the
relationship between them. Effective means for knowledge distribution should therefore support this
model for representing the knowledge. In this respect it is also interesting to analyze the emerging
knowledge distribution means according to how they support this model.
Through our study we will do the following:
• make a survey of emerging knowledge distribution means; and
• analyze the characteristics of these knowledge distribution means.
1.3 Definitions and limits of the thesis
Before going any further there is a need for making clear what we mean by different terms used
throughout this study. We use the term distribution to define a process of either sharing something,
adopting something, or both as in exchanging something, in our case knowledge. The term emerging is
4
indeed troublesome and rather vague. It means that a new "thing" has come forth, been born, or had its
existence enabled. It also means that an existing "thing" is experiencing a rise in its usage or
distribution. Wijnhoven (1998) defines organizational memory means as the procedures and media
used for effectively interacting with the organizational memory. Following this view we define a
knowledge distribution means as being concerned with:
the media for the distribution of the knowledge; the structure, form or language that the
knowledge is represented in; and processes or procedures for its management and creation.
The aim of the study is to make a survey of and to characterize different emerging knowledge
distribution means. This is a descriptive study and it does not aim to find a set of prescriptions or
norms for choosing the best way to distribute knowledge under specific conditions, even though the
results from the study could be used for this very purpose. The study will yield three primary results.
The first result is the survey of different emerging knowledge distribution means, yielding a list of
computer mediated distribution means that individuals use, thus showing us what these may be. Then,
there is the definition we use for identifying them. The definition of the term knowledge distribution
means is an important part of the result since it not only outlines the object of study but also helps us
to raise questions and identify issues that we may want to explore. The definition may also be used for
identifying other distribution means or similar phenomena than those we have identified here. In order
to analyze and characterize these distribution means we created a conceptual framework, based on
existing theories, for the identification of relevant issues and characteristics. This framework or model
is the second result and it may be used when analyzing or exploring other means for distribution or
other similar phenomena. The framework could also be used, within academic perspective, for
creating theories or test hypotheses related to questions of usage of different knowledge distribution
means or similar phenomena. The third result is the actual analysis or characterization of the different
knowledge distribution means. The characterization may be used, in an academic perspective, for the
development of methods for better and effective knowledge distribution and raise further issues in
need of exploration. Within an industrial perspective the characterization may be used to reveal
potentials for support and effective management of knowledge distribution means.
The most obvious limitation in this investigation concerns the area or objects of investigation, i.e. the
knowledge distribution means. The study is limited to knowledge distribution means that are
computer-supported, although other types of media, such as paper may very well support some of the
identified knowledge distribution means. This limitation is consciously made on our part. The focus on
computer mediated distribution is made since it is there we have experienced the greatest development
during the last years and had the greatest impact on how individuals use these technologies as
described in a previous section. It is there that most of knowledge distribution means have been most
emerging in their usage, seen their greatest rise in the consciousness of individuals, and had their
broad usage enabled.
The second limitation concerns the choice of theories used for the characterization. There exist a
wealth of literature concerning different aspects why and how people communicate and distribute
information. The aim in identifying relevant literature for characterizing the means has not been
towards making a thorough literature survey but to pick out relevant and central theories related to
issues concerning the distribution of knowledge, as will be discussed in chapter 4.
1.4 Research method
Following the work of Galliers (1992) and Weick (1984), Cavaye (1996) distinguishes between the
terms research strategy and research method. She defines research strategy as a way of going about
one's research in making it operational, embodying a particular style and different methods. Research
method is defined as a way to systemize observation, describing ways of collecting evidence and
indicating the type of tools and techniques to be used during the data collection. Here we have also
distinguished between strategy and the method that we used in conducting the research. We begin by
defining our research strategy and then we follow up with a discussion of the process used in our
research.
5
Research within the scientific field of information systems exhibits a plethora of research strategies,
epistemological foundations of deriving knowledge, methods and tools for gathering, interpreting and
reporting findings pertaining to the research undertaken. However, there exists no cohesive basis from
which to launch research within this field. Visala (1991) notes that information systems research is
mostly fragmented and mostly uncoordinated, and that cumulative research traditions are rare. The
field is divided into a set of schools and trends, which have not always made explicit their basic
assumptions. The field of information systems research seems to be caught in a pre-paradigmatic state
as noted by Culnan (1987). Keen (1991) goes as far as stating that the field of information systems
research is a self-defined community rather than a field or discipline.
The questions raised and the phenomena or objects that are targeted for study help in laying the
foundation of the research strategy and choosing a research approach. They determine to a great extent
also which methods, techniques, and tools to employ for gathering, processing and presenting data and
results. What determines fruitful research, in terms of defining appropriate objects or phenomena for
study and questions to raise concerning these, is the paradigm of the scientific field (Kuhn 1970). The
established paradigm states what good science (normal science) is and defines the field of research.
The choice of research strategy or approach is as much a personal style, as Keen (1991) puts it, as it is
the unintentional choice due to the socialization into the research tradition that the researcher has gone
through, from a junior researcher working on the dissertation to the matured and well established
senior researcher (Orlikowski 1991).
What have guided the development of our research strategy are the issues and objects, or phenomena,
targeted for research. The first aim of our research, to make a survey of emerging knowledge
distribution means, implies an explorative research approach. Patel and Davidson (1991) states that
most research and research strategies may be classified according to how much is known in advance of
the problem area before the research commences, ranging from uncharted territory to established
theories. An explorative research aims to gather as much knowledge as possible about a specific
problem area by shedding light to as many aspects of the problem area as possible. Since this kind of
research aims to establish knowledge that may lead to further studies, creativity and potential for
innovation are important characteristics of the methods employed in this strategy. Usually several
techniques for gathering information or data are used. Therefore we find that an explorative research
approach is appropriate for the identification of these means since little is known in advance of what
these may be. We must determine their environment or setting as to establish where to look for them.
This narrows the definition of what these means are. Then we must define the criteria used for
selecting them.
The second aim of our research, to analyze the characteristics of these knowledge distribution means,
implies a descriptive research approach for characterizing the means. A descriptive research is
appropriate, according to Patel and Davidson (1991), when the problem area is somewhat known and
there exists a need to systematize the knowledge in the form of descriptive models. Descriptions made
may concern relationships of past tense or present tense. In a descriptive research strategy the methods
used concentrate on a subset of the aspects of the phenomena under study. The descriptions of these
aspects aim to be as detailed and thorough as possible. Here a description of each aspect by itself or a
description of the relationship between the aspects may be appropriate. However, we also apply a
conceptual research approach since in order to do the characterization we must create a framework
within which we may analyze the means and on which to base the identification of relevant
characteristics.
This type of investigation and characterization has its methodological predecessors. Other researchers
have made similar types of surveys and characterizations. Conklin (1987) made a survey of different
hypertext applications and characterized them according to a number of specific features. This offered
a model by which the different hypertext application could be evaluated against each other or towards
a specific norm. Rudström (1995) has made a similar type of survey and characterization regarding
applications of machine learning.
6
In order to investigate the emerging knowledge distribution means we have first established the setting
or the context for them. Since individuals distribute knowledge in order to enhance or promote
learning, either individual or organizational, we defined the setting to be within an organizational
learning context and the means we wished to study were reduced to those that distribute knowledge
within that context.
The definition of knowledge distribution means was also narrowed or sharpened to those that are
computer mediated. This was done since computer mediated distribution and communication has
become ever more important to organizations, therefore pending greater attention than other forms of
distribution. In order to launch the investigation of computer mediated knowledge distribution means
we narrowed the area for the investigation to the distribution means that can be found through the
Internet. The Internet as a medium for communication or distribution has become overly popular and
has had a tremendous impact in society and organizations. The technological platform of Internet has
in many organizations been incorporated into the existing technological platforms of communication
and distribution as Intranets, and has in some cases become the dominating ones, e.g. Internet-based
organizations such as Bokus.Com and CNET. Therefore the distribution means that can be found
through the Internet are representative of those that are used within organizations.
During the investigation we selected knowledge distribution means according to two criteria. First
they should be general types and yet distinct, those that resembled each other too much were reduced
to the most representative one of them. The means should exhibit unique features or distribute
knowledge in a distinct manner. Their degree of generality was also a factor that was taken into
consideration. Second they should show some degree of popularity, promise, or usage, we didn't want
to select any that only had a marginal usage. This was accomplished by investigating organizations,
such as CNET, and companies, such as the Microsoft ® Corporation, that used the Internet for
distributing knowledge. During the investigation the means were selected according to how many
references they had, how great their availability was, or how widespread their usage was in terms of
volume or audience. This produced a list of seven distinct types of knowledge distribution means,
which is presented in chapter 6.
After the investigation we continued with the construction a conceptual framework for identifying the
characteristics that we used for characterizing the distribution means. Since the setting of the
knowledge distribution means was established to be that of organizational learning we turned to Huber
(1991) and his model of organizational learning constructs. Each of these constructs for organizational
learning is interpretable from different levels. Here we used a model of interpretive levels of
knowledge systems (Lundberg 1999) to analyze the different constructs of organizational learning.
The foundation of the framework is presented and elaborated on in chapter 3.
Within the framework we then identified different theories and concepts related to the different
constructs of organizational learning and the levels of interpretation, although we remained mostly
focused on the knowledge distribution. The aim here was not to make a comprehensive literature
overview but to identify what we perceived to be relevant issues concerning knowledge distribution.
This was made through a literature study on different aspects of knowledge distribution that has been
published during recent years. Relevant issues were selected based on the importance and attention
that had been given them by other researchers, e.g. Huber (1991), Walsh and Ungson (1991), and Fulk
and Boyd (1991).
Based on the theoretical framework we then identified characteristics (or features) from the concepts
and theories related to knowledge distribution and the interpretive levels, and discussed possible
interpretations of them. These characteristics are presented in chapter 5. We then grouped the
characteristics according to the three levels of interpretation and constructed three models of the
knowledge distribution means. These then became a three level view of the knowledge distribution
means that we used for the analysis and characterization. These views are presented in chapter 7.
7
1.5 Overview of content
Here we will give a brief outline of what will be discussed and presented in the rest of the thesis.
In chapter 2, following this chapter, we approach the concept of knowledge through a literature study
in order to define it and differentiate it from the concept of information. We go on discussing and
presenting other aspects or activities related to the distribution of knowledge such as creation and
acquisition of knowledge (where does it come from?), managing of knowledge (how to handle it?),
organizational memory (where is it located or stored?), and finally organizational learning (why
distribute it?).
In chapter 3 we lay the foundation for constructing a framework in which we will identify concepts
and characteristics drawing on the work of Huber (1991) concerning the constructs of organizational
learning and the different levels for interpreting or analyzing knowledge systems.
In chapter 4 we present theoretical approaches in characterizing the different knowledge distribution
means. We present theories and concepts related to media choice and use for distributing knowledge,
the concept of genre (Yates and Orlikowski (1992), communication contexts, knowledge
transformation processes (Nonaka and Takeuchi 1995), and knowledge exchange (Goodman and Darr
1998).
In chapter 5 we start analyzing the framework in order to identify features and characteristics that we
will use for characterizing the knowledge distribution means. We start with an examination of the
concepts that are related to the framework and proceed to examine the different levels of interpretation
of the knowledge distribution.
In chapter 6 we present the results from the explorative investigation of different emerging knowledge
distribution means and we describe each of the identified distribution means.
In chapter 7 we proceed to analyze and characterize the different means for knowledge distribution by
relating them to identified features and characteristics from the framework by examining each level of
interpretation thus producing three views of characterizing the knowledge distribution means.
In chapter 8 we present and discuss the conclusions we have drawn based on the findings from the
study and finish by pointing upon future research directions.
8
2 Theoretical background on knowledge
The purpose of this chapter is to make clear what we mean by different concepts and aspects relating
to knowledge that we use throughout this thesis. There is a need to define what we mean with
knowledge. It is a central concept to our work and therefore pending a definition, especially in
relationship to the concept of information (which sometimes is used interchangeable with knowledge).
This is necessary since we later make use of characteristics of knowledge in our characterization.
Throughout the work we make references to activities and processes of creation, acquisition and
management of knowledge. Therefore they too need to be given an overview and proper attention. We
also make references to the concept organizational memory as the locus where the distributed
knowledge resides and what enables and enhances knowledge distribution (in a time perspective) and
organizational learning. Therefor we will discuss and define it here. Since we aim to investigate
knowledge distribution means within an organizational learning context (a driving force behind
knowledge distribution and exchange) it is only natural that we also here make clear what
organizational learning is all about.
2.1 Knowledge
What knowledge is and is not has been debated since the time of the old Greeks. When we think of
what knowledge is we somehow intuitively think of it as that which is known (the sum or range of
what has been discovered, perceived and inferred) or that it has something to do with learning the facts
about objects in the world about us. That it is processed information that enables us to act, to know
how something works, and etc.
There exist many common definitions of knowledge, for example: "that part of a person's information
which is in accord with established facts" (Dictionary of Psychological Terms, English-English, 1958,
p.385). Other definitions regard knowledge as process, the act of knowing, for example: "the state of
knowing" (American Heritage Dictionary of English Language, 1975, p.725).
Wiig (1996) defines knowledge as the insights, understanding, and practical know-how that we all
possess. Knowledge is the fundamental resource that allows us to function intelligently. Over time,
considerable knowledge is also transformed to other manifestations such as books, technology,
practices and traditions, within organizations of all kind and in society in general. These
transformations result in cumulated expertise and, when used appropriately, increased effectiveness.
Knowledge is one, if not the, principal factor that makes personal, organizational, and societal
intelligent behavior possible.
According to Baskerville (1998) knowledge is a characteristic of the living mind. A person must
interpret and internalize it. It becomes embedded in human routines and norms. Alter (1996) states
further that knowledge is a combination of instincts, ideas, rules, and procedures that guide actions and
decisions. People need knowledge to use information effectively.
Wijnhoven (1998) sees knowledge as a collection of concrete experiences, or a set of abstract
conceptualizations. Concrete experiences consist of stories, feelings, data, and opinions about what has
been observed. Abstract models may be science (containing laws, theorems, and procedures accepted
as valid knowledge) or judgment (containing workable knowledge in the form of policy-rules,
probabilities, and heuristics). Science is public and accepted knowledge, whereas judgment is
uncertain, often private and untested knowledge. According to this definition knowledge encompasses
both know-how and know-why.
Knowledge can be viewed in respect to two dimensions: the epistemological dimension, i.e.
knowledge is either tacit or explicit; and the ontological dimension, the social interactive dimension
(ideas/knowledge is created with the individual and interaction holds grave importance).
From the epistemological perspective, knowledge is linked to beliefs: "In knowledge, a belief is linked
to the fact believed; without this linkage there may be true belief but there will not be knowledge" (R.
9
Nozick. Philosophical Explanations, 1981, p.81). Nonaka (1994) endorse the same epistemological
view of knowledge but adds that it is "justified true beliefs" and thereby recognizing not only the
absolute, static and non-human nature of knowledge but also recognize it as a dynamic processes in
which the individuals strive to justify personal believes.
Information (or knowledge) can be viewed in both a syntactic and a semantic aspect. The syntactic
aspect deals with the amount of information. Here one may refer to the work of Shannon and Weaver
(1949). The semantic aspect deals with the meaning or content of information. Here one may well
refer the infological equation of Langefors (1973), defining information as a function of the data
observed, background knowledge of the observer and amount of time for observation, i.e. I = i(D,B,dt)
and D = i-1(I,B,dt).
Brookes (1980) regards knowledge as a structure of concepts linked in their relations and information
as a small part of such a structure. The knowledge structure can be subjective or objective. Brookes
(1980) express this relationship by what he calls the fundamental equation: K(S) + dI = (S + dS),
where K(S) = the knowledge structure, dI = the increment of information, and S= the effect of the
modification.
Nonaka (1994) further states that the informal way of thinking of knowledge is that it is processed
information that enables us to act; to know how something works; and etc. Within the perspective of
Information Systems field the interest lies in how to express or represent knowledge, distribute the
knowledge or information, and how to create knowledge (through innovation, i.e. defining and
creating a problem and developing knowledge in order to solve the problem).
We adhere to the definitions of Wijnhoven (1998) and Nonaka (1994) stating that knowledge is
justified true beliefs of tacit or explicit nature and that may it be viewed as both know-how and knowwhy.
2.2 Knowledge and information
Many definitions of knowledge seem to be interchangeable with the definition of information and
many use the word information and knowledge almost synonymously. Therefore there is a need to
reflect a bit on the difference between the two. Machlup (1983) defines information as a flow of
messages or meanings that might add to, restructure or change knowledge. Dretske (1981) elaborates
further by stating that information is that commodity capable of yielding knowledge and what
information a signal carries is what we can learn from it. Dretske (1981) goes on by stating that
knowledge is identified with the information-produced belief, but the information a person receives is
relative to what he or she already knows about the possibilities of the source.
Nonaka (1994) sums it up by stating that information is a flow of messages, while knowledge is
created and organized by the very flow of information, anchored on the commitment and beliefs of its
holder. Further, that knowledge is related to human action as it is directly guiding it. Information is a
necessary medium or material for initiating and formalizing knowledge and can be viewed from
syntactic and semantic perspectives. Syntactic aspects of information can be best described by the
work done by Shannon and Weaver (1949) concerning the amount of information with regard to its
meaning or value.
Dretske (1981) points out that a genuine theory of information would be a theory about the content of
our messages, not a theory about the form in which this content is embodied. The semantic aspects of
information should be more relevant as it focuses on the conveyed meaning, i.e. seen from a semantic
point of view information means literally that it contains new meaning. Thus the semantic aspect of
information is indeed appropriate to study as it deals with the conveyed meaning so important in the
knowledge creation out of the chaotic and equivocal state of information.
10
2.3 Knowledge creation and acquisition
How then is knowledge created? Where does it come from and how is it acquired?
How knowledge is acquired has something to do with the process of gathering and learning. Levitt and
March (1988) define the process of knowledge acquisition within organizations as the gathering and
placing of knowledge into memory stores. Wijnhoven (1998) defines the following adoption
mechanism: learning from direct experience, where trial and error experiences are recorded; learning
from experiences of others, demanding that a search process exists in order to locate the desired
knowledge, however taking over knowledge derived from the experience of others may pose a
problem since the acquired knowledge has no universal validity; and learning through transactions of
knowledge, memory content is acquired through a purchasing process, the desired knowledge's value
must be measured somehow in order to buy it and the payment may be other memory content,
services, etc.
Nonaka (1994) states that there are three dimensions tied to the individual with respect to the
knowledge creation process: the intention or the values; degree of autonomy, self-motivation, or
liberty; and degree of fluctuation, interaction, change, contradictions, or anomalies (c.f. Kuhn's work
concerning the paradigm and revolution, 1970). In an ontological dimension it is only individuals or
group of individuals that can create knowledge.
Knowledge is created through the act of transmission or conversion. Knowledge made explicit is
knowledge that has been brought forth and represented in one form or another as opposed to tacit
knowledge. Knowledge is thus created through the act of transmitting/converting it to/from
tacit/explicit form. Here Nonaka (1994) contributes with a model of knowledge transmission or
conversion, taking its basis on the distinction Polyani (1958) makes between tacit and explicit
knowledge.
This creates four possible knowledge creation processes, se figure 2.1: 1) tacit to tacit, i.e.
socialization, without language as in an apprenticeship relation ("on the job training"); 2) tacit to
explicit, i.e. externalization of knowledge, representing the knowledge in an external formal form,
codifying and sharing it; 3) explicit to explicit, i.e. combinations such as formal meetings, telephone
calls, or computer processing systems; and 4) explicit to tacit, i.e. internalization, the process of
learning, assimilating external knowledge to the internal knowledge structure.
To
Tacit
Tacit
1) Socialization
Explicit
4) Externalization
Explicit
3) Internalization
2) Combination
From
Figure 2.1. Knowledge transformations (Nonaka 1994, p.19).
According to Nonaka (1994) there is a need within the socialization process to create an arena for
interchange of knowledge. With respect to the management of organizational knowledge creation he
proposes that within the arena one should strive for creative chaos, redundancy of information and
requisite variety. The knowledge creation process that we have touched on here will be explored
further in a later chapter.
2.4 Managing knowledge
Knowledge management is concerned with the representation and processing of knowledge by
humans, machines, organization, and societies. However knowledge management is not only about
managing the processes that includes knowledge work but also the managing of methods for these
11
processes. It is also at a higher/academic level concerned with the methods for managing the
knowledge management methods themselves.
Zwass (1998) states that knowledge management is the employment of organizational methods,
procedures, and information systems, used to collect the knowledge and experience of the members of
the organization and to bring them to bear on present problems and opportunities. Whereas Laudon
and Laudon (1996) views knowledge management as the process of systematically and actively
managing and leveraging the stores of knowledge in an organization.
Angus and Patel (1998) brings us a four-process view of what knowledge management is about,
stating that it consists of four major processes: gathering (including activities such as data entry, OCR
and scanning, voice input, pulling information from various sources and searching for information to
include), organizing (including activities such as cataloging, indexing, filtering, and linking), refining
(including activities such as contextualizing, collaborating, compacting, projecting and mining), and
disseminating (including activities such as sharing, alerting, and pushing). One might suspect that they
use the concept of information and knowledge interchangeable since none of the above mentioned
activities are specific for knowledge work but applicable to information work in general.
2.5 Organizational memory
Organizational memories come into existence more from evolution than from design because
organizations need a variety of information, skills, and knowledge (Weick 1979). Stein and Zwass
(1995) define organizational memory as the means by which knowledge from the past exerts influence
on present organizational activities. This memory preserves the experience the firm has accumulated
in delivering its products and services to the marketplace. Thanks to this memory the firm can
continue its operations in the face of employee turnover. This is due to the fact that memory is retained
not only in the minds of the firm's employees but also in the firm's structure that casts these employees
into appropriate roles, in the business processes of the firm (e.g. order processing), and in the
corporate culture ("this is the way we treat the customers around here"). Elements of the
organizational memory have more and more come to reside in the software, the data and knowledge
bases of companies' information systems.
Walsh and Ungson (1991) define organizational memory as stored information from an organization's
history that can be brought to bear on present decisions. This information is stored as a consequence of
implementing decisions to which they refer, by individual recollections, and through shared
interpretations. The information can be considered as both the decision stimulus and the organization's
response that is preserved in particular storage bins (problem-solution "catalogues") and that has
behavioral consequences when retrieved.
The content of an organization's memory can limit the organization's perspective and alternatives, or
can expand the visible options when shared and integrated. Organizational memory can be used to
both develop and answer questions. The ways in which memory is structured and stored will affect its
accessibility and usefulness. Walsh and Ungson (1991) identify three critical elements to an
understanding of organizational memory: the retention structure of memory, i.e. the locus of
organizational memory; the processes by which the contents of organizational memory are
manipulated, acquired, stored, and retrieved; and the ways in which the use of organizational memory
affects group processes, organizational outcomes, and performance.
Wijnhoven (1998) defines the content of organizational memory as knowledge and information that
can be applied in operational activities, the know-how; and knowledge and information that gives the
theoretical, conceptual, and background understanding of know-how called know-why. Both knowhow and know-why are familiar from the earlier definitions of knowledge and constitutes the
operational memory. Wijnhoven (1998) further defines the content of organizational memory to
include knowledge and information about value and quality of existing skills, asset capabilities, and
information, affecting the need for improving and changing existing know-how and know-why (i.e.
12
meta-memory), and knowledge and information that are important for retrieving and using operational
memory and meta-memory (i.e. memory information).
Wijnhoven (1998) defines the means for organizational memory as both the processes and media for
processing. According to Stein and Zwass (1995) the processes for managing organizational memory
are acquisition (the input), retention (the storage), search or retrieval, and maintenance or update.
Wijnhoven (1998) adds yet another process and that is the dissemination when considering modern
network environments.
According to Walsh and Ungson (1991) the central component of organizational memory is
information about critical decisions made and problems solved. The information may include issues
that trigger an event, how the situation was handled, and the consequences of an organizational
response. Walsh and Ungson (1991) identify six different storage facilities or retention bins,
constituting the organizational memory, which this information is usually distributed across. The first
is individuals, containing experiences, professional skills, evaluation criteria and results, explanation
of procedures, decision rules, personal ethics and beliefs, performance criteria, individual routines, and
formal records. The second is culture such as inhibiting language, symbols, stories, frameworks,
personal networks, schemes, external communications, cultural routines, and norms base. The third is
transformations such as tasks, experiencing process knowledge, standard operation procedures,
agendas, technology, and patents. The fourth is structure such as organizational structures, job
descriptions, patterns, titles, role labels, social structure, formal structure, and communication
structure. The fifth is the ecology within the organization such as physical layout of workplace and
building architecture. The sixth is external archives such as client and market characteristics,
competition profiles, list of knowledgeable people and organizations, technology of competitors,
former employees, observers, news media, and regulatory bodies. Electronic repositories for
organizational memory need to draw from each of these facilities and needs to store a diverse set of
information. Here readers are referred to Walsh and Ungson (1991) for a further discussion on storage
requirements for the different storage facilities. According to Walsh and Ungson (1991) knowledge
about the decision stimulus can be retrieved from only two of the bins: individuals and culture.
Wijnhoven (1998) also include systems such as planning and decision systems, process control
systems, GroupWare, computer aided design systems, knowledge-based systems, and administrative
systems, as one media where organizational memory is embedded.
Walsh and Ungson (1991) point at three roles that organizational memory serves: an informational
role, a control role through reducing transaction costs by focusing on relevant options, and a political
role through the dependence and influence on actions of others. Memory acts to frame a new situation.
Decisions made within such a framework are likely to be more effective and will meet less resistance
than those considered without reference to organizational memory.
Levitt and March (1988) conclude with that not everything will or can be recorded in the
organizational memory. There will always be a good deal of experiences remaining unrecorded
because the costs are just too great.
2.6 Organizational learning
When discussing knowledge distribution within organization one must reflect on the purpose of the
distribution. There is always the purpose of individual learning, e.g. the matching of solutions to
problems, but within an organizational context knowledge distribution organization-wide is about the
learning of the organization as a whole.
There exist many definitions of organizational learning. Argyris (1993) talks about organizational
learning as a process of detecting and correcting errors. Huber (1991) states that an organization learns
if, through its processing of information, the range of its potential behaviors is changed. He bases
organizational learning on four related constructs: knowledge acquisition, obtaining knowledge;
information distribution, sharing information that leads to new information; information interpretation,
13
information is given one or more commonly understood meanings; and organizational memory, means
by which knowledge is stored for future use. Senge (1990) states that as an organization acquires
knowledge and modifies its behavior to reflect this, the organization is said to be learning.
Organizational memory, increasingly embedded in information systems, perpetuates what has been
learned and is a point of departure for further learning.
Huber (1991) point out that cognitive deficiencies of human memory, personnel turnover, information
overload, lack of knowledge of the existence or location of data and other such factors are often
suggested as limiting the effectiveness of organizational memories. When organizational members due
to such limitations or specialization of duties do not know what they know then organizational and
individual learning is retarded. Learning is a primary goal to more effectively processing of
information. Organizational memory is the primary tool for such learning.
Levitt and March (1988) argues similarly to Huber (1991) with the suggestion that an organization
learns by encoding inferences from history into routines that guide good behavior. According to Fiol
and Lyles (1985) organizational learning deals with either cognitive changes or behavioral changes by
the organization. Cognitive changes result in newer shared understanding and conceptual schemes by
organization members. Behavioral changes refer to those as described by Huber (1991). The work of
Huber will be explored further in a later chapter.
According to Daft and Huber (1987) there are two basic perspectives on organizational learning: the
system structural perspective and the interpretive perspective. The system structural perspective
emphasizes the acquisition and distribution of information as a resource that is necessary for learning.
Here the focus is on the processes of learning. Message routing, distribution of messages, and message
summarizing, compression of ideas without loss of meaning, are important processes. The interpretive
perspective focuses on a deeper reason for information exchange. Information has utility because it
can reduce uncertainty, the absence of information, and equivocality, the existence multiple and
conflicting interpretations about a situation, and can therefore change one's understanding about the
external world (Daft and Macintosh 1981). Collecting and organizing information into shared
repositories, such as an organizational memory, may reduce both uncertainty and equivocality.
Organizational learning can be viewed as the acquisition and sharing of assumptions and cognitive
maps among organizational members. Learning occurs when individuals compare their own
assumptions and maps to what actually occurs. Argyris and Schön (1978) argue that this comparison
leads to two types of learning, single-loop learning and double-loop learning. Single-loop learning
refers to that misconceptions are corrected, facts are changed, within a fixed infrastructure. Doubleloop learning is here defined as the questioning and reconstruction of existing perspectives;
organizational norms, assumptions, strategies, structures; interpretation frameworks; or decision
premises that underlies one's action and basis for understanding why and how we act in a certain way.
Hoffer and Valacich (1993) give us the following analogies concerning single-loop and double-loop
learning within a database. In single-loop learning the content of a database is altered whereas the
table and relationships remain the same. In double-loop learning new relationships, as well as content
is defined or redefined
Goodman and Darr (1998) view organizational learning as a process by which one unit, work-group or
individual, acquires knowledge from another unit within the same organization or in a different
organization. Learning on the individual level occurs when solutions from one unit are matched to
problems of an individual from another unit, i.e. the exchange of problems and solutions. Learning on
the organizational level occurs when the problem-solutions exchanges and consequences are
communicated and known by other organizational members. There must be some form of memory that
stores problem-solutions exchanges and consequences, such as defined by Walsh and Ungson (1991),
and there must be a mechanism for organizations to share their interpretations about the problemsolution exchanges and to update the organizational memory about their experiences.
14
3 Towards a framework for characterization
In this chapter we start building a framework for analyzing or characterizing the identified knowledge
distribution means. We do so basing this on the work of Huber (1991), concerning the constructs of
organizational learning where knowledge distribution is but one, and the levels of knowledge system
analysis (Lundberg 1999).
We find it appropriate to base our study within the context of organization learning since the main
reasons, the driving force or purpose, for distributing knowledge, within units or between units, are
either to promote organizational or individual learning. Organizational learning is the prime motivator
for distributing knowledge, unless we are in a situation of selling knowledge, then money is the
driving force. We note that the differences between knowledge distribution and a similar distribution
process, information distribution, here become apparent. If the purpose for distributing knowledge is
to promote learning, then information distribution within an organization, an essential part of the
organization's information processing, is done in order to reduce uncertainty and resolve ambiguity
(Daft and Lengel 1986).
Huber (1991) does not distinguish between individual learning as opposed to organizational learning.
He states that an organization learns if any of its units acquire knowledge that it recognize as
potentially useful to the organization. The organization learns even if not every one of its units is
partaking in the learning. As long as an individual is a member in an organization and learns
something that is useful to the organization then the organization is said to be learning. Here, however,
Nonaka and Takeuchi (1995) would not agree. They see organizational learning as part of the
knowledge creation process starting with the individuals of a unit, self-organized team, working
together to solve a problem. The knowledge created and experiences drawn must then be leveraged
through the rest of organization in a "spiral process" for the learning to take place. The knowledge
learned in one team must be made unrestrictedly public to the different units and individuals, at
different levels, throughout the rest of the organization so that they may or may not adopt the
knowledge in their own work at their own choosing.
3.1 Constructs of organizational learning
Huber (1991) argues that a unit learns if, through its processing of information, the range of its
potential behaviors is changed. This definition holds true whether the unit is a human, group,
organization, or society. The information processing involves activities such as acquiring, distributing,
and interpreting information. Here Huber (1991) uses the word information and knowledge almost
interchangeably.
Huber (1991) characterizes organizational learning in term of the attributes: existence, breadth,
elaborateness, and thoroughness. Existence of organizational learning refers to that an organization
learns if any of its units acquires knowledge that it recognizes as potentially useful to the organization.
An organization learns even if not every one of its components learns that something. Breadth of
organizational learning refers to that more organizational learning occurs when more of the
organization's components obtain this knowledge and recognize it as potentially useful. Elaborateness
of organizational learning refers to that more organizational learning occurs when more and more
varied interpretations are developed. Thoroughness of organizational learning refers to that
organizational learning occurs when more organizational units develop uniform comprehension or
understandings of the various interpretations.
Organizational learning, according to Huber (1991), is based or constituted of four different learningrelated constructs: knowledge acquisition, knowledge distribution, information interpretation, and
organizational memory.
Knowledge acquisition is the process by which knowledge is obtained. Here five processes through
which an organization acquires information or knowledge are identified: congenital learning,
experiential learning, vicarious learning, grafting, and searching. First, congenital learning refers to
15
that the organization's creators impart knowledge. An organization's congenital knowledge is a
combination of the knowledge inherited at its conception and the additional knowledge acquired prior
to its birth. Second, experiential learning refers to that organizations acquire some of its knowledge
through direct experience. This learning could sometimes be a result of intentional and systematic
efforts but is more frequently unintentional and unsystematic. Experiential learning is enhanced by the
ability to analyze the feedback between organizational actions and outcomes. Third, vicarious
learning, acquiring second-hand experiences. Organizations often attempt to learn about the strategies,
administrative practices, and especially technologies of other organizations in order to either imitate it
or incorporate it into their knowledge base. Fourth, grafting referring to that organizations increase
their store of knowledge by acquiring or grafting on new members who possess knowledge not
previously available within the organization. The fifth is searching and noticing. Searching refers to
the following processes: scanning, i.e. the relative wide-ranging sensing of the organization's external
environment; focused search, i.e. active search in narrow segment of the organization's internal or
external environment in response to actual or suspected problems or opportunities; and performance
monitoring, i.e. monitoring the effectiveness in fulfilling the organization's pre-established goals or
requirements of stakeholders. Noticing refers to the unintended acquisition of information about the
organization's external environment, internal conditions, or performance.
Information distribution is the process by which information from different sources is shared and
thereby leads to new information or understanding. This process is a determinant of both the
occurrence and breadth of organizational learning. Occurrence of organizational learning refers to that
organizational components often develop new information by piecing together items of information
that they obtain from other organizational units. Breadth of, or broadly based, organizational learning
refers to that when information is widely distributed in an organization so that more and more varied
sources for it exist, retrieval efforts are more likely to succeed. Organizational leaning is thus greatly
enhanced when units and individuals are able to effortlessly access the information or knowledge that
is spread throughout the organization.
Information interpretation is the process by which distributed information is given one or more
commonly understood interpretations, i.e. information is given meaning, events translated, and shared
understanding is developed. More learning has occurred when more and more varied interpretations
have been developed, because such development changes the range of the organization's potential
behavior. Learning has also occurred when more of the organization's units understand the nature of
the various interpretations held by other units, leading either to greater cooperation or inhibiting
cooperation. Huber (1990) points out that in either case an organization's range of potential behaviors
have changed and the organization has learned. The extent of shared interpretation of new information
is affected by a number of variables. First, the uniformity of prior cognitive maps possessed by the
organizational units, i.e. belief structures or frames of reference of individuals. Second, the uniformity
of the framing of the information as it is communicated. If the information is not uniformly framed,
e.g. with common language, prior distribution then uniform interpretation is less likely. Third, the
richness of the media used to convey information referring to the medium's potential to convey shared
understanding. Fourth, the load of information on the interpreting units may inhibit their capacity to
interpret new information if it is too great. Finally, the amount of unlearning that might be necessary
before new interpretation could be generated, i.e. changing previous beliefs, constraints, and behaviors
that may affect the process of interpretation in a non-optimal manner.
Organizational memory is the means by which knowledge is stored for future use, whether they are
computer supported databases, standard operating procedures and routines, or the minds of the
members of the organization, and whether the information or knowledge is hard or soft. Variables that
affect organizational memory are membership attrition, information distribution and organizational
interpretation of information, the norms and methods for storing information, and the methods for
locating and retrieving stored information. In emphasizing the importance of organizational memory
Huber (1991) states the following that information acquisition depends in many instances on attention,
which is directed by previous learning retained in memory. Information distribution is affected by
organizational decisions made, which are applied using information contained in memory. Information
16
interpretation is greatly affected by cognitive maps or frames of reference, which are undefinable
except in terms of a memory. The basic processes that contribute to the occurrence, breadth, and depth
of organizational learning depend on organizational memory. In conclusion Huber (1991) notes that as
a result of specialization, differentiation, and departmentalization organizations frequently do not
know what they know. By incorporating into the organizational memory computer facilities this kind
of problem can be reduced. Electronic stores of knowledge can be more easily spread and made
accessible to different units throughout the organization and thus enhance organizational learning.
3.2 Interpretive levels
When trying to characterize the different knowledge distribution means one may look on the different
levels the knowledge distribution means can be interpreted from. Lundberg (1999) describe three
different levels of a knowledge system (basically a social system interacting with artifacts such as
computer-based information systems or paper-represented information) can be interpreted from. These
levels are:
• The knowledge level, referring to the knowledge content, the ideas conveyed, and the knowledge
processes of a system.
• The representation level, referring to the representation systems used in the knowledge system,
such as the languages and other representational forms employed in the system.
• The technical level, referring to the media and artifacts that are used to convey the knowledge in
the systems.
These levels may be used to describe other types of systems. Communication theorists, such as Rogers
and Agarwala-Rogers (1976), may study the communication system from a technical level (the media
used for transmission of a message from a sender to a receiver), a representational level (the message
as such in terms of its language and structure), and a knowledge level or process level (studying the
effects on behavior the message has on the receiver and the feedback that it yields as proof that
message has been correctly received and understood).
Eco (1976) defines a code system, the s-code, through his illustrious example the "Watershed", as a
system built in three levels. A code consists of a set of signal governed by combinatorial laws, then a
set of states of notions or "contents", i.e. semantic system, and last a set of behavioral effects. Here a
binary electrical signal directing a warning lamp, with the possible values of "off" or "on" at the
syntactic level, may denote either "all is safe" or "danger" at the semantic level, and may connote at
the behavioral level the effects of either "rest easy" or "raise the alarm".
Texts may be studied from lexical, i.e. the symbols and signs, syntactical, i.e. structure and
combinatorial rules for the signs and symbols, and semantic levels, i.e. the meaning that is given to the
different structures and combinations of signals and symbols. According to Vickery and Vickery
(1992) an utterance or a sentence may analyzed in individual letters (or sounds, if it is spoken), words,
phrases, or clauses. At the representational level linguists distinguish lexemes, vocabulary words that
can take various forms, and morphemes, which are the smallest segment of a word that has semantic
significance. At the knowledge level sememes are found, which are the concepts represented by
lexemes or morphemes.
Vickery and Vickery's (1992) model of meaning transfer and personal knowledge show us a similar
approach as the one we are advocating. Here a message or a query used for transferring meaning is at
the technical level viewed as a set of linguistic symbols, which refers to the concept at the knowledge
level. At the representational level the message or query is viewed as a knowledge-structure made up
by combination of the linguistic symbols. Then at the knowledge level we find the concept that is the
object for the meaning transfer.
These different levels of interpreting the phenomena within respective scientific field are all akin to
the interpretive levels of technical (artifacts or means used for conveying or representing),
17
representational (structure and norms), and knowledge (the processes and content), and most types
information systems may be analyzed or studied from these three levels.
These levels of interpretation are also appropriate when it comes to describing and characterizing the
different knowledge distribution means. One may look on them from a technical level, (discussing
issues such as e.g. the media used for conveying the knowledge), a representation level, (discussing
issues such as e.g. language and structure), or from a knowledge level, (discussing issues concerning
aspects of content and processes supported).
3.3 The framework
Here we have brought together the constructs of organizational learning from Huber (1991) and the
levels of the knowledge system (Lundberg 1999) to form a framework with which we may use to
identify concepts, theories, and features with which we may study, interpret, and characterize the
different distribution means from, see figure 3.1. The exploration of concepts and theories within the
framework will be done in chapter 5.
Constructs of
organizational
learning
Knowledge
acquisition
Knowledge
distribution
Knowledge
interpretation
Organizational
memory
Levels of the knowledge system
Technical
Representational Knowledge
Figure 3.1. The theoretical framework.
Each of the different constructs of organizational learning may be interpreted or studied from
technical, representational, and knowledge level. At the technical level each constructs is related to
issues concerning technical aspects such as physical artifacts, e.g. hardware, and their capacities or
limits. At the representational level they are related to issues of structural aspects such as e.g. form and
substance, and language. Lastly, at the knowledge level each of the constructs is related to issues
concerning processes and content, such as e.g. adoption and contribution (Goodman and Darr 1998).
Looking at the different constructs we notice that knowledge acquisition will deal with activities
related to retrieval and search (Walsh and Ungson 1991) but also the matching of the knowledge to
ones needs (Goodman and Darr 1998). Knowledge distribution will deal with activities related to
transmission of the knowledge, either as a reply to a request or for contribution to the organizational
memory. Knowledge interpretation will deal with issues concerning overcoming inhibitors to
knowledge exchange (Goodman and Darr 1998) such as language, culture and alienation, and
ambiguity resolving (Daft and Lengel 1986). Organizational memory deals with issues related to the
storage of knowledge in order to make it available throughout the organization to all its members, such
as e.g. indexing techniques, maintenance and updating. Within the framework it is thus possible to
identify different features and characteristics, related to the means of the different constructs, at every
level of interpretation within the knowledge system.
Our focus in the study is on the knowledge distribution means and thus we are mainly interested in
identifying features and characteristics that are concerned with the knowledge distribution construct.
However sometimes the boundaries are not as strict as they would appear from the framework. There
are issues related to the different knowledge distribution means that concern the other constructs or
overlap them. Features or characteristics applicable to the knowledge distribution may well concern
18
the other constructs. Also, as we will see further on, the dividing line between the different levels of
interpretation may in some cases not be as clear as the framework would have it at a first glance. Some
features or characteristics may be applicable to several levels at the same time, e.g. the concept of
feedback (which will be explored later) is both applicable to the technical level (dealing with the
capacity of the media for transmission) and the knowledge level (dealing with aspects concerning the
communication process). In cases where there are theories and concepts identified which touch on
other constructs than the knowledge distribution it is because the constructs are also related to each
other and do not occur in isolation. Interpretation of knowledge underpins the activities of both the
distribution and the acquisition since the knowledge is dependent on the context in which it is
contributed and adopted, especially between different cultural units within organizations or in-between
organizations. Organizational memory is where the distributed knowledge is stored or will be stored,
as to make it available to other members or units throughout the organization, and it is where
organizational members or units go to acquire knowledge (c.f. Walsh and Ungson 1991). The whole or
sum of constructs constitute organizational learning and it is therefore very difficult to just take one
construct and study it in isolation due to the relations it has to the other constructs within the
organizational learning context.
In the following chapter we will present different theories dealing with different aspects and features
of the knowledge distribution from which we will identify characteristics that are relevant in
describing the different knowledge distribution means from different levels of interpreting the
knowledge system.
19
4 Theoretical background on characterization
In this chapter we will present different theories and concepts covering different aspects, features and
characteristics of the knowledge distribution all related to the levels through which we may interpret
and characterize the different means for distributing knowledge.
We have divided this chapter on the theoretical background into three different sections, grouping the
theories and concept according to issues relating to knowledge distribution they present and according
to the levels of interpretation from the framework. This division of the theoretical background is not
done strictly according to the levels of interpretation within the framework, since many of the theories
and concepts presented will cover aspects from different levels of interpretation at the same time.
Instead we will present them in the order, although it resembles that of the interpretive levels, that we
have grouped them according to the issues and concepts they are mainly concerned with.
First group of theories and concepts relate to media choice and use. Here we explore technical aspects
(as well as some behavioral aspects) of media and means for distributing knowledge. Communication
theory is rich with discussions and studies concerning media choice and use for distributing and
interpreting information (Daft and Lengel 1986, Fulk and Boyd 1991, and Poole and Jackson 1993).
The same issues here are relevant for the distribution of knowledge. Huber (1991) has identified media
richness and interactivity as concepts related to the knowledge distribution and knowledge
interpretation.
Second group of theories and concepts relate to structures and contexts used for distributing
knowledge as well as concepts related to the communication process that affects the knowledge
distribution and interpretation. The relation between the concept of genre and information distribution
within communication theory has been established (Yates and Orlikowski 1992, and Poole and
Jackson 1993). The concept of genre is relevant to us since recurrent situations of knowledge
distributions may involve means sharing the same form and substance to the extent that we may talk
about genre of knowledge distribution affecting which means are used. Huber and Daft (1987) talk
about that ambiguous information can be interpreted, creating shared understanding and promoting
learning, through feedback, use of rich media, and interactivity affected by the communication
environment (or context).
Third group of theories and concepts relate to the different processes for creating knowledge and
exchanging knowledge and factors affecting the exchange. The work of Nonaka and Takeuchi (1995)
has been included here as it goes beyond the notions of organizational learning and concentrates on the
more wide-ranging concept of organizational knowledge creation through transformation processes of
knowledge. They focus on the importance of individual learning as the basis for organizational
learning. Finally the work of Goodman and Darr (1998) has been identified as relevant because of the
concepts of adoption and contribution within the knowledge sharing in support for organizational
learning. They discuss inhibitors for organizational learning and how technology can help overcome
them. Their work has strong parallels to that of Huber (1991) since many of the activities and concepts
they discuss, within aspects of organizational learning, are related to those the organizational
constructs.
4.1 Media choice and use
Media choice deals with how individuals choose medium or means for contributing or adopting
knowledge, as Goodman and Darr (1998) point out, when one or more are available. A first guess
would be that some degree of efficiency is wanted, in the sense of time required by the contributor to
codify and to transmit the knowledge. However, efficiency is just one factor that accounts for the
choice individuals make.
Palme (1989) has in an attempt to explain successful usage of media pointed out some factors that
contributes to this success or criteria that must be met for this success:
• That the user has prior knowledge of how to use the media.
20
•
•
•
•
•
That the user has access to the media in his or her workspace.
Individuals that the user sees as important to communicate or share knowledge with, e.g.
superiors, experts or other authorities, actively use the media.
The user experiences a real need to communicate with the group of people, which can be reached
through the knowledge distribution, especially when sharing knowledge.
The feeling of togetherness and community with one or several groups of persons that actively
uses the media for knowledge distribution.
That the amount of communication or degree of knowledge distribution that is offered is in itself
large enough (critical mass). Often this is equated by the amount of people that can be reached is
large enough.
Poole and Jackson (1993) and Rudy (1996) have identified the following theories as relevant when
discussing media choice and effects of media choice: information richness, social influence, social
presence, symbolic meaning, critical mass, interactivity, and bindingness. These theories emanate all
from discussions of communication media choice for distributing information but are applicable to us
when it comes to distributing knowledge in the sense that we deal with the same issues when
determining the best way of distributing knowledge. It is here interesting to note whether the choice of
a particular knowledge distribution means is related to the choice of a particular media by which the
knowledge is distributed. Included here is also a discussion of the theory of adaptive structuration as it
is an important theory in explaining behavioristic aspects behind the choice as well as use of media for
distribution. We also point on the work of Rice (1987) that has defined or identified different
dimensions with which to describe or understand media.
4.1.1 Information richness theory
This has been the most influential theory when it comes to explaining the choice of media. Daft and
Lengel (1986) defined information richness as the ability of information to change understanding
within a time interval. Communication transactions that overcome different frames of reference or
clarify ambiguous issues, in order to change understanding within a certain time limit, are considered
rich. Those transactions that fail to do so are considered poor. In a sense, richness pertains to the
learning capacity of a communication.
The choices of media for the communication transaction are based on its capacity or ability to
facilitate shared meaning or convey information among communicators. The proposed media can be
ranked in ascending order of information richness, independently of situation or context. Daft and
Lengel (1986) ranked the following media from poor to rich: numeric documents, impersonal written
documents, personal documents such as letters or memos, telephone, and face-to-face meetings. The
ranking of a medium is based on a blend of these five criteria:
• feedback, richer media have instant feedback, allowing a more interactive discussion;
• multiple cues, richer media have more cues, such as words, voice inflection and body gesture;
• language variety, richer media are able to cope with a wider variety of language and hence are
better at conveying meaning, for example natural language can convey a broader set of ideas and
concepts than numbers;
• personal focus, richer media allow personal feelings and emotions to be conveyed; and
• bandwidth, the number of different channels used for conveying the message.
Trevino, Lengel, and Daft (1987) altered this ranking a bit by including e-mail as being a bit poorer in
richness than telephone. They also excluded bandwidth as one of the criteria to base the ranking on.
Daft and Lengel (1986) tackle the question of why organizations process information. They argue that
organizations process information in order to reduce uncertainty and equivocality. Uncertainty is here
defined as not having the data you need to answer particular questions. Galbraith (1977) gives us a
further definition by saying that it is the difference between the amount of information required to
perform the task and the amount of information already possessed by the organization. The amount of
information needed for management coordination and control. Equivocality is defined, with respect to
21
the definition given of uncertainty, as not knowing which question to ask. There exist multiple
interpretations that are conflicting about an organizational situation, i.e. high degree of ambiguity
(Daft and Macintosh 1977). High equivocality means confusion and lack of understanding and to
resolve equivocality structural mechanisms employed, such as a specific media, must enable debate,
clarification, and enactment more than simply provide a large amount of data. Trevino et al. (1987)
point out that equivocality is a central feature of organizational life. For an organization to act
effectively it must reduce equivocality and create shared frameworks of meaning.
The theory developed by Daft and Lengel (1986) predicted that good managers would choose a
medium for a particular message on the basis of the equivocality of the message or situation.
According to Daft and Lengel (1986) good managers would choose an information-rich medium for
messages or situations of high equivocality, where a medium with low richness would not be capable
of conveying the necessary information due to reduced number of cues and restricted feedback. They
would then choose information-poor medium for messages or situations of low equivocality. Here a
medium with low richness is quite capable of conveying the information required, such as well
understood messages and standard data, and to use an information-rich medium would be overkill.
It has been shown that information richness or media richness theory as a model for explaining media
choice is not always appropriate. Markus (1994) points out that individuals choose media on the basis
of being seen to be behaving appropriately, i.e. conforming to a set of social norms established within
organization. This provides with another perspective on the choice of media.
Lee (1994) argues that the richness of a medium is not just a property of the medium itself, but also of
the organizational context in which the medium is used. In his work he pointed out that the usage of a
relatively poor media, such as e-mail, for sending a message within an organization brought about
wealth of interpretations. The receiver appropriates social constructs through a hermeneutic process
thus making the poorer media richer, giving the message a context for interpretation. The receiving
and interpreting a message is thus far from a simple passive process and could explain why there are
deviations from the norm given by the information richness theory.
Both the works of Markus (1994) and Lee (1994) are consistent with and relates to the social influence
model that will be discussed next.
4.1.2 Social influence model
Rudy (1996) points out that one important criticism of the theory of information richness is that the
theory doesn't take into account the effects that social influence from colleagues and organizational
norms has on an individual's choice of media. A number of criticisms of the theory of information
richness have been made (Fulk, Schmitz, and Steinfield 1990, and Fulk and Boyd 1991) since it has
failed to properly explain the anomalies found when studying choices of communication media. It has
been pointed out that it assumes that a medium has fixed properties, that individuals make their choice
independently of the people around them, that choice making is a purely cognitive, prospective and
objectively rational process, and that the choice is motivated solely by efficiency.
Emanating from the work of Fulk, Schmitz, and Steinfield (1990), and Fulk and Boyd (1991) an
alternative model based on social influence of technology use has been presented. This model
proposes that media choice depends on an individual's evaluation of media, the evaluation of the task
at hand, the media skill and experience, and a host of situational factors such as individual
characteristics, accessibility of media and time constraints. The model makes the following
assumptions that perceptions of electronic media vary across individuals in systematic ways and that
this variance is as important for media selection as is any constant component identified in the media
richness model. Media perceptions, such as richness are in part socially constructed. The process
involves at least four forms of social influences on both media perceptions and use: direct statement by
coworkers in the workplace, vicarious learning, norms for how media should be evaluated and used,
and social definitions of rationality. There will be considerable variation in the perceptions and use of
communication technology and that this variation will be linked to the social processes in play within
22
the organization. Since work groups are important sources of social support and regular interaction
there will be similar patterns of media use within groups and different patterns of media use across
groups. Any mismatching of media use and task ambiguity will occur where social forces come into
play to influence selection.
Fulk, Schmitz, and Steinfield (1990) point out that the evaluation of the media and of the task is partly
objective and partly socially influenced. Support for this theory comes from the empirical findings of
Schmitz and Fulk (1991) where they found that the media choices of an individual are influenced by
co-communicators and immediate superiors.
4.1.3 Social presence
Short, Williams, and Christie (1976), and Rice (1984) define media in terms of their social presence,
which refers to the degree which media allow a communicator to establish a personal connection with
others, the sense of psychological closeness achievable between partners using the media, and the
extent to which one feels aware of the person one is communicating with.
Different types of media are ranked according to the degree of social presence they bring about during
the communication act, from impersonal to personal. The degree of social presence varies according
the degree of restrictions that media has for conveying non-verbal cues. Here the importance of nonverbal cues in conveying emotions and subtleties of meaning may explain the difference between
media in the perception of social presence. Rice (1984) ranked five media in increasing degree of
social presence: business letters, telephone, multi-speaker audio, television, and face-to-face
discussion. Social presence may explain how different media make it easier to or harder for one party
to persuade another of their point of view.
Both Poole and Jackson (1993), and Rudy (1996) note that although both information richness and
social presence seem rather similar. The differences are that social presence distinctions are based on
interpersonal concerns such as the warmth of the communication, dealing with the effects of media
choice. Information richness distinctions are based on information-processing considerations such as
equivocality or uncertainty reduction, dealing with the media choice
4.1.4 Symbolic meaning
Rice (1984), and Trevino, Daft, and Lengel (1990) distinguish between media in terms of symbolic
meaning. There are two levels of symbolic meaning of media. A medium can carry symbolic meaning
that transcends the explicit message. Written media may symbolize formality and the need for
conveying formality may override the need to use a richer medium to deal with an ambiguous
communication situation. Media in themselves are symbols. The usage of a hi-tech media may show
others that the user is positive towards technological development and progressive. Thus, the effect of
symbolic meaning is to deflect media choice behavior away from the rational matching of task
ambiguity and media richness in some situations.
Poole and Jackson (1993) note that high social presence media are generally regarded as more
appropriate for emotional or conflictive communication, whereas low social presence media tend to be
selected for task oriented communication. High-information-richness media are better for highly
equivocal tasks, whereas low-information-richness media are better for low equivocality tasks. Poole
and Jackson (1993) point that we mustn't forget the impact of context as it plays an important role in
determining the social presence, information richness, and symbolic meaning, and thus the most
appropriate medium. An e-mail message from one manager to another may be of low social presence
if the first one refuses to meet the second personally, but a message from two who have never met may
be ranked higher in social presence. Rudy (1996) note that even though the theory of social presence
seems to be an explanatory factor behind choice of media it has not been proven to be a major one.
4.1.5 Critical mass theory
Another explanatory factor behind media choice is the theory of critical mass. Markus (1990)
developed the concept of critical mass in choosing a particular medium. She points out that it is not
23
worth any individual switching from traditional media to a new one until a reasonable number of the
people they want to communicate with are also using it. This suggests that in order for a medium to be
used successfully for the communication act there must be a fair number of users in the audience,
constituting a minimal critical mass. It has also been suggested that the critical mass may not only
relate to the mass of users but also of the mass of services accompanying the media (cf. Steinfield
1992).
4.1.6 Interactivity
Zack (1993) made a study of e-mail and face-to-face communication during the work periods in which
editions of a newspaper were created. He found out that rich media, such as face-to-face, was used
when establishing shared context and understanding, and that poorer media, such as e-mail, was used
in the working periods between for coordinating. However Zack (1993) didn't just discuss his findings
from a media richness theory perspective when looking at the choice of media. He introduces the
concept of interactivity, taking its basis from the work of Goffman (1981) concerning the importance
of interaction in communication and human life.
Although there is a lack of a clear and agreed on definition of the concept of interactivity Zack (1993)
describes several key characteristics of interactive communication. Some of these characteristics are
continuous exchange of information by both parties simultaneously, the use of non-verbal cues, the
unpredictability of the path of the conversation, and the ability to interrupt.
Zack (1993) notes that although in some ways interactivity is similar to media richness, in that it is
multi-dimensional and shares some of the dimensions of media richness, it goes beyond it in also
emphasizing the structure of the conversation. It is clear that the structure of communicative act, such
as the conversation, is far different when comparing face-to-face with poorer media, such as written
documents.
Zack's (1993) particular work shows how face-to-face meetings were preferred when conversations
requiring high interactivity, as in the process of establishing a shared context, were occurring. Once
this shared context was established a poorer medium could be used.
4.1.7 Bindingness
According to Ong (1982) and Innis (1972) bindingness distinguish the degree to which media binds
time and space, the degree that media provide permanent public records, such as writing, from those
that carry transient messages, such as speech. Media with high degree of bindingness, in this sense,
may better preserve past events, decisions and actions for future use. Media also differ in the degree to
which they are capable of joining geographically separated actors and areas, binding space, and how
tight this binding is in terms of how fast interchanges can be managed and how tightly they can be
monitored.
4.1.8 Adaptive structuration theory
Another aspect that highly influences the usage of media is of course how the media technology is
adopted and perceived by the group who utilizes it for communication (these ideas come close to the
social influence model). Advanced communication technology may be understood in terms of the
structures, rules and resources, which they provide for group interaction. Technological effects arise
not from the technology itself but from the choices that group members make about what features of
the technology to use and in what fashion. The choices are closely linked to ongoing group dynamics.
Poole and DeSanctis (1990) state that current research on effect has not adequately managed to
capture the complex interaction of technology and group processes. The research objectifies
technology by treating it as independent of the user. Users constitute and give meaning to
technologies. Until a user in a specific context applies the technology it is simply dead matter.
Poole and DeSanctis (1990) give us the theory of adaptive structuration stating that technology use is a
social process involving the construction and recreation of context and social world. The foundation of
the theory is based on studies of information exchange processes in group decision support systems,
24
research on decision development in small groups, and social theory on structuration processes. The
focus in the theory is on the interaction among technology, context, group processes, and social
structure.
Group structure is a dynamic entity that varies as a group engages in social interaction. The concept of
structuration (Giddens 1979) is the major explanatory force behind this dynamism. Structuration is
here the process by which systems are produced and reproduced by members' use of rules and
resources. Structure is both a medium and an outcome of action. A system is the group itself, structure
is the rules and resources that sustain the system, and structuration is the process of applying structure
to create and recreate systems.
Four relationships are highlighted through the theory of adaptive structuration. The first is how
different dimensions of the technology, such as asynchronous vs. synchronous meeting support, faceto-face vs. dispersed meeting support, degree of structure imposed on the group by the technology, and
so forth, affect how groups use the technology. The second is how dimensions of context, such as
group size, nature of group's task, degree of agreement on values, affect how group structures, such as
rules and resources, are used. The third shows that the uses of technologies and group structure
produce an appropriation process for the technology. It is the particular combination of structural
features used by groups in their practices. Fourth, appropriation affects group outcomes in that the
social system, constituting the group, is affected, which in turn affects the appropriation process by
reproducing it. In sum, technological dimensions affect use of technology and context dimensions
affect use of group structure. Both use of technology and use of group structure produce appropriation
processes that affect the social system, which reproduces the appropriation processes.
In order to understand technology effects, one must first understand how groups in the pursuits of their
own social practices use the technology.
4.1.9 Dimension of communication media
According to Rice (1987), all communication media, from dyadic face-to-face communication,
through business memo, telephones, mass media, and new media such as electronic and voice mail,
can be described as having more or less a set of specific attributes, either objectively or perceived by
users. These attributes belong to the following four broad dimensions: constraints, bandwidth,
interaction, and network factors.
Constraints refer to the extent the user can identify the sender, must know the other communicator's
address, is able to overcome communicator's selectivity, maintain privacy, must be temporally
proximate, must be geographically proximate, has access to the medium for initiating the process, is
able to store the message, can retrieve the content at a later time, is limited in message length, can
reprocess the content of the message, or can perform media transformations.
Bandwidth is the extent to which the medium allows the representation of different communication
modes such as physical distance, gestures, tone, ton, volume, or non-grammatical sounds. It is also the
extent that connotation, i.e. meaning or denotation, i.e. data, can be represented. Bandwidth also
concern social presence, or information richness, i.e. the extent which the message communicates the
presence of the participants, richness of interpersonal interaction, social and nonverbal cues,
ambiguity, etc. Bandwidth refers to both real technical limitations, the capacity of the medium, and
social or perceptual influences on the meaning communicated, the same content may involve many
more contextual cues, and may be far more subject to various interpretations, in a face-to-face
interaction compared to a business memo.
Interaction refers to the extent to which the medium allows two communicants to quickly respond and
complete a message exchange. It refers to the extent of control over pace and repetition user has, i.e.
slow down, speed up, pause or even terminate the communication process, but also to what extent the
user can ensure he is communicating with the intended participant. It refers also to the extent users
may exchange roles in the sequence of interaction being both senders and receivers, not being
25
restricted to communicate in chronological sequence, and replying in the same interaction as
receiving.
Network factors are the extent to which the medium involves different communication flows, such as
one-to-one, one-to-many, few-to-few, or many-to-many. It is also the extent which it involves the
necessity of a critical mass of other users interacting for the medium to have a value to a specific user,
different communication roles such as gatekeepers or liaisons might be necessary or possible, and
distortion in the communication due to that the pattern of communication inherently delay
communication, overload users, or allow one user to edit and then forward a message from another
user.
4.2 Theories relating to structure and context for knowledge distribution
In this section we present theories related to the substance and form of knowledge distribution means
such as the concept of genre from Yates and Orlikowski (1992). We also explore different
communication contexts wherein knowledge is distributed (Tubbs and Moss 1994). Drawing on
Goffman (1981) and Zack (1993) we explore the concepts of feedback and interactivity in the
communication process.
4.2.1 Genre of knowledge distribution
Reder and Schwab (1988) point out that if we look at users in real organizations we will observe
different genres of work and that media use will depend heavily on the genre. Genre, here, include
one-to-one meetings between supervisee and supervisor, project status meetings, brainstorming
meetings, and unscheduled and informal meetings to maintain social links. Reder and Schwab (1988)
point out further that in some genres one particular medium will be preferred and in others several will
be used depending on the users, switching from one medium to another, perhaps choosing a particular
medium for strategic reasons.
Yates and Orlikowski (1992) contribute to this line of reasoning in their work concerning genre of
organizational communication. They take their basis from the work of Giddens (1984) when studying
the evolving nature and role of organizational communication as individual actors interacts with social
institutions over time. This ongoing interaction between individuals and institutions can be seen as an
instance of what Giddens (1984) called structuration. Structuration theory involves the production,
reproduction, and transformation of social institutions, which are enacted through individuals' use of
social rules. These rules shape the action taken by individuals in organizations, while at the same time
by regularly drawing on the rules individuals reaffirm or modify the social institutions in an ongoing
recursive interaction.
Concept of genre, adapted form rhetorical theory and Miller's (1984) concept of rhetorical genres, is
here used by Yates and Orlikowski (1992) to explain organizational communication as a structuration
process. In the context of organizational communication it is applied to recognized types of
communication, such as letters, memoranda, or meetings, characterized by structural, linguistic, and
substantive conventions. These genres can be viewed as social institutions that both shape and are
shaped by individuals' communicative actions.
Drawing on Millers' (1984) concept of rhetorical genres Yates and Orlikowski (1992) propose similar
concept: genres of organizational communication. A genre of organizational communication is here
defined as a typified communicative action invoked in response to a socially defined recurrent
situation. The recurrent situation includes the history and nature of established practices, social
relations, and communication media within organizations. Similar substance and form characterize the
resulting genre. Substance refers to the social motives, themes, and topic being expressed in the
communication and form refers to the observable physical and linguistic features of the
communication. There exist at least three aspects of form in organizational communication: 1)
structural features, e.g. text-formatting devices, such as lists and fields, and devices for structuring
group interactions, such as an agenda and a chairperson for a meeting; 2) communication media, e.g.
26
pen and paper or face-to-face; and 3) language or symbol system, e.g. including linguistic
characteristics such as formality and the specialized vocabulary of technical or legal jargon.
Drawing on Giddens' (1984) notion of social rules Yates and Orlikowski (1992) posit that genres are
enacted through rules (genre rules), which associate appropriate elements of form and substance with
certain recurrent situations. When organizational members write business letters or engage in
meetings, they implicitly or explicitly draw on the genre rules of business letter or meeting to generate
substance and form of their documents or interactions. Genre rules may operate tacitly, through
socialized or habitual use of communicative form and substance, or they may be codified by an
individual or body into specific standards designed to regulate the form and substance of
communication.
Yates and Orlikowski (1992) elaborate further on the concept of genre and state that inherent to the
notion of genre is the issue of level of abstraction. Genres exist at various levels of abstractions, from
the very broad to the very specific, e.g. from the genre of letter it is possible to identify business letter
as a sub-genre within that genre. Related to level of abstraction is the issue of normative scope. This
issue deals with how extensively shared the social norms of a recurrent communicative situation,
along with characteristic subjective and formal features, must be to qualify as a genre, i.e. how high
must the degree of universality be. Since recurrent situations may be socially defined at any level
within an organization, from individual to inter-organizational, the normative scope of the genre may
vary with the same degree. Yates and Orlikowski (1992) conclude that a genre with broad normative
scope also is more likely to be at a high level of abstraction and vice versa.
Yates and Orlikowski (1992) observed that the notion of communication media is used variously and
inconsistently by different researchers and has often been confused with that of genre, but genre is not
media. Stohl and Redding (1987) identified this problem using the term format rather than genre. They
noted that the dividing line between medium and format is ever so fuzzy, e.g. between telephone, print
and oral media on the one hand and conversations, interviews, committee meetings, letters and inhouse presentation are all formats on the other hand.
Yates and Orlikowski (1992) argue that medium may play a role in both the recurrent situation and the
form of a genre. A recurrent situation may include a specific medium, where e-mail messages
typically evoke an e-mail response. Alternatively, a medium may be conceived as an aspect of a
genre's form; e.g. letters are traditionally conceived of as paper-based. Media are the physical means
by which communication is created, transmitted, or stored. Though a genre's form may at some point
include the medium, that genre may also expand into other media as with accounting records that have
mitigated from clay tablets, to ledger books, to punched cards, and nowadays electronic files. The
distinction between genre and media raises question about media richness continuum, which combines
media and genre on a single scale. Memo and bulletin are different genres traditionally associated with
the same medium and should occupy the same point on the scale. The fact that they do not do that
indicates or suggests that genre mediates the influence of communication media.
A question to raise here is whether it is possible to explain usage of different knowledge distribution
means from a genre perspective as to identify or explain different genres of knowledge distribution
whose substance and form are more appropriate to specific distribution means. The choice of a
specific knowledge distribution means may not then be an arbitrary choice but may stem from socially
influenced and recurrent communication actions.
4.2.2 Communication contexts
Tubbs and Moss (1994) define communication broadly as the sharing of experiences and in a sense
every living organism is to some extent sharing experiences. What is specific to human
communication is our ability to create and use symbols enabling us to share experiences indirectly and
vicariously. Communication, in broad sense, then becomes a process of creating a meaning between
two or more people. Rogers and Agarwala-Rogers (1976) define communication as the process that
27
transfers an idea from a source to a receiver with the intention of changing his or her behavior. Such
behavior may then encompass a change in knowledge or attitude as well as in overt behavior.
Organizational communication, according to Rogers and Agarwala-Rogers (1976), is communication
that takes place within an organization and between an organization and its environment. Tubbs and
Moss (1994) continue with that organizational communication is the flow of messages within in a
network of interdependent relationships, concerned not only with the effectiveness of the individual
communication but with the role of communication in contributing to or detracting from the effective
functioning of the total organization. Organizational communication is an essential part of the
organization's information processing (Daft and Lengel 1986), concerned with the reduction of
uncertainty and resolving of ambiguity. According to Tubbs and Moss (1994) organizational
communication may take place in any of the following communications context.
First, two-person, dyadic, or one-to-one communication is the basic unit of communication and
includes most of the informal, everyday exchanges that we engage in. Second, small-group or few-tofew communication is the process by three or more members of a group exchange verbal and
nonverbal message in an attempt to influence one another. In small groups the degree of intimacy,
participation, and satisfaction tend to be lower than in a two-person communication. Within smallgroup communication the focus is on ways to improve the work, i.e. problem solving and decision
making. Both two-person and small-group constitute a communication framework for interactive twoway transactions that emphasizes the personal exchange of messages between individuals or groups
who can give each other direct and immediate feedback. Third, mass or one-to-many communication
involves communication that is mediated. The source of a message communicates through some print
or electronic medium. The messages are intended for an audience that is relatively large,
heterogeneous, and anonymous to the source. Communication is here modeled as a one-way
transmission with delayed or indirect feedback, if any since opportunities for feedback are severely
limited as compared to two-person or small-group communication.
This gives us three different communication modes or contexts, one-to-one, few-to-few, or one-tomany, influencing the choice of knowledge distribution means and knowledge distribution processes.
These modes or contexts can be used to classify the distribution means since they may be regarded as
supporting a specific context or be a product of a specific context in which the knowledge is
distributed.
Intercultural communication aspects permeate or underlie the other communication contexts to
varying degree within organizational communication. Tubbs and Moss (1994) define intercultural
communication to be communication between members off different cultures, whether defined in
terms or racial, ethnic, or socioeconomic differences, or as a combination of these differences. Culture
is defined as a way of life developed and shared by a group of people passed down from generation to
generation. Cultural aspects of organizational communication play an important role when distributing
knowledge within units, between units in an organization, or between organizations. Goodman and
Darr (1998) also discuss effects of cultural communication or contexts on the knowledge distribution.
They talk about the inhibitors such as language barriers and alienation affecting the processes of
adopting or contributing knowledge between different units within the organization and between units
in different organizations. Cultural aspects on adopting are the degree of anonymity of the source and
the degree to which the knowledge is verifiable. Cultural aspects on the contributing may be lack of
incitement to help strangers when feedback may be lacking.
Rice (1984) point out that the emergence of new advanced communication technologies seem to
combine aspects of both interpersonal interaction, one-to-one and few-to-few, and mass media, one-tomany, e.g. e-mail or electronic conference systems. The line between the various contexts begins to
blur and it is unclear whether models based on mass media or face-to-face contexts are adequate.
However, these two models are still the most predominating one for describing or modeling
communication processes.
28
4.2.3 Feedback in the communication process
When we talk about feedback we often think of it as an answer to a question, a response or reaction to
a stimulus (request, statement, or change in condition), or as Luft (1969) defines it: "the return to you
of behavior that you have generated" (p. 116).
This view of feedback refers to the concept of response in an unchained dialogical unit (Goffman
1981) and is somewhat non-contextual in the sense that it reduces the notion of feedback to something
that can be automatically and mechanistically retrieved from a database. Here feedback is nothing
more than a response to a stimulus. This view of feedback is very unproblematic but also at the same
time not a very useful notion in describing the full range of feedback that might take place in a
knowledge distribution process.
One may argue that feedback in the form of an answer to a question may be regarded as feedback and
should be sufficient to be called feedback. Even though, as Goffman (1981) argues, that the question is
made expecting the answer, looking forward to it and is dependent on it, and the answer returned is
totally dependent on the question asked it still nothing more than a stimulus-response sequence. The
strong coupling between question and answer in this case doesn't solely justify the concept of
feedback.
When talking about feedback it seems to be more fruitful to talk about it within in a historical context
or of how the dialogical units are chained (Goffman 1981). Within a historical frame and context each
response, evoked by a stimulus is dependent on what has gone on before. We come here very close to
Goffman's (1981) definition of interactive natural conversation where one aspect is how the dialogical
units are chained two each other, i.e. what the responses refer to.
Within a historical context the questioning and answering is done depending on and relating to that
which has gone before. A question posed a time t+1 is dependent of what the question posed a time t
was. The answer to the question at time t+1 is returned transiently dependent on the question at time t
since the question posed at time t+1 may be evoked by the answer received at time t. A question posed
at time t+2 is thus directly dependent on the question at time t+1 and transiently or indirectly
dependent of the question at time t. The question at time t+2 would then have been posed differently
had it been directly dependent on the question at time t.
This depicts a series of statements and responses, feedback, that are highly inter-chained, taking into
account the historical context. We note that feedback in human conversation or dialog is seldom
offered outside a historical context. Here we come close to a definition of feedback resembling the
concept of interactivity of both Zack (1993) and Goffman (1981).
4.2.4 Interactivity in the communication process
When Goffman (1981) talks about everyday conversation he defines it to be a chained set of dialogical
units or terms of talk, i.e. consisting of chained set of statements and responses. Statements are here
called references of response since they are not restricted to any specific type of utterance, verbal or
non-verbal. A reference of response may even be a previous response.
Responses may be verbal, nonverbal, action, or even self-reflective, i.e. self-responding. Response
may focus on the content or the process of the conversation. Responses may break frame of talk or
context and refer to aspects of how previous statements were referred to. Messages may not only refer
to the previous message but also to the way previous message referred to earlier ones. A reply to a
question is just one type of response. The chaining of dialogical units now becomes a chaining of
responses to their references whatever they may be.
All these characteristics, states Goffman (1981), is what distinguishes highly interactive everyday
natural conversation from the uninterrupted and non-interactive "perfect talk" of newsreaders, or
written communication such as theater plays or novels.
29
4.3 Theories on knowledge content and knowledge processes
This section deals with characteristics of distribution means that either facilitate the knowledge
creation process, i.e. learning or sharing; or help reducing or overcoming inhibitors to the knowledge
distribution. The main theories explored in this section are the ones contributed by Nonaka (1995)
concerning knowledge creation or transformation processes, Goodman and Darr (1998), concerning
activities of adopting and contributing of knowledge related to organizational learning.
4.3.1 Knowledge creation processes
Here we look on the work of Nonaka and Takeuchi (1995) related to organizational knowledge
creation. They have studied how organizational knowledge is created and leveraged within companies.
By organizational knowledge creation they mean the capability of a company as a whole to create new
knowledge, disseminate it throughout the organization, and embody it in products, services, and
systems. Understanding of this process is the key in understanding how innovation may come about.
In laying the foundation for explaining the organizational creation process Nonaka and Takeuchi
(1995) have first drawn on Polyani's (1966) notion of tacit knowledge to point on the difference and
distinguish between implicit and explicit knowledge from an epistemological aspect. Tacit knowledge
is personal, context-specific, and therefore hard to formalize and communicate whereas explicit or
codified knowledge refers to knowledge that is transmittable in a formal and systematic language.
Nonaka and Takeuchi (1995) state that knowledge is created through a conversion between these two
types. The interaction of these two types of knowledge, called knowledge conversions, yields four
modes or processes. Socialization (tacit to tacit) is a process of sharing experiences and thereby
creating tacit knowledge such as shared mental models and technical skills. The tacit knowledge can
be acquired directly from others without using language, e.g. in the case of an apprentice learning a
craft from a master. Externalization (tacit to explicit) is a process of articulating tacit knowledge into
explicit concepts, thus creating new explicit knowledge from tacit knowledge through the usage of
metaphors, analogies, hypotheses, or models. Combination (explicit to explicit) is the process by
which concepts are systematized into a knowledge system. Individuals exchange and combine
knowledge through such media as documents, meetings, telephone conversations, or computerized
communication networks, e.g. a controller at a firm gathering facts and figures from different sources,
such as databases and documents, and put them all together in a report. Internalization (explicit to
tacit) is a process of embodying explicit knowledge into tacit knowledge. This process, the authors
note, is closely related to learning by doing where an individual re-experiences other people's
experiences. However internalization could also take place when reading or listening to a success story
makes some members of the organization feel the realism and essence of the story, thus changing the
experiences of the past into tacit mental models.
According to Nonaka and Takeuchi (1995) the content of the knowledge created by each mode of
knowledge conversion is naturally different. Socialization yields what can be called sympathized
knowledge, externalization outputs conceptual knowledge created through the usage of metaphors and
analogies, combination giving rise to systemic knowledge, and internalization producing operational
knowledge.
Nonaka and Takeuchi (1995) argue that it is the individuals who perform this interaction not the
organization itself because it cannot create knowledge devoid of individuals. However if the
knowledge cannot be shared with others or is not amplified at the group or divisional level, then
knowledge does not spiral itself organizationally. The spiraling process of knowledge across different
ontological levels, from individual, to unit, to organizational level, is one key in understanding the
organizational knowledge creation process.
According to Nonaka and Takeuchi (1995) the knowledge creation modes are not independent of each
other, but their interaction produce a spiral when time is introduced as the third dimension, besides
epistemological and ontological dimensions. The spiral starts with the conversion from tacit to tacit,
the socialization mode, by building a team whose members shares their experiences and mental
30
models. Then successive rounds of meaningful dialogue trigger the externalization mode. Team
members use metaphors and analogies to help them articulate their own perspectives revealing hidden
tacit knowledge otherwise hard to communicate. The combination mode is facilitated when the
concepts formed by the team is combined with existing data as well as with knowledge that resides
outside the team in order to create more shareable specifications. Entering the internalization mode
team members internalize the new explicit knowledge that is shared throughout the organization in
order to broaden, extend and reframe their own tacit knowledge.
The organizational knowledge creation process takes place at the group level but the organization
provides the necessary enabling conditions, such as contexts or devices that facilitate group activities,
as well as the creation and accumulation of knowledge at individual level. Nonaka and Takeuchi
(1995) identified five enabling conditions that are required at the organizational level to transform the
four knowledge creation modes into a knowledge spiral. These are intention, autonomy, fluctuation
and creative chaos, redundancy, and requisite variety. Intention refers to an organization's aspiration to
its goals. This effort to achieve intentions often takes the form of strategy within a business setting,
where the most critical element of the corporate strategy is to conceptualize a vision about what kind
of knowledge should be developed. The organization may increase the chance of introducing
unexpected opportunities by allowing its members to act as autonomously as circumstances permit.
Autonomy may also help increasing the possibility that the individuals will motivate themselves to
create new knowledge. When fluctuation is introduced into an organization its members face a
breakdown of routine, habits or cognitive frameworks, that forces them to question the validity of their
basic attitudes towards the world. The continuous process of questioning and reconsidering existing
premises by individual members of the organization fosters organizational knowledge creation since
its members turn to social interaction, in the spirit of Berger and Luckmann (1966), for defining or
creating their new reality. Chaos is generated naturally when an organization faces a crisis situation.
Intentional chaos or creative chaos increases the tension within the organization and focuses the
attention of organizational members on defining the problem and resolving the crisis situation.
However, the authors point out, creative chaos can only be realized when organizational members
have the ability to reflect on their actions. Redundancy refers to the existence of information that goes
beyond the immediate operational requirements of organizational members. Sharing redundant
information promotes the sharing of tacit knowledge. Requisite variety refers to that an organization
internal diversity must match the variety and complexity of the environment in order to deal with
challenges posed by the environment.
Nonaka and Takeuchi (1995) have then pointed out that the organizational knowledge creation process
is nonlinear and highly interactive. They have then proposed a five-phase model of the process, which
consists of the sharing of tacit knowledge, the creation of concepts, the justification of concepts, the
building of archetypes, and the cross leveling of knowledge. In the sharing of tacit knowledge we need
a field where individuals can interact with each other through face-to-face dialogues sharing their
experiences and resolving differences. Typically a self-organizing team consists of members coming
from various functional departments working together in order to reach a common goal. In creating
concepts the members of the team articulate shared mental models through further continuous
dialogue and crystallize them into explicit concepts using multiple reasoning methods such as
deduction, induction, and abduction. In justifying concepts the newly created concepts are determined
if they are truly worthwhile for the organization and society. The organization conducts this
justification in order to see if the organizational intention is still intact, i.e. the new concepts are
relevant and useful according to the goals. In building archetypes the justified concepts are converted
into something tangible or concrete, an archetype or a prototype, that is built by combining newly
created explicit concepts with existing explicit knowledge. In the cross leveling of knowledge the new
concept moves on to a new cycle of knowledge creation at a different ontological level. It is here
essential that organizational units have the autonomy to take the knowledge developed somewhere
else and apply it freely across different levels and boundaries. The authors state that internal
fluctuation such as frequent rotation of personnel may help this cross leveling.
31
Nonaka and Takeuchi (1995) point out that this model is cyclic and across levels within the
organization. The first four phases move horizontally, but the fifth phase moves vertically across
organizational levels. They conclude with that the creation of organizational knowledge is a neverending iterative process that may go through any number of cycles continuously leveraging new
knowledge.
It is here interesting to ask whether the usage of a specific knowledge distribution means can be
explained from the aspect of how well they support the different knowledge conversion modes, such
as the externalization and internalization, or the phases of the knowledge creation process, such as the
sharing of tacit knowledge and the cross leveling of knowledge. One may as well question if the usage
is unintentional in the sense that specific means of knowledge distribution are inherently related to
different activities of the knowledge creation process.
4.3.2 Adopting and contributing knowledge
When looking on the characteristics of different knowledge distribution means one may look on the
work of Goodman and Darr (1998) where they have studied how computer-aided systems may help to
overcome inhibitors in organizational learning processes.
Goodman and Darr (1998) states that for organizational learning to occur there must be a decision to
contribute and to adopt knowledge. Contributing means that individuals who have discovered
solutions to problems are willing to share these with others in different locations within the
organization either as a response to a direct request or to some form of organizational memory.
Adopting means that one has a problem and is willing to search for possible solutions in other
locations within the organization either by targeting an employee or group of employees from another
unit, or some other form of organizational memory such a database or library.
Goodman and Darr (1998) state that in the decision to contribute there are two related activities. One
must formulate what has been learned and deliver it, i.e. presenting it in a way that is meaningful to
others. In these two activities there are inherent costs. First these activities take time and energy.
Providing help could evoke additional clarifications or more requests for assistance, thus increasing
the cost. Second, why should anyone contribute with solutions to problems in a distributed
environment? Absence of feedback about whether the advice was adopted or successful, which is
likely in a distributed environment, will affect the motivation for contributing. Rewards for helping
strangers will not be as strong as when one helps others who have similar problems or are connected to
the contributor. Third, learning how to contribute takes time and energy in keeping track of formatting
and other delivery requirements, time and energy that could be otherwise devoted to business.
However, shared values of cooperation or citizenship through the organization may motivate the act of
contribution unless a competitive organizational environment is encouraged.
Goodman and Darr (1998) continue with that in the decision to adopt one looks outside one's unit for
solutions to problems. They argue that despite the obvious benefits of possibly finding a solution there
are inherent costs or inhibitors related to the adoption. Searching for possible solutions outside one's
local environment requires time and effort. The matching of problems to solutions may require a
translation between two or more local representations emanating from usage of different language in
different units when describing problems and solutions, thus increasing the load of time and effort.
The adopter must rely on people who are not well known and on information not easily verifiable.
Also asking for help may create awareness with an unknown audience that there is a problem.
Reluctance to show a weakness may inhibit the motivation for adoption. All this is contrasted with the
situation where you go searching for solutions in your own work environment, your own backyard.
Here the people are known and due to perhaps similar work the matching of problems to solutions
should be easier. Being amongst friends and colleagues makes the information received easier to
verify.
Goodman and Darr (1998) also examine how the nature of that which is to be exchanged may affect
both the contribution and adoption. The problems and solutions are characterized in terms of the
32
environmental conditions surrounding the problem statements, the nature of the solution set, the rules
to implement the solutions, and the nature of the results of the solutions. Any variations of these
dimensions may affect the exchange and increase the difficulty. The complexity of the problem
inhibits exchange. Multiple attributes are needed to describe a complex problem and the
environmental conditions surrounding it. There may be many solutions and many rules to implement
any particular solution. For a contributor it will be much more difficult to formulate and deliver a
complex solution. As the complexity of the problem-solution increase so does the cost of both
contribution and adoption.
Drawing on the work of Polyani (1966) Goodman and Darr (1998) state that the difficulty in
exchanging problems and solutions deals also with the distinction between explicit and tacit
knowledge. They point out that some problem statements and their environmental conditions may be
too difficult to articulate. The rules for implementing a solution may be understood but it may be
difficult to articulate why certain rules fit some solution and not others. They conclude here with that
since the contributor cannot articulate the key elements for the exchange, contribution and adoption
become difficult for geographically distributed units because there is no common meeting place for
creating and sharing understandings and interpretations of the nature of the problem-solution.
Goodman and Darr (1998) have studied how features of a computer-aided system (CAS), as defined in
chapter 1, can play an important role in facilitating the decision to contribute and decision to adopt
knowledge and help overcome or reduce inherent costs related to the problem-solution exchange. They
looked on how communication capabilities built in to the system can help overcome costs related to
contribution. Here for example features related to media richness (Daft and Lengel 1984) are explored.
They have looked on how capabilities related to the creation of an organizational memory (Walsh and
Ungson 1991) could support the storage of problems and solutions, an action related to organizational
learning, in order to make these accessible to others. They have looked on capabilities related to the
searching and matching of solutions to problems. Capabilities related to updating of the organizational
memory about the consequences of problem-solution exchanges, and capabilities implemented to
elicit, categorize and store new insight supporting the act of contributing have also been explored.
Goodman and Darr (1998) also point out that organizational context variables, besides technological
factors, are important to examine. Context variables can be viewed as increasing or decreasing the
costs inherent in the decisions to contribute or to adopt. One factor that effects the distribution of
problems and solutions is their characteristics. An organization's problem and solution characteristics
will be a function of its core technology. If the core technology is low in task variety and high in
analyzability then the distribution of problem-solution features may be low in complexity and
tacitness, thus reducing some of the costs of adoptions and contributions. Perceived adoption and
contribution costs will be affected by organizational factors such as culture and reward system. If an
organization encourages high degree of competition between units then the incitement for an exchange
of solutions and problems, i.e. helping other units, will be low as both the culture and reward systems
do not favor such actions. Another contextual factor that may affect a computer-aided systems ability
to have an impact on the reduction of costs is the existence of other communication and memory
systems that function independently of the computer-aided system. Any organization has a variety of
formal and informal communication systems that may affect the exchange of problems and solutions.
Finally Goodman and Darr (1998) have postulated four theoretical assumptions:
• Costs are inherent in the decisions to contribute and adopt knowledge and they reduce the
likelihood for these actions to occur.
• Characteristics of problem-solution, such as the degree of complexity and level of tacit
knowledge, affect these inherent costs. The likelihood of contributing and adopting knowledge
will decrease as the degree of complexity and level of tacit knowledge increase.
• The feature of a computer-aided system may reduce costs related to both contribution and
adoption, and that capabilities of such a system should match the problem-solution complexity and
tacitness.
33
•
Organizational context may moderate the impact a computer-aided system has on organizational
learning. As the features of a computer-aided system help reducing the inherent costs
organizational learning should increase, but if the organizational reward system doesn't encourage
sharing behavior then its impact on cost reduction will be hampered.
Here it is interesting to ask if the choice or usage of a certain knowledge distribution means is related
to the degree it help reduce or overcome inhibitors or costs inherent in contributing and adopting
knowledge.
34
5 Identified characteristics within the framework
In this chapter we will from the presented theoretical background, from the previous chapters, identify
concepts and features that relate to the different constructs and levels within the framework. This is
done in order to identify the characteristics by which we will describe the different knowledge
distribution means. These are dealt with in a coming chapter. The chapter, here, is divided in to two
sections where we first identify concepts and theories and then identify characteristics that we will use
for describing the different knowledge distribution means.
5.1 Concepts and theories within the framework
Within the framework se figure 5.1, we have identified the following concepts and theories as being
relevant for organizational learning based on the presented theories.
Constructs of
organizational
learning
Knowledge
acquisition
Knowledge
distribution
Levels of the knowledge system
Technical
Representational
Media [9, 11]
Communication context [10, 11],
Genre [6], Social influence [15], and
AST [12]
Media [9, 11], and Interactivity Communication context [10, 11]
[4, 11]
Genre [6]
Social influence [15] and AST [12]
Knowledge
interpretation
Media richness [3, 9], and
Interactivity [4, 11]
Organizational
memory
Retention media/bins [2, 13],
Media bindingness [8], and
Indexability [1, 3, 14]
[1] Goodman and Darr 1998
[2] Walsh and Ungson 1991
[3] Huber 1991
[4] Zack 1993
[5] Nonaka and Takeuchi 1995
[6] Yates and Orlikowski 1992
[7] Goffman 1981
[8] Innis 1972 and Ong 1982
Knowledge
Acquisition mode [3, 13],
Knowledge transformation
process [5], and Genre [6]
Adopting - contributing [1],
Broadcasting [1], Knowledge
transformation process [5],
Distribution mode, and Genre [6]
Matching [1], and Genre [6]
Genre [6], Communication context
[10], AST [12], Shared frames [2,
3], Interactivity [4, 7], and Feedback
[7, 9]
Genre [6], Best practice (problemUpdating and maintenance [1, 13,
solution) [1], and Decision stimulus 14], Storage, retrieval, and
and organizational response [2]
retention [2, 13, 14], and Genre
[6]
[9] Daft and Lengel 1986
[10] Tubbs and Moss 1994
[11] Rice 1987
[12] Poole and DeSanctis 1990
[13] Wijnhoven 1998
[14] Stein and Zwass 1995
[15] Fulk, Schmitz and Steinfield 1990
Figure 5.1. Concepts and features within the framework.
Knowledge acquisition deals with, at the technical level, the means, media, tools, artifacts and
technology that members of an organization utilizes for acquiring knowledge according to such
activities and processes as those defined by Huber (1991) and Wijnhoven (1998). In examining these
the discussion of media choice is relevant in the same manner as Daft and Lengel (1986) do when
talking about media choice being subject to the capacity it has according to its characteristics. Rice's
(1987) dimensions of media are also applicable here since they too concern aspects of media choice
for acquiring information or knowledge.
At the representational level, dealing with the structures and language used for acquiring knowledge,
the discussion of communication context (Tubbs and Moss 1994) is appropriate since they affect the
different acquisition processes by giving them frames or structures within which the processes for
acquisition take place. Rice's (1987) discussion of network factors, as a dimension of media is
appropriate here as well as it concerns the structural aspects of the acquisition processes. The concept
of genre (Yates and Orlikowski 1992) when talking about the form (structure and language) of
recurrent situations of knowledge acquisition is relevant. Finally, the form of representation is set into
a socially influenced context. Here the work of Fulk, Schmitz and Steinfield (1990) concerning social
35
influence models of technology use is most useful. From Poole and DeSanctis (1990) we get yet
another view of how the structures used are influenced by social contexts as well the technology
through their theory of adaptive structuration.
At the knowledge level we encounter the processes for acquiring knowledge. Walsh and Ungson
(1991) state that is it only individuals that acquire knowledge. Huber (1991) defines the processes for
acquiring as congenial, grafting, experimental, vicarious, and scanning and monitoring. Wijnhoven
(1998) concurs with Huber but talks about only three modes of acquisition, i.e. vicarious,
experimental, and transactional. This leads us to talk about different acquisition modes, which roughly
corresponds to the processes identified by Wijnhoven (1998). Here we may also analyze the
knowledge transformation processes (Nonaka and Takeuchi 1995) that are related to the acquisition
processes when discussing how the knowledge acquired was created or transformed. The concept of
genre (Yates and Orlikowski 1992) is once again appropriate since it, at this level, deals with the
substance (content) of the recurrent acquisition processes.
Knowledge distribution deals with at the technical level roughly with the same issues as those that
concerning the acquisition. However some other aspects concerning the media are relevant to explore
as well. Here we find that the concept of interactivity such as both Rice (1987) and Zack (1993), to
some extent, are fruitful to explore when looking at the capacity the media has for mediating the
"natural conversation" of Goffman (1981) in distributing knowledge or information.
At the representational level, we again experiences the same issues (and for the same reasons) as the
ones encountered in the knowledge acquisition. Rice (1987) also discusses the dimension of
constraints that media has on the distribution and sharing of knowledge between individuals that may
affect the structure, such as the procedures and rules, for the distribution.
At the knowledge level we encounter the processes of adopting and contributing of knowledge that
Goodman and Darr (1998) have identified and defined. These two processes deal with how and why
individuals share and distribute knowledge and are thus relevant two explore. Goodman and Darr
(1998) also talks about the concept of broadcasting as being a key functionality tied to organizational
memory. Here it is applicable because it concerns the dissemination of knowledge and therefore it
concerns the distribution process. The knowledge transformation processes of Nonaka and Takeuchi
(1996) and the concept of genre of Yates and Orlikowski (1992) are also fruitful to explore and for the
same reasons as in the knowledge acquisition process. At the knowledge level we may also talk about
distribution mode, which will be defined in the following section, which characterizes the knowledge
distribution process as either pushing, active pulling, or passive pulling. This concept concern the way
knowledge is exchanged and is thus worth exploring here.
Knowledge interpretation at the technical level deals with aspects of the means, media, tools, artifacts,
and technology in their capacity for conveying shared understanding and promoting learning. Here we
encounter issues concerning media richness, such as numbers of cues, language variety, possibility of
immediate feedback and interactivity (Daft and Lengel 1986, Rice 1987, Huber 1992, and Zack 1993).
These all deal with the capacity media has for conveying shared understanding and is a key concern
when distributing knowledge between units within or between organization where we may not
experience the same contextual frame as when distributing within same units between colleagues and
friends.
At the representational level knowledge interpretation concerns mostly the same issues as discussed
before, but they are now angled towards enhancing the shared understanding of knowledge, such as
the ability to use common language, common frames of references, and interpretive schemata as both
Huber (1991) and Walsh and Ungson (1992) point out. The concept of genre (Yates and Orlikowski
1992) as it relates to the form of recurrent situations of, for example, knowledge distribution that may
identify common structure and language used throughout the organizations is important to explore
here as it creates a platform for shared understanding. Here, we also encounter the concepts of both
feedback (Daft and Lengel 1986, and Goffman 1981) and interactivity (Goffman 1981 and Zack 1993)
36
in the communication process for creating or enhancing the interpretation of the knowledge. These are
important structural capabilities, although also process oriented to some extent, for sharing knowledge.
In a face-to-face situation for knowledge sharing these capabilities have their greatest capacity as
opposed to e-mail communication, which Zack (1993) points out in his study. Therefore we find these
two concepts relevant for the knowledge interpretation and fruitful to explore.
At the knowledge level we again find the concept of genre (Yates and Orlikowski 1992) to be fruitful
and for the same reason as at the representational level. Goodman and Darr (1992), when they discuss
the process for adopting of knowledge, identify the matching of the retrieved knowledge as an
important function when the individual try to find a solution to a problem situation.
Organizational memory at the technical level concerns, as both Walsh and Ungson (1991), and
Wijnhoven (1998) have pointed out, the different media where knowledge is retained. Innis (1972) and
Ong (1982) also point upon the bindingness of the media, i.e. its capacity for capturing the knowledge
in a more permanent form that is not as transient as oral communication in a face-to-face context. Here
we also encounter the concept of indexability. This dimension refers to the extent to which the
knowledge is searchable and retrievable by other means than just browsing. This implies that the
knowledge must be recorded by computer facilities and stored in a computer-supported memory. If we
are interested in storing the distributed knowledge for future reference and sharing it with other
members of the organization through electronic means then indexability becomes a key feature of the
knowledge distribution means. A knowledge distribution means is either indexable or not, there is no
gray zone explored here as to what extent it is indexable. Goodman and Darr (1998), Stein and Zwass
(1995), and Huber (1991) support this characteristic through their discussion of computer support for
organizational memory. When talking about the possibilities for storing and retrieving knowledge
effectively the indexability becomes a key concept or feature.
At the representational level issues concerning the structure for representing organizational memory
are dealt with. When we talk about the structure of the knowledge retained we may look at such
structures as the "best practice" (problem and solution along with the consequences) structure, that
Goodman and Darr (1998) discuss, or the decision stimulus and organizational response structure, that
Walsh and Ungson (1991) discusses. Once again the concept of genre (Yates and Orlikowski 1992) is
applicable since it concerns the structure and language, i.e. form, that the retained knowledge derived
from recurrent activities may yield.
At the knowledge level processes such as storing, searching, retrieving along with maintaining or
updating knowledge in the organizational memory are identified and discussed by Walsh and Ungson
(1991), Stein and Zwass (1995), Goodman and Darr (1998), and Wijnhoven (1998). Theses all
concern the vital functionality of making the knowledge available to organizational members
especially, as Stein and Zwass (1995) among others argue, if knowledge from past (historical
knowledge such as past experiences) is to have an impact on the future activities and guide good
behavior. Therefore these activities are important to explore here.
Having identified concepts and theories, which may be used in exploring the different interpretive
levels of organizational learning, we have arrived at framework that give us a workable foundation for
identifying the characteristics we will use when analyzing the different knowledge distribution means.
The identified characteristics and features will be presented and discussed in the next section.
5.2 Characteristics of knowledge distribution
In this section we have come to identify the characteristics, from the framework, which we will use to
describe the different knowledge distribution means. Our main interest in this study is the knowledge
distribution and as such we find that not every construct of organizational learning is of equal interest
and applicability to our study of the distribution means. The organizational learning construct of main
interest and focus is the knowledge distribution followed by the knowledge interpretation,
organizational memory, and knowledge acquisition. This ordering is reflected mostly through the
identified concepts or characteristics. Most of them are related to the different levels of interpretation
37
related to knowledge distribution and interpretation. However some of the identified characteristics are
also related to the other constructs because they are related to or affect the knowledge distribution
means. We have grouped them according to their level of interpretation and identified what scales or
units they are measured or described in.
5.2.1 Characteristics at the technical level
Here we have identified the following:
Indexability. This dimension refers to the extent to which the knowledge is searchable and retrievable
by other means than just browsing. This implies that the knowledge must be recorded by computer
facilities and stored in a computer-supported memory. If we are interested in storing the distributed
knowledge for future reference and sharing it with other members of the organization through
electronic means then indexability becomes a key feature of knowledge distribution means. A
knowledge distribution means is either indexable or not, there is no gray zone explored here as to what
extent it is indexable. This characteristics is supported by Goodman and Darr (1998), Stein and Zwass
(1995, and Huber (1991) through their discussion of how computer support for searching and
retrieving knowledge in an organizational memory.
Media richness (or information richness, Daft and Lengel 1986) relates to the capacity of the media for
conveying shared understanding through distribution and interpretation of knowledge (Huber 1991).
The concept comes from the media richness theory and the extent to which the media will contribute
in solving ambiguity (Daft and Lengel 1986) and therefore we talk about low respective high degree of
richness. However this is a rough measurement. Huber (1991) focuses on the number of cues and
degree of immediate feedback of the media. From the concept of media richness we adopt the
characteristics of capacity for cues, language variety, and feedback.
Cue capacity. This is the capacity of the media for transmitting different types of cues as defined by
Daft and Lengel (1986). The number of cues, such as somatic besides vocal, will facilitate the sharing
of framework for interpreting knowledge, this has also been discussed by Huber (1991). Individuals
engaged in communication through e.g. e-mail find that they are stripped of almost all the cues
normally used for communicating except linguistic style. Through the use of linguistic features such as
language style and use of symbols other than letter and numbers and capitalization of letters many
more cues may be transmitted. We grade capacity for transmitting cues as either high, medium
(experiencing to an extent some capacity for multiple cues), or low.
Language variety. This is the capacity of the media for conveying rich language instead of just
numbers. The greater the capacity of the media is for transmitting knowledge in rich language the
greater the capacity is for the knowledge distribution means to create shared understanding and
establishing common views. Rich language facilitates the matching (interpreting) of knowledge in
adopting it as well as the formulation of the knowledge in contributing it (Goodman and Darr 1998).
We grade the capacity of language variety as either rich, mediocre (experiencing to an extent some
variety), or poor.
Feedback. Huber (1991), Goffman (1981), Rogers and Agarwala-Rogers (1976), and Daft and Lengel
(1986) all examine the concept of feedback but from slightly different aspects and angles. It concerns
the capacity of the media used for distributing knowledge and to give immediate feedback (response
from the receiver and proof of that the knowledge has been properly received and understood).
Feedback at the technical level is an important characteristic for establishing shared understanding and
enhancing learning. Capacity for feedback may either be high, medium (experiencing to an extent
some feedback), or low.
Interactivity. This concept, coming from Zack (1993) and Rice (1987) (at this level also to some extent
from Goffman 1981), refers to the extent media support it in promoting shared understanding among
those involved in the knowledge distribution. This concept is similar to both feedback and cues as has
been pointed out by Zack (1993), but also by Goffman (1981) when defining the key characteristics
38
for the "natural conversation". However this concept transcends feedback in the sense that it can go
out of frame, which both Zack (1993) and Goffman (1981) has pointed on. As with both cues and
feedback this characteristic is important for the interpretation of knowledge. Capacity for interactivity
has also been graded as either high, medium (experiencing to an extent some interactivity), or low.
Bindingness. This characteristic refers to the extent media may bind or capture knowledge into more
permanent records as opposed to oral communication in a face-to-face situation as defined by Innis
(1972) and Ong (1982). This characteristic is also important as it relates to the extent that
organizational memory is supported by the distribution means. We classify the media capacity for
bindingness into either high, medium (experiencing to an extent some bindingness), or low.
5.2.2 Characteristics at the representational level
Here we have identified the following:
From the concept of genre (Yates and Orlikowski 1992) we get form, generated from recurrent
situations of the knowledge distribution means. Form deals with both the structure and language of the
knowledge distribution means and it is these two characteristics we are interested in here.
Language. Here it refers to the type of language used for representing knowledge and, at the
interpretive level, conveying shared understanding. A shared language between organizational units is
a key characteristic for distributing knowledge and it affects the incitement for both the adoption and
contribution of knowledge, as defined by Goodman and Darr (1998). Language, here, refers also to the
interpretive schemes and frames people use when adopting or interpreting knowledge (Huber 1991,
and Walsh and Ungson 1991). Language may be classified as either jargon or expert, casual or strict,
or domain specific or broad and general. Any one of these classifications may be used. We have not
given a specific set of classifications for the language but will characterize it when we analyze the
knowledge distribution means.
Structure. This is the structure the distributed knowledge has or the structure of the knowledge
distribution means. It refers not only to the structure as such but also to procedures, rules and norms
for distributing knowledge. Common views of knowledge (in the form of common ways of structuring
it and handling it as dictated by norms and rules) are important for the interpretation of knowledge
because they facilitate the formulation of knowledge when contributing and the matching when
adopting (Goodman and Darr 1998). Structure may be in the form of problem-solution, stimulusresponse, scenario oriented, hierarchical, tree, web, or incremental. Whatever structure used we have
not classified it here but will explore it when analyzing the knowledge distribution means.
Communication context. This is the type of communication context (Tubbs and Moss 1994) that the
knowledge distribution means is situated in or related to. Communication context is one factor that
affects the frames that individuals use for interpreting knowledge, which Huber (1991) among others
discusses. Here we distinguish between mass median (one-to-many and one-way), dyadic (one-to-one
and two-way), or group (few-to-few). These affect the extent of feedback and interactivity in the
communication or distribution process.
Degree of interactivity and feedback. This characteristic refers to the degree of interactivity, as both
Zack (1993) and Goffman (1981) has defined it, and degree of feedback, such as Goffman (1981) talks
about it, in the communication process that is supported by the knowledge distribution means. This is
a joint characteristic since they share many similarities, although interactivity goes beyond feedback in
the sense that it may go out of frame. This joint characteristic is both a structural aspect and a
procedural aspect for creating shared understanding of knowledge distribution and knowledge
interpretation. Although it looks like this characteristic would belong to the knowledge level as some
dimensions of it concern the knowledge exchange process, we have identified it to belong mainly with
the structural aspects of knowledge distribution and interpretation. We grade this characteristic as
either high, medium (to some extent), or low.
39
5.2.3 Characteristics at the knowledge level
Finally, at the knowledge level we have identified the following:
Knowledge transformation process. Nonaka and Takeuchi (1995) has identified and defined four
transformation processes wherein knowledge is created, thus leading to either individual
organizational learning: socialization, externalization, combination, and internalization. These refer to
how knowledge moves to and from tacit and explicit form when individuals interact. The adoption of
knowledge is akin to the internalization process and the contribution of knowledge is akin to the
externalization process. However new knowledge is not just created in the previous activities but also
when putting pieces of external knowledge together as to produce e.g. a report of some sort, thus
resulting in a combination process. We use the same categories as Nonaka and Takeuchi (1995) when
classifying the knowledge transformation process that is supported or mediated through the knowledge
distribution process.
Distribution mode. Angus and Patel (1998) identified pushing as an activity related to knowledge
distribution and pulling from information sources as an activity related to knowledge acquisition.
However, we regard the pulling for information sources to be an activity related to the distribution and
we make further distinction about these activities or modes of distribution. Distribution as an activity
can either be active or passive, pulling from sources or pushing knowledge or information on
individuals. Active pulling of knowledge from a source means to know your need and locate and
lookup, search and retrieve, the knowledge matching your needs. Passive pulling means to filter a
stream or flow of knowledge or information according to your explicit needs. Active pushing means to
define a receiver or group of receivers and based on either their explicitly made needs or that you
judge, or perceive that they should be receivers, disseminate relevant knowledge or information to
them.
Knowledge distribution mode could also be regarded in terms of adoption and contribution (Goodman
and Darr 1998). The different means may be discussed in relation to how they reduce the cost
inherited to the knowledge distribution, e.g. when distributing across boundaries of units within an
organization or between organization one often encounter cultural barriers such as language, but also
an increasing degree of alienation to the source.
Acquisition mode. This concept refers to the different ways of acquiring knowledge (c.f. Huber 1991
and Wijnhoven 1998) and can be grouped roughly into vicarious learning, experimental learning, and
transactional learning i.e. buying knowledge or hiring new members to the work force thereby grafting
new knowledge into the organization (Huber 1991).
Substance. From the concept of genre (Yates and Orlikowski 1992) we have adopted substance as to
be a characteristic of the knowledge distribution means. Substance refers to the content of recurrent
distribution means and as such we look on what type of knowledge is distributed. A common way that
knowledge may be distributed as is as a problem-solution pair. Here we have the concept of "best
practice" from Goodman and Darr (1998), and the decision stimulus and organization's response from
Walsh and Ungson (1991). Other may be a scenario describing a problematic situation and its possible
solutions. We may also look on the type of knowledge that is distributed from an epistemological level
and may basically distinguish between know-how and know-why knowledge. Know-how relating to
the gaining of procedural knowledge of how to go about in doing something. Know-why relates to the
understanding of the states of nature and value systems used. The division of knowledge into these
basic categorize has parallels with the way Argyris and Schön (1978) divides organizational learning
into single-loop, relating to know-how, and double-loop learning, relating to know-why. We have not
graded or classified the substance of the knowledge distribution means in advance but wait until we
analyze them.
Purpose. Underlying the knowledge distribution there is a purpose that transcends that of individual or
organizational learning, and increasing the effectiveness of the organization as a whole. The purpose
for knowledge distribution could either be commercial, i.e. we sell knowledge or tie knowledge to
40
products in order to make them more competitive, pedagogical or educational, i.e. distribution is done
in order to educate, give norms or train other members of the organization, service oriented, i.e. giving
support or updating common knowledge resources, or aimed towards idea generating and problem
solving by introducing innovative solutions to different problems, thus expanding the frame of
interpretation of individuals.
41
6 Identified knowledge distribution means
In this chapter we present the identified computer mediated knowledge distribution means and
describe them. Some of these, one may argue, seem to be new and specific to the computer as media
and others may not. Some may seem to be more closely related to media and others more closely
related to the processes of distributing or creating knowledge. One may also argue that some, that
would seem obvious such as e-mail, are missing when they could or should be included and some that
should not have been included at all. Reasons for not including a specific means for distributing
knowledge could be that they on the conceptual level are already covered but through another means,
such as discussion groups and conference systems.
The aim and criterion for identifying different knowledge distribution means has been to identify those
that represent a broad variety of knowledge distribution means, ranging from media related to process
related, and representing unique aspects of the knowledge distribution. The criterion for selecting has
been that the means should have at least some "critical mass" of usage, i.e. have a fairly broad scope of
usage. They should thus not be totally new and alien but be recognized and accepted within the
consciousness of individuals who distribute knowledge. This last criterion was met by investigating
public and popular sites on the Internet supporting professional activities such as development and
research, e.g. Microsoft ® and CNET amongst many others.
The list presented below is not definite. Other distribution means may be included or existing ones
excluded if we are to change the definition or because in the future new means may emerge and old
ones fall out of practice. However, the list is representative of the knowledge distribution means used
today.
6.1 FAQ
FAQ (frequently asked questions) is a way of distributing knowledge to a community of knowledge
workers within a specific area/field. FAQ consists of a typical standardized question that a knowledge
worker may raise, coupled then with a following passage of text describing the solution/answer to the
raised question, i.e. problem-solution. Both the question and the solution are given in free text, in a
narrative form (experiencing some degree of formal structure and language), although using those
technical terms that the knowledge community uses, thereby demanding more than a brief
acquaintance with the knowledge domain.
The type of knowledge is practical know-how knowledge in explicit form and is meant to give
pragmatic solutions to real problems. Often there may be given several different FAQs within a
specific area in order to cover as many different aspects as possible of the area/field. The FAQ given
are not meant to give answer to every possible problem or question that might arise but to give general
guidelines that will give enough knowledge for the reader to further draw his/her own conclusions and
provoke a learning process.
Although the knowledge of the FAQ or of the collection of FAQs may be expressed in different forms
when comparing different ones, they share a couple of features that makes them distinct:
• They are always expressed as a general and common question coupled with an answer.
• They are created by the experts within the field/area and are directed towards the knowledge
worker within the field/area.
• They are looked-up, matching the questions or problems of the individual with the general and
common questions of the FAQ, in search for the answer or solution.
• They are always explicit know-how knowledge and the receiver is meant to internalize the
knowledge, convert it to tacit form through comprehending it.
The FAQs are very common in accompanying products that require knowledge to use, such as
programming and system development environments. Below follows an example of a FAQ, figure 6.1,
from Microsoft ®.
42
Visual Studio 6.0 Service Pack 2
Frequently Asked Questions
Q: If I did not install Service Pack 1, do I need to install it prior to installing Service Pack 2?
A: No. The service packs are cumulative, so SP2 contains SP1. You will get all of the SP1 fixes in SP2.
Q: What products are covered by the Service Pack?
A: The Service Pack contains fixes for the following products:
• Microsoft J++® 6.0 (Professional, and Learning Editions)
• Microsoft Visual C++® 6.0 (Enterprise, Professional, and Standard Editions)
• Microsoft Visual Basic® 6.0 (Enterprise, Professional, and Learning Editions)
The Service Pack includes all of the fixes in Service Pack 1 and specifically resolves an issue that can
cause some third party software to behave unexpectedly or crash after the installation of Visual Studio
6. This service pack also contains the required Visual J++ files needed to comply with the preliminary
ruling issued by Federal District Court Judge Ronald H. Whyte. We recommend that all Visual Studio
6.0 developers install the Service Pack prior to releasing their applications, so that they redistribute
the latest runtime files.
Figure 6.1. Example of a FAQ from Microsoft ®.
6.2 Success or failure stories
A success or failure story is another way to disseminate know-how or experience-based knowledge in
an explicit way, although in narrative form allowing very varied language. A success story is a
narration describing how someone went about in achieving a specific goal. The success story may
therefore serve us as a good example to us when we try to achieve similar goals. This is learning from
the experience of others, the practical know-how knowledge. The language of the case descriptions is
the individuals' own and does not belong to the terminology of the elite expert group within the
knowledge domain. There exists no obvious structure of the expressed knowledge, such as the one of
FAQ. The stories are structured in order to allow browsing through them, but can be indexed and
categorized, according to problem areas/domains, in order to allow for a more effective search and
retrieval.
The opposite of a success story is a failure story. At first one may argue against the usage of these
because who is interested in failures. However, they are of equal importance when conveying practical
know-how. They contain the wisdom of what not to do or describing the circumstances when an,
otherwise successful, action might fail. The failure stories share exactly the same characteristics as the
successful ones.
The following examples, figure 6.2 and figure 6.3, of failure and success stories are collected from the
infamous web-site called "Tom's hardware guide", a virtual paradise for all hardware freaks seeking
advice and wisdom concerning different aspect of PC-hardware. At Tom's there are many examples of
both how-to-do stories and success or failure stories.
These stories are gathered from the vast collection of successful and unsuccessful experiences people
have made when trying to overclock their computers (speed them up by altering the clock-frequency
of the CPU and/or the bus-frequency). At Tom's they have gone to great pains in organizing this huge
collection of wisdom of their readers, allowing each and every one to fill in a form describing first
some common features with a restricted language, thus making it possible to classify and index the
stories according some keywords, and then allowing the usage of the natural language in a
commentary to describe the scenario/problem in a narrative form. This makes the collection easy to
browse and search.
43
CPU
Motherboard
Cooling
Clock Speed
Multiplier
Bus Speed
Comments
Pentium MMX 166 MHz, SL27n/a 2.8 V Stepping: O Ceramic Pin Grid Array
Asus TX97, Award, ver. Not important, Intel Triton TX 430TX Chipset, SDRAM
Brand: Toshiba
Big Heat Sink and Fan,
166 MHz to 290.5 MHz
2.5 x to 4 x
66 to 83 MHz
Voltage: 3.0 V, System crashes after a short while. First of let me say that my CPU is
SL27K, Stepping 3, I think I need to use thermal compound, multiplier of 5.5 to reach
456.6Mhz....FOR ANYONE THAT DOESN'T KNOW ABOUT THIS...The new TX97
have a jumper named BF2 on them...beside Socket 7...IF you have this jumper you
have a multiplier of 5.5...please try it ...thanx
Figure 6.2. Example of a failure story.
CPU
Motherboard
Cooling
Clock Speed
Multiplier
Bus speed
Comments
Pentium MMX 166 MHz, SL000 2.8 V Stepping: C Ceramic Pin Grid Array
Shuttle Hot-565 v1.51, Award, ver. wiq0e, Intel Triton TX 430TX Chipset, SDRAM
Brand: Hyundai 8ns
Standard Heat Sink and Fan,
166 MHz to 290.5 MHz
2.5 x to 3.5 x
66 to 83 MHz
Voltage: 2.9 V, This is SL23X !!! What a great CPU is this! At 2.9v, it's not even warm
with the on-chip heatsink and fan. Nothing could be better, I guess. Completely stable
in every condition on Win95 and NT4. This CPU should be better than most 233mmxs.
There are clock-crippled versions out there, though. I'm waiting for a reliable 100mhz
mo'bo soon.
Figure 6.3. Example of a success story.
6.3 Step by step
Step by step or how-to stories tries to give good examples of how to go about in accomplishing
something. They are often written with the intention of being pedagogical and represent typical knowhow knowledge represented in a strict clear language without lengthy technical explanations and
burdening technical terms that require thorough definitions.
Knowledge distributed is typical know-how and presented in a language structured often from a
pedagogical angle. The structure of the knowledge is made into steps. For every step taken the
knowledge base of the individual is augmented, i.e. an incremental process of learning, and the
previously learned is often exemplified, thus becoming more embedded within the individual.
Typically these kinds of knowledge repositories accompany development products, this in order to get
the individual on the correct track as fast as possible. There is an underlying pedagogical intention of
fostering the individual into a "good tradition", ensuring that all individuals taking part of this
knowledge learns a common way to express themselves and how to think about similar problems, thus
having formed a knowledge community of some sort. Step-by-step or how-to stories are for example
those tutorials or educational programs that are offered by Sun Microsystems through the Java
Developer Connection, and Microsoft®.
6.4 Reviews
Another way to seek or distribute knowledge is through review or evaluations. The reviews are
statements by some alleged expert or authority on a subject within a specific knowledge field and are
typically concerned with some product, book, or service. Individuals seek out these reviews or
44
evaluations because it is just too time-consuming or bone tiring to earn the wisdom through brute
experience. One seeks an overview of something presented in a compact format, presented by an
expert, an individual or an organization, or what we hope is an authority of some sort. If the seekers
are e.g. looking for reviews concerned with a group of products, they want to find some set of criteria
on which these products have been evaluated or compared so that they can form their own opinion.
The knowledge given is know-why type of knowledge since most reviews base their argumentation on
understanding the underlying facts.
The review or evaluations are often located with the "expert in the field". Knowledge seekers locate
these stores and browse through them. This demands that the individual wanting to take part of a
review or evaluation must at least be able to locate these stores and that the content of them are
somewhat known. Knowledge presented has no specific structure except for the topic-centered
structure. The language used is not formal but the technical level of terminology, which belongs to the
"authority", tries to match that of the broad audience.
6.5 Subscriptions
This way of distributing knowledge sets itself apart from the other ones. To subscribe to something
means that you have registered, at the source of the information or the knowledge distribution, the
needs you have. Typically we think of e-mail subscriptions in form of mailing lists that we subscribe
to. However equally common is it to register yourself with a vendor whenever you bought something
and receive upgrade information or other product information.
Here you have to let others take part of your needs and let them play the active part of supplying
knowledge to you whenever something that matches your needs can be found. The mode of
knowledge acquisition is here a typical pushing. Knowledge is pushed on someone by another one
who knows the first one's needs. The process here is a directed distribution process towards selected
receivers, based on their explicit made and shared needs. This is opposed to the other distributing
means where the individuals must take an active part, i.e. the individuals must of course know his/her
own needs and make the effort of looking up the knowledge or pulling it from the source.
6.6 Bulletin boards
These are nothing more than a public location where an individual may go and post something that
he/she thinks others should take part of. On the other hand the individual can visit the location and
browse through what others have posted there to see if something interesting shows up. Messages can
be read many times as there often is no news control. The search for knowledge can either be a search
for something specific according to one's needs or just an idle browsing. Bulletin boards are not totally
unstructured but may be divided into different categories, thus making browsing more feasible.
A public bulletin board would be too unrestricted to be really useful if one searches for something
specific. There would be just too much non-relevant stuff. However, the restricted use of one within a
specific knowledge community ensures that the members visiting the board know roughly what its
content will be.
Weather to consider bulletin boards as a medium or means for other knowledge distribution means or
as a knowledge distribution means on its own might be difficult to discern. On one hand a bulletin
board offers a public "meeting" place, enabling people to exchange knowledge or information of any
kind. As such it may act as a medium or means for the other types of distribution means. On the other
hand bulletin boards may be regarded as distribution means on its own since members of a specific
knowledge community utilizing the board both for disseminating and retrieving knowledge comes to
create a common knowledge store, albeit a bit unstructured.
6.7 Discussion groups
As a knowledge distribution means this one distinguishes itself from the previous ones by introducing
the dialog as a communication model for the knowledge exchange. Discussion groups are typically
45
structured to be concerned with a specific problem area. The individuals taking part in the discussion
group usually raise different issues concerning the problem and reply to each other (they may even
make replies to the replies and in some sense form sub-issues). Messages posted in a discussion group
are made accessible to all the participants.
Different aspects of an issue together with the replies constitute the collective knowledge of the
participants regarding a specific issue. The conversation, which is a series of messages that are directly
or indirectly replies to each other, form a tree-structure or a web if a reply is allowed to address several
issues. Usually the discussion group is intended for reaching a common understanding or debating an
issue, equalizing the individuals anxiety and expectations, of that specific issue.
The discussion groups are not public in the sense that bulletin boards are. The discussion groups can
either be open, where participants can enter on whim, or private in the sense that the individuals have
to be invited to the discussion group in order to take part of it, i.e. there is a control on who may
participate. Due to the inherent structure of issues and replies, following an ongoing discussion is
rather easy even if a participant is invited in the middle of the discussion. There is usually a storage
facility coupled with the discussion group that enables new participants to follow up on what has been
discussed before. Discussion groups usually have some sort of news control mechanism that keeps
track of where the participant is in the conversation, which messages are read by whom, and which the
participant has yet to read. Examples of discussion groups are news groups, conference systems, and
similar.
6.8 Hypertext documents and traditional manuals
In contrast to the above-presented means of distributing knowledge there stand the hypertext
documents and traditional manuals. Hypertext in its basic form is actually not a means to distribute
knowledge but rather a means to organize knowledge as opposed to traditional linear representations
such as manuals. Manuals are in this context all forms of linear books, such as technical manuals, class
books, encyclopedias, and etc., whatever media they may be presented on.
Hypertext is an approach to information management in which information/knowledge is stored in a
network of nodes connected by links. The combination of nodes, links, and supporting indexes for any
particular topic is often called a hypertext document. The difference between a hypertext document
and a conventional paper lies in the relationship between the documents logical and physical structure.
In most paper document the physical and logical structures are very closely related and the reader
proceeds through it in the order prescribed by the author. However, some paper documents such as
encyclopedias and other references works operate differently. They assume that the user will identify a
topic of interest, read about it, and then use cross-reference to find other information of interest related
to it. Hypertext documents were designed to function in this way (Alter 1996). The point of the
hypertext documents is that they were meant to be represented electronically thereby borrowing the
power of the machine when it comes to browsing through the document, thus providing greater
flexibility than the traditional paper encyclopedias.
One may well argue that hypertext documents should be regarded as an emerging knowledge
distribution means when considering the explosion of usage of Internet browser of today. However
one can just as easily argue that hypertext as a phenomenon, even though sharing many characteristics
with the other above mentioned distribution means, is nothing else but a "cross-referencing
encyclopedia" with electronic support for browsing. Then, a similar argumentation could also be used
about the other above described means.
46
7 Characterization and analysis of the distribution means
In this chapter we will characterize the different knowledge distribution means according to the
different levels of interpreting the knowledge system, using the concepts or features we have identified
as relevant and of interest in earlier chapters. We present three different views of the knowledge
distribution means characterized according the identified characteristics of chapter 5 starting with the
technical level of interpretation, then with the representational level, and finally with the knowledge
level. We will also turn our attention towards the model for knowledge representation of Walsh and
Ungson (1991) that was pointed upon in chapter 1. Walsh and Ungson (1991) argue that effective
means for knowledge representation should support this model and in this respect we continue with an
analysis of how the emerging knowledge distribution means support their model.
7.1 Characterization at the technical level
At the technical level we have characterized the knowledge distribution means as shown below in
figure 7.1.
Characteristics Indexability
Means
Yes
FAQ
Cues
Language variety
Low
Medium (strict
structure)
Rich
Success/failure
stories
Reviews
Yes
Low (linguistic features only)
Yes
Low
Step by step
Yes
Low
Subscriptions
N/A
Low
Bulletin boards Yes
Discussion
groups
Yes
Medium (linguistic features
and number of responses)
Medium (linguistic features,
number of responses, and
rapidity of response
Feedback and
interactivity
Low
Low
Medium (strict
language)
Medium (strict
language and structure)
Medium to rich (some
strict language)
Rich
Low
Low to
medium
Medium
Rich
High
Low
Figure 7.1. Characteristics related to the technical level.
At this level of interpretation we have found that all the different knowledge distribution means are
indexable, thus allowing for both efficient search and retrieval of knowledge from computer supported
retention stores. The only one that is not searchable and retrievable, due to the very nature of how the
knowledge is distributed, is the subscriptions and therefore indexability is not applicable here.
If we look at the capacity of cues, number of different concurrent signals that are transmitted through
the media or distribution means, we note that for all of them the capacity is low. This is due to the fact
that the content of the distribution means is mostly textual and the distribution means doesn't really
allow for anything more, although some linguistic features may be used (such as capitalization, use of
symbols such as smiley, and etc.). Such linguistic features are common with the success/failure stories
due the nature of their content. However success/failure stories are still considered to be low in
capacity for transmitting cues. The distribution means that deviate from low capacity grading are the
bulletin boards and discussion groups. Here, besides linguistic features other cues may be transmitted
as well that help in establishing a context for the knowledge. Number of responses a message may get
is one cue that both means share, which may yield a sense of importance, relevance, interest, or
popularity that is associated with the message. Rapidity of response to a message, which only
discussion groups exhibit due to the nature of being "online", is yet another cue that may help further
in establishing a broader context. Therefore these two distribution means are graded as medium.
Richer media for distribution are able to cope with a wider language variety and hence are better for
conveying meaning. A rich language is better suited to convey broader set of ideas and concepts than
mere numbers allow for. All of the distribution means could therefore be considered rich in this aspect
47
were it not for the fact that some of them exhibit features that restrict them and thus make them less
rich. The FAQ exhibits a strict structure that must be followed, that of question-answers, although this
in a sense creates a common view of interpreting the knowledge. The reviews exhibit constraints of
language dictating that it must be clear and broad so that it may reach a broad audience. This makes
the usage of a rich language full of technical terms and jargon not applicable here and thus lessens its
capacity for language variety. The step-by-step, due to its pedagogical nature and its structure for
building knowledge is also restricted in the same sense as both the FAQ and the reviews are. Finally,
the subscriptions are also considered to be somewhat medium to rich in its capacity due to the degree
of formality (clearness or strictness) of the language. This imposes further restrictions, as with
reviews, on the variety of the language used.
Both degree of or capacity for feedback and interactivity, that is associated with the knowledge
distribution means is here considered rather low for the FAQ, success/failure stories, reviews, and
step-by-step. This is due to that the nature of these knowledge distribution means doesn't allow very
much for either feedback or interactivity in interpreting the knowledge. After the knowledge has been
distributed to an individual he/she sits rather in solitude without the possibility of interacting or
responding with the contributor of the knowledge. However for subscriptions, bulletin boards, and
discussion groups the light seems to shining a little different at the end of the tunnel. More capabilities
are allowed for here in relation to feedback and interactivity of the distribution means. These are
indeed increasing as is reflected in the grading. You may opt to quit the subscription, you may answer
or as the case is post a reply to the bulletin board, and you may in discussion groups interact and give
feedback almost, but not quite, at a conversational level such as in a face-to-face situation.
The capacity for bindingness of the distribution means is another characteristic that was identified in
chapter 5 as applicable at the technical level. However, this characteristic is not shown in figure 7.1
since all of the distribution means were found to have the same rating. The capacity for bindingness is
considered to be high since the distribution is computer-supported and therefore allows for easy
storage of the knowledge in more permanent form. The capacity is considered high even though the
binding within time and space is not necessarily high for all of the distribution means, expect for
discussion groups where some degree of bindingness of the distribution means within time and space
exists.
7.2 Characterization at the representational level
At the representational level we have characterized the knowledge distribution means as shown below
in figure 7.2.
Characteristics Language
Means
Technical and
FAQ
strict
Jargon and domain
specific
Pedagogical and
broad and general
Pedagogical and
Step by step
strict
Technical and
Subscriptions
domain specific
Bulletin boards Any
Jargon, expert, and
Discussion
domain specific
groups
Success/failure
stories
Reviews
Structure
Answer-question
Stimulus-response
Problem-solution
Scenario or problem
centered
Topic centered and
criteria oriented
Topic centered and
incremental
Topic centered
Topic centered
Topic centered in a
web or hierarchical
Communication Degree of interactivity
context
or feedback
One-to-many
None
One-to-many
None
One-to-many
None
One-to-many
None
One-to-many
Can quit subscription
Many-to-many
Few-to-few
May post response
Dialog process
Figure 7.2. Characteristics related to the representational level.
From the concept of genre (Yates and Orlikowski 1992) we derived language and structure as
important for the representational level of knowledge distribution means in yielding shared
48
understanding and enhancing learning. If we look at the language we note that a variety of language
types seem to be used throughout the different distribution means. For the FAQ we see that they
exhibit a rather technical oriented language with strict formulations. The nature of FAQ is such that it
requires the structuration of the language to be in the form of a question-answer. It is also domain
specific requiring of the user that he/she has prior knowledge of the terminology used. This is also the
case of success/failure stories, being domain specific, but the structural restrictions imposed on the
language are here forgone. The narrative language form of success/failure stories allows for usage of
jargon and thus enables a richer and more varied language for conveying meaning. The purpose of the
reviews is educational, in an evaluative aspect, and therefore they exhibit a more pedagogical
language. Since reviews often address a larger and broader audience the language is more broad and
common. Step-by-step are also pedagogical in their nature but they do not address such a large
audience as reviews do. They aim to educate and serve as norm giving and thus exhibit a stricter
language. Subscription often require of its audience prior knowledge of the domain that the specific
subscription is intend for and therefore the language tend to be more technical. Bulletin boards have
no structural restrictions on language, any type of language is usable due to the broadness of the
audience of this knowledge distribution means, and therefore does not seem to exhibit a "genre" of
knowledge distribution from the language perspective. Discussion groups are usually not that open as
bulletin boards are and therefore they tend to be more domain-specific with the "group" using it.
Although there are no restrictions in the use of jargon, technical or expert oriented language may occur
within the knowledge communities formed by the groups.
The structural aspect of genre concerns recurrent structures for representing the knowledge. If we look
on the FAQ we see that the structure is quite clear. A FAQ is always in the form of a question-answer.
This is a clear structure similar to a stimulus-response or a problem-solution pair. This structure is
indeed a recurrent one for this type of knowledge distribution means and as such has a clear meaning
to everyone that uses it. For the success/failure stories there exists no such distinct structure, as with
the FAQ, but they are typically scenario or problem situation oriented. They usually describe how or
what someone did in achieving a solution for a problem situation and as such may be considered
recurrent. The reviews usually have a clear topic that is concerned with an evaluation of a specific
item or issue according to some set of criteria. Therefore we may say that they have a recurrent
structure that is topic centered and criteria oriented. Step-by-step have a typically topic centered
incremental structure. They move step by step in building and expanding the knowledge of a specific
area, moving from simple examples to increasingly more difficult ones, incorporating the previous
knowledge. Subscriptions are also topic centered since a user subscribe to a specific knowledge
community or source which have a clear focus, but otherwise exhibit no specific structure. The same
holds true for discussions groups but here there is a clear structure. The way members of a discussion
group may respond to each other and post message, i.e. interact with each other, may form a web or
hierarchical structure of ongoing discussions or conversation around specific topics.
The communication context of the knowledge distribution means dictates many of the structural
aspects of how shared understanding may be enhanced. Here we primarily find four types that are
applicable to us: one-to-one (face-to-face or two-way communication), on-to-many (the mass median
one-way distribution), few-to-few (group centered two-way or "few-way" distribution), and many-tomany. If we look on the FAQ, the success/failure stories, the reviews, step-by-step, and subscriptions
we see that the distribution is directed as a mass median distribution not allowing for either feedback
or interactivity (or is severely crippled). This way of distributing, one contributor distributing to an
audience of adopters in a one-way communicational context is natural for these distribution means and
not accidental, but it has impacts on the structural aspects of enhancing shared understanding. The
knowledge distribution means, here, must hold all the information, concerning language and structure,
for the interpretation to be successful. If we look on the bulletin boards we have a completely different
situation and context of distributing knowledge. Due to the broadness of the audience, in terms of
numbers and variety of participating members, we have a situation of many-to-many. The nature of
bulletin boards is such that posted message may yield several responses and a response may address
several messages. The situation is the same with discussion groups, however they are usually more
restricted regarding numbers of members participating and the message and responses are structured in
49
webs or hierarchies. Discussion groups are also more topic-oriented, as has previously been shown,
which also restricts the numbers of members to a degree (only the interested in the topic discussed will
participate). Therefore we have, as Tubbs and Moss (1994) defined a group communication context
where more of the group dynamical aspects come into play in the knowledge distribution process. This
is a typical few-to-few distribution encouraging high degree of feedback and interactivity.
If we look on the degree of interactivity of feedback within the communication process or distribution
process, as a structural aspect of enhancing shared understanding, we note that none except
subscriptions, bulletin boards, and discussion groups actually exhibit any such capacity. With
subscriptions you always have the ability to quit the subscription, even though this may seem to be a
rather crude way of implementing feedback or interactivity. On the bulletin boards you may post
responses and message of your own. In discussion groups structural capabilities for having a dialog, of
some sort, are implemented. The degree of interactivity and feedback, within the distribution process,
appears to be more coupled with the media aspect of the distribution means than the structural ones.
7.3 Characterization at the knowledge level
Finally, at the knowledge level we have characterized the knowledge distribution means as shown
below in figure 7.3.
Characteristics Knowledge
Means
transformation
Internalization
FAQ
Success/failure
stories
Internalizationexternalization
Reviews
Internalization
Step by step
Internalization
Subscriptions
Internalization
Distribution
mode
Active pulling
(adopting)
Acquisition
mode
Vicarious
Active pulling
(adopting or
contributing)
Active pulling
(adopting)
Vicarious or
experimental
Active pulling
(adopting)
Pushing
Experimental
Vicarious
Vicarious
Bulletin boards Internalizationexternalization
Active pulling
(adopting)
Vicarious
Internalizationexternalizationsocialization
Active pulling
(adopting)
Vicarious
Discussion
groups
Substance
Purpose
Know-how or
Service and problem solving
know-why
Best practice
Know-how
Service and problem solving
Problem scenario or idea generating
Know-why
Evaluation
oriented
Know-how
Educational
Know-how or
know-why
Any type
Know-how or
know-why
Any type
Know-how or
know-why
Any type
Service or commercial
Pedagogical or educational
Service
Service
Problem-solving, idea
generating, pedagogical or
educational
Figure 7.3. Characteristics related to the knowledge level.
When we look on knowledge transformation processes for classifying the different processes of
knowledge distribution we note that all of them can be regarded as internalization since they all
involve a transformation from explicit knowledge to a tacit form. The knowledge distributed is done in
an explicit form where the learning takes place when the individual has interpreted and understood the
knowledge, i.e. internalized it or as Huber (1991) states have had his/her range of possible behavior
altered. We note that there are some cases of externalization here as well. The nature of success/failure
stories is such that the individuals that use these could be encouraged to expand the knowledge base of
these stories by contributing with their own experiences, which is the case with "Tom's hardware
guide" where this process has been given a rather elaborated structured support. Externalization
processes may also be found in both bulletin boards and discussion groups where explicit message are
formulated, either to serve as reference for response or as a response. The discussion groups could also
be seen as a case of the socialization process since individuals attending these groups may be
socialized into a good communicative behavior, something that is typical for interacting groups.
50
The distribution mode concerns how the knowledge is distributed. We have here distinguished
between pushing, i.e. making specific knowledge available to a specified audience, active pulling, i.e.
intentional search, retrieval and matching of knowledge from source according to ones needs, and
passive pulling, that is filtering a flow of knowledge. Here we the found that the distribution mode of
active pulling is represented through all the different knowledge distribution means. The only
exception is the subscriptions where the knowledge is distributed through pushing. When an individual
has subscribed to a source, knowledge or information will at a regular basis be pushed on him/her.
Subscribing here means that an individual has identified a need or interest in the knowledge contained
at that source and this, then, serves as the basis for the distribution. Related to the distribution modes is
the process of adoption, as defined by Goodman and Darr (1998), either through the active pulling or
the pushing mode. In the case of success/failure stories we also find the process of contribution
(Goodman and Darr 1998), which has been shown before. Contribution may here be seen as a
distribution process, though we do not actually push anything on anyone. However the process of
contribution makes new knowledge available to others for adoption and may be regarded as an indirect
distribution process.
The acquisition mode concerns how individuals acquire knowledge. The FAQ, reviews, subscriptions,
bulletin boards, and discussion groups are all clear cases of vicarious learning, i.e. learning from the
experience of others through the process of adoption. There is actually not that many different forms
of acquiring knowledge that are related to our identified distribution means. However, with the
success/failure stories we have a case of both vicarious learning, quite obviously, and experimental
learning. Individuals using these distribution means for acquiring knowledge are encouraged to try out
the "recipes" given in the problem situation scenarios. Step-by-step also encourages the experimental
mode of learning. This seems to be the very nature of this type of knowledge distribution, learning by
doing in a pedagogical form.
From the concept of genre (Yates and Orlikowski 1992) we also derive the substance pertaining to the
content of the knowledge distribution means. The substance can be viewed from an epistemological
dimension and here we distinguish between know-how and know-why type of knowledge. The FAQ
may contain both know-how, the procedural aspects of reaching a solution such as the "best practice"
of Goodman and Darr (1998), and know-why, reasons for why the "state of nature" is as it is. The
success/failure, containing problem scenarios, is more oriented towards practical know-how derived
from trials and errors. Reviews, on the other hand, are oriented towards explaining "the state of nature"
through their more evaluative criteria based content. Step-by-steps are primarily aimed at giving
practical know-how, e.g. norms or prescriptions of what to do in order to achieve a specific goal, by
following guidelines or good examples. However step-by-steps may yield know-why knowledge in
elaborating on the underlying reasons of specific norms or prescriptions. Subscriptions, bulletin
boards, and discussion groups may contain any type of knowledge. They do not exhibit any specific
constraints in this regard. Their purpose is that of spreading knowledge, as a service, to those who
subscribe to the specific knowledge or information source, visit the bulletin board, or participate in a
specific discussion group.
The underlying purpose of the knowledge distribution means that transcends that of efficient
organizational learning is quite varied as observed among the different knowledge distribution means.
The FAQ is primarily being viewed a service that have a problem solving purpose. Individuals use
FAQs in order to find possible solutions or answer to questions raised in their own domains. This is
clearly reflected in the form and substance of the knowledge distributed as a question-answer. The
success/failure stories can also be seen as serving as a problem-solver or idea-generator due to both the
substance and form of the knowledge. The success/failure stories provide for quite innovative
solutions to a range of different problems situations described in a rich narrative form. Reviews are
regarded as having a service purpose, same as the FAQ or success/failure stories, where products or
artifacts are critically evaluated or a commercial purpose for promoting the same products and artifacts
for sale. The purpose of step-by-step is clearly educational and pedagogical, by providing a structure
for incremental learning using a pedagogical language as seen in both the substance and the form. The
purpose of subscriptions is usually service oriented. Individuals choose to subscribe to different
51
sources of knowledge or information order to get regularly "updated". Within a knowledge community
this could ensure that everyone, thus interested or concerned with the same issues, subscribing to a
specific source, get the same knowledge or information, serving as a platform for creating shared
understanding. Bulletin boards are usually regarded as service oriented. Individuals visiting bulletin
boards usually do so to utilize it as a service for posting "messages" of any type. The purpose of
discussion groups could actually be any of the before mentioned ones, except service and commercial
because rarely is discussing in a group regarded as a service and even less is there a commercial
motive behind it unless less we participate in a computer supported sales-meeting.
7.4 A model for knowledge representation
The model for knowledge representation of Walsh and Ungson (1991) captures and represents
decision information, i.e. the cause or the stimulus that triggers a decision (problem), and the effect or
the response of the organization (solution), as well as an analytical understanding of the relationship
between the cause and effect. Walsh and Ungson (1991) point out that the rational for the existence of
the relationship between the cause and effect may only be understood within the context of an
organization's history. The model structures knowledge in the form of a problem, matched with a
solution, and a context for the problem-solution relationship giving us an explanation to why a specific
solution is given to a specific problem. In this sense the model encompasses both know-how
knowledge, in the form of a solution to a problem, and know-why knowledge, in the form of the
analytical understanding of the problem-solution relationship. We may look upon the model both from
a structural as well as process related aspect. From a structural aspect the model is supported by what
structural features or mechanisms for representation that the knowledge distribution means offer. From
the process related aspect we see that their model is supported through the very process of distributing
the knowledge. We will first take a look at how the different knowledge distribution means from a
structural aspect support this model and then follow with the process-related aspect.
A knowledge distribution means that would appear to be a natural candidate for supporting the
structural aspects of knowledge representation is the FAQ, supporting the knowledge structure in the
form of a question-answer. This is very similar to the model of Ungson and Walsh (1991). However,
the FAQ doesn't generally support the "why" unless, of course, it is embedded in the answer since
rarely does the question reveal the rational for its own existence. The FAQ is the knowledge
distribution means that comes closest to the model, in terms of structure, but the purpose of the FAQ
seems not to aim for capturing the "why". Though the model of Walsh and Ungson (1991) aims to
capture both the practical know-how aspects as well as the analytical understanding in the form of
know-why knowledge, the FAQ, however, doesn't explicitly support the latter one. The FAQ support
know-how knowledge in the sense that it gives practical solutions to practical problems. There is no
analytical understanding embedded yielding know-why knowledge. Success/failure stories allow for a
scenario-centered structure in describing in a narrative form the reason for success or failure in
achieving a specific goal. The scenario could capture the knowledge pertaining to the "why" but
doesn't exhibit a specific structure for it. The other knowledge distribution means seem to support the
structural aspects in an even lesser degree. One can only note that they seem to be used for a different
purpose, in a structural perspective, than that of the model of Walsh and Ungson (1991).
The distribution means that comes the closest to the model, in terms of process, is the discussion
group. The nature of knowledge exchange in a discussion group is such that, due to the high degree of
interactivity and feedback, the "why" may evolve through the discussion taking place. Discussion
groups allow for an analytical understanding through feedback and interactivity. Here both the knowhow and know-why may be shared between participants in a discussion. Discussion groups support in
this sense the concept of a shared "arena" that Nonaka and Takeuchi (1995) has talked about. From a
process aspect we see that discussion groups as a means for distributing knowledge is the one that
comes closest in supporting the model of Walsh and Ungson (1991) for knowledge representation as a
participant may pose a question and receive an answer from another participant, and then through a
dialog arrive at an analytical understanding of the rationale behind the question-answer. None of the
other knowledge distribution means seem to give such a support for the model, from a process related
perspective, as the discussion group does. Even though the bulletin boards seems to share many
52
features or characteristics with the discussion groups, their "arenas" for knowledge exchange or
distribution do not exhibit any mechanisms for supporting a dialog-based process as the discussion
groups do.
What we see here is that only one knowledge distribution means, the discussion group, seems to
support the model of knowledge representation of Walsh and Ungson (1991), and then from a process
related aspect. The others only support it in part, both from structural and process related aspects.
Even though the FAQ would seem to be a natural candidate for supporting the model it lacks any
structural features or mechanisms for representing the "why". The FAQ, as a means for distributing
knowledge, must be developed, structurally, as to also accommodate for the know-why. Not until then
may we state that their model has been given full support both in term of structure and process for
effective knowledge representation, which Walsh and Ungson (1991) have called for.
53
8 Conclusions, future research and final observations
8.1 Conclusions
The following conclusions are drawn from the results of this work:
Knowledge distribution is not an isolated phenomenon and need to be framed within a wider context
such as organizational learning.
We base this conclusion on the fact that we have found that it is difficult to study knowledge
distribution means in isolation without touching on the other aspects or constructs of organizational
learning, such as the knowledge acquisition, knowledge interpretation, and organizational memory. It
may be due to that these are all integrated aspects of the organizational learning as Huber (1991) has
defined it. Issues concerning the interpretation are intertwined with the distribution, which is reflected
through our characterization. The same holds true for the other constructs. These do not occur as
separate activities or entities within the frame of organizational learning. Therefore many of the
characteristics that are identified and used also concern these as well. However we do not regard this
to be a problem or limit in our study. Quite the opposite, it contributes to a better understanding and
richer picture of the different knowledge distribution means. It is no secret that individuals distribute
knowledge in order to promote learning and as we have seen organizational learning involves many
integrated activities and aspects. Therefore it is only natural that our characterization has come to
reflect these as well. Knowledge distribution is not an isolated phenomenon since it touches too much
on other related phenomena such as knowledge interpretation, retention (organizational memory), and
acquisition.
The importance of an organizational memory becomes apparent when one notes that many of the
presented knowledge distribution means deals with a knowledge acquisition process directed towards
learning from the experience of others, i.e. vicarious learning (Chew, Leonard-Barton, and Bohn
1991). It seems to be far easier to learn from the mistakes or experience of others than to reinvent the
wheel, even though the danger of this type of knowledge acquisition is that knowledge may lack
universal validity (Wijnhoven 1998). If this type of learning is to be effective then the knowledge, the
experiences of others, must be retained in some form of organizational memory if it is to be made
available to other members of the organization. Thus we once more must emphasize and draw the
conclusion that knowledge distribution as an activity cannot be studied separately but must be studied
in a wider context such as the organizational learning as Huber (1991) has discussed.
The concept of genre is not only fruitful but an important concept to explore when studying, analyzing,
or developing means for knowledge distribution.
This conclusion is based on that the different knowledge distribution means that we have identified
seem to represent each a category of specific types of distribution means. However, the comparative
analysis has shown us that even though we see distinct features of both substance and form that suits
generic activities or situations of knowledge distribution we cannot say that they are solely designed
for a specific purpose alone. Each of the identified distribution means does not represent a specific
category of distribution means that would be best for a specific type of knowledge distribution. We
may only state that it seems that some are more appropriate for a specific task than others, e.g.
discussion groups seems more suited for knowledge exchange occurring in teams and FAQ seems
more suited for distribution of knowledge related to problem-solution situations. However, the
characterization has shown us that we may talk about genre of recurrent knowledge distribution
means. Here the concept of genre (Yates and Orlikowski 1992) suits us very well and we may talk
about recurrent situations of knowledge distribution yielding what could be considered as a specific
genre of knowledge distribution that have the same substance and form.
Through the characterization we have seen that the different knowledge distribution means seem to
exhibit properties that are similar to the properties of a genre. There seems to be possibilities for
54
identifying recurrent distribution means and to talk about genres of organizational knowledge
distribution. The FAQ is one such instance of a genre of knowledge distribution. Whatever the
actually content of the FAQ is they all seem to be exhibiting the same form and substance, i.e. the
structure and content of question-answer or problem-solution pairs, and therefore we may say that the
FAQ is indeed one genre of knowledge distribution. Clearly related to the FAQ is the "best practices"
of Goodman and Darr (1998) since they exhibit the same form and substance. However both the FAQ
and "best practices" may also be regarded as sub-genres (Yates and Orlikowski 1992) that inherits
both the substance and form from what we call a "problem-solution" genre. Thus the concept of genre
(Yates and Orlikowski 1991) is not only fruitful but an important concept to explore when studying,
analyzing, or developing means for knowledge distribution. The concept of genres of knowledge
distribution serves as a foundation for understanding and establishing shared frames of interpretations.
The concept of genre is indeed interesting since it creates possibilities for a shared context wherein the
adoption and contribution processes of knowledge (Goodman and Darr 1998) are facilitated by giving
common platforms for creating understanding, thus facilitating learning. Interesting to explore here is
whether we may talk about genres of organizational learning. Is it possible to identify recurrent
distribution means within each of the other constructs of organizational learning so that we may talk
about genres of knowledge acquisition, knowledge interpretation, and organizational memory? We
will come back to this later when discussing future research directions.
The investigated distribution means only partly support a model for knowledge representation that
captures both the problem-solution as well as an understanding of their relationship. In this respect
existing distribution means must be enhanced or new ones developed if we wish to endorse such a
representational model.
This conclusion relates to the analysis we made concerning the model for knowledge representation of
Walsh and Ungson (1991) in chapter 7.4. We stated that effective means for knowledge distribution
should support this model and we analyzed the knowledge distribution means according to how they
supported this model. The conclusion we draw here is that the model for knowledge representation is
still rather unsupported (even though discussion groups may support it from a process-related aspect).
The FAQ, even though it might be considered a natural candidate for supporting the model as a
knowledge distribution means, doesn't capture the know-why aspects of the knowledge. However in
the light of the problem-solution model of Goodman and Darr (1998) the FAQ is adequate since it
clearly captures and represents the know-how aspects of knowledge. If we are to take the model of
Walsh and Ungson (1991) seriously we must develop better support for it, not only for the distribution
of knowledge but also for its acquisition and retention.
8.2 Future research
If we look towards the future and turn our attention on what to do next we find, based on the results
from our investigation, one specific issue that is not only worth exploring but demanding it, and that is
the possibilities for identifying genres of organizational learning.
Is it possible to identify genres of organizational learning, similar to the concept of genres of
organizational communication that Yates and Orlikowski (1991) have discussed? May we, based on
the construct of organizational learning that Huber (1991) has identified, talk about genres of
knowledge acquisition, knowledge interpretation, and organizational memory (we have already
touched on the concept genre of knowledge distribution)? Perhaps it is more fruitful to talk about
organizational learning as a whole and not divide it up into its specific constructs, especially since we
have found in our study that it is quite hard to discuss knowledge distribution in isolation. If we
explore the concept of genre of organizational learning what could we possibly hope to achieve and is
it fruitful?
The problem within organizations is that much knowledge or experience of individuals, or groups of
individuals, is not available for others within the organization or between organizations. The wheel, so
to speak, may well be invented not only twice but also thrice or more times, mistakes being repeated
55
over and over. Solutions to critical problems and good practices may live and die in isolation that
would otherwise effectively enhance the behavior of the organization if they ever met. Also
individuals come and go, maybe as part of natural personnel turnover or due to other circumstances,
taking with them a lot of experiences and tacit knowledge that the next fellow must re-conquer. There
seems to be a need for finding ways or means of making it easier for people to express what they know
and make it available to others.
Effective leveraging of knowledge is vital for bringing about innovations and making organizations
more competitive, which Nonaka and Takeuchi (1995) have discussed. Now this seems to create, in a
sense, a view of knowledge as a commodity that needs to be managed effectively like concrete
artifacts and other products within the organization. One mustn't forget that knowledge is a property of
the mind of the individuals, interacting in a social system, expressed through their behavior. The point
here is that good knowledge management is about creating "arenas", as Nonaka (1994) talks about,
where individuals may "meet" and exchange knowledge. These arenas for interaction do not only
imply face-to-face situations but in broader sense any common grounds for shared understanding and
interpretation that enhances the learning, creating incitements for both contribution and adoption, as
Goodman and Darr (1998) talk about. Here the concept of genre of organizational learning seems
fruitful since it focuses on recurrent situations of organizational learning, thus creating a common
platform or frame of references for individuals, within different units, to efficiently exchange
knowledge and learn from each other, overcoming such barriers as anonymity, language, culture, etc.
(Goodman and Darr 1998).
In sum, to follow up on this study we would go on investigating or exploring the concept of genres of
organizational learning. The first step here would be to define what these genres are and how the
concept might support the learning of an organization as to effectively change the range of an
organization's behavior in making it more competitive or efficient. In the second step to investigate or
explore the means used for supporting or making these genres possible through a descriptive study.
Then follow up with a discussion or analysis of what these means should support as to be most
effective, in this respect the constructs of organizational learning (Huber 1991) could serve as one
ideal model.
8.3 Final observations
The first observation we make regards the characterization of the different distribution means and how
this result may be used. Since we have chosen to explore the knowledge distribution means through a
descriptive study in which we characterize them, the results from this characterization may be used in
an evaluation and selection process. The characteristics we have identified are possible to use as
criteria for choosing the best means for distributing knowledge under certain conditions or for a
specific purpose. In this respect one must first make clear how one values the different characteristics
and the different views in order to make this evaluation possible and meaningful. In this study none of
the identified characteristics or derived views has been given a special value as to say that they are
more important than others are and thus should receive more attention in an evaluation process. The
framework in itself may here also be used for characterizing other or similar distribution means, which
we have not identified, and thus function as a generic model for evaluating and choosing.
The theoretical framework that we have constructed, based on the levels of interpretation and the
constructs of organizational learning, is far from complete. The concepts and theories that we have
identified within that framework are not the only ones that are relevant. Here much more theoretical
exploration is needed than we have done to fill the gaps that may exist. It has not been the primary aim
in this study to make a complete and final model of all the relevant issues and theories, but to serve as
a basis for identifying those issues that help us in identifying relevant characteristics. The framework
must be seen as an instrument or tool for generating debate and exploration on issues that are relevant
to different aspects of organizational learning.
The three views that are derived from the framework in relation to the construct of knowledge
distribution show us different levels from which we may view, compare, and interpret the distribution
56
means, ranging from technical issues to structural and process related ones. These views are, in the
same sense as the framework, also instruments for exploration and debate. Other characteristics may
be identified and analyzed, thus creating other views of the different knowledge distribution means. It
is of course possible to create the same kind of three-level view of the other constructs thereby
arriving at a very detailed and rich picture for analysis of organizational learning. In this respect the
framework serves as a generic model together with the possible three-level views that may be used in
analyzing organizational learning as a whole, both from a top-down (or a holistic) angle and from a
bottom-up angle, through the different constructs.
All of the different levels from which we have made the comparative analysis have shown us that
many of the issues are similar to those of communication and information exchange, such as the ones
relating to the media at the technical level. Here many of the issues are not unique to the knowledge
distribution alone but also for e.g. the acquisition of knowledge (actually any form of transmission
dependent on technology). Analyzing the distribution means at this level reveals nothing new except
that many of the considerations that are related to the technological platform of the organization for
distributing information are valid here as well. Issues of technology related to richness of media may
be a bit more emphasized since the knowledge distribution seems to require richer media allowing for
shared understanding, as Huber (1991) has discussed. We have not explored issues concerning the
technology for organizational memory since we have remained mostly focused on the distribution, but
these issues are still important to throw light on when developing computer-mediated support.
At the representational level, dealing with issues concerning representation of knowledge through
structure and language, we find that all of the knowledge distribution means favor some form of rich
language, be it technical, jargon or whatever. Type of language used for representing the knowledge
depends on the purpose of the distribution that we find at the knowledge level. The same is also true
for the structure related to how the knowledge is represented. Here we find different structures,
ranging from stimulus-response to "conversational". These too are dependent on the purpose of the
distribution and what type of knowledge that is distributed. Knowledge distribution means that don't
allow for much interactivity and feedback, between contributor and adopter, must use a language and
structure that facilitates the interpretation and the creation of a shared understanding. It seems that
one-way transmitted knowledge must bring with it the very means for interpreting it based on its
content alone. For knowledge distribution means allowing for higher degree of interactivity, such as
discussion groups, this need is not that apparent and both language and structure of the knowledge
may be more relaxed. However, here we exchange structure of the knowledge with structure for the
exchange of it.
Structural issues also concern the context of the distribution and structural features that allows for
interactivity and feedback in the distribution process. Here we saw that all of the distribution means
are mostly related to mass median distribution. Perhaps this is the most natural since the distribution of
knowledge is based on impersonal sources through one-way transmissions. Language and structure is
not unique to knowledge alone but are also part of the representational issues concerning information
at large, especially when we consider the concept of genres of organizational communication (Yates
and Orlikowski 1991) that we have discussed earlier in chapter 4.
Finally at the knowledge level, dealing with issues such as the content and the processes, we found
that these are more uniquely tied to knowledge distribution and do not concern information
distribution in general. The issues of knowledge content, type of processes for creation/transformation
and distribution were compared and we found that to a great extent they were very similar with little
variations. The processes for transformation concerned mainly the internalization of knowledge
through similar distribution (except in the case of subscriptions) and acquisition modes. This could be
due to that the knowledge distribution means that we had identified handled explicit knowledge from
explicit sources. The different distribution means were not really distinguished until we looked on
substance and purpose of distribution. Here they seemed to serve specific functions such as problem
solving, idea generating, or education among others.
57
When looking at the distribution of knowledge one mustn't forget that there are two sides of the coin.
Distribution concerns both the process of contribution, there is always someone or something that has
gone through great pains in formulating and making available source of knowledge to others, and the
process of adoption, where individuals seek out these sources and retrieve knowledge in order to, as
Huber (1991) stated, make changes in the range of their possible behaviors.
Discussion groups accommodate for a distribution of knowledge in a "conversational" form. The
"distance" between contributor and adopter is relative short and the relationship between them allows
high degree of interactivity. With this type of knowledge distribution means there exists a clear basis
on which to form a shared interpretation between contributor and adopter. Discussion groups seem
most adequate in this respect for knowledge distribution that demands a close interaction for
exchanging tacit knowledge, almost akin to the knowledge transformation process of socialization that
Nonaka and Takeuchi (1995) talk about.
With the other types of knowledge distribution means the process of contribution is not that apparent
and may seem very remote to the adopter. The relationship between contributor and adopter is not
tight and hence the degree of interactivity is low, making shared understanding harder. This may be
overcome by using a means for distributing the knowledge that support characteristics which
facilitates the knowledge sharing such as rich language and multiple cues. Goodman and Darr (1998)
talks about inhibitors to the processes of contribution and adoption such as language, culture,
anonymity, and lack of feedback that reduce the incitement for knowledge exchange. These may easily
be overcome if the knowledge is distributed through means that favor possibilities for creating shared
frames of reference for interpretation and exhibit relative high degree of interactivity in the relation
between contributor and adopter.
Goodman and Darr (1998) also talks about the characteristics of the problem-solution, such as the
degree of complexity and level of tacitness of knowledge that, if they are high, may inhibit the
processes of contribution and adoption. Such inhibitors may be overcome by using knowledge
distribution means that allows for rich language and structural features that may help capture and
support both complexity of the task and the tacitness of knowledge.
In discussing the relation between the contributor and adopter of knowledge one inevitably comes to
touch on the interpretation of knowledge. An activity that is naturally embedded within the process of
contribution is that of formulating the knowledge, making it explicit. If individuals share a frame of
reference or common platform of understanding then the process of explicating the knowledge is much
facilitated due to a common language and way of expressing oneself. Also, a shared frame of reference
has effects on the knowledge adoption since an individual searching for solutions to a problem
situation will find the process of matching a solution to his/her problem much facilitated. These
processes of formulating and matching concern the interpretation of knowledge. This shows us that the
constructs of knowledge distribution and knowledge interpretation are closely linked.
There are several issues that we have not touched on within our study that are interesting to explore
but which we have judged to lie outside the scope of our study. However, there are some issues among
them that we would like to throw some light on. There are several administrative issues that need to be
resolved if we are to effectively leverage knowledge through the organization. These issues are mostly
concerned with the retention or storage of the knowledge. Wijnhoven (1998) has identified some.
First, how to give access rights to different groups of individuals, thereby restricting knowledge or
information that contains industrial intellectual properties of strategic importance. Second, how to
guarantee that valuable knowledge and expertise can be used in the future. The index structure of the
knowledge stores must be an unambiguous one so that all the available and relevant knowledge can be
retrieved from the correct locations. Third, how to allocate responsibilities for different knowledge
retention stores and whom to place as the gatekeeper of particular locations. Not allocating these
responsibilities seems to be the surest way of losing the value of the knowledge and information
involved. One answer that Wijnhoven (1998) gives in order to solve these issues is to locate the
knowledge within the shared organizational memory.
58
Another issue related to knowledge retention is the knowledge maintenance, especially when the
knowledge is updated at one site while at another site people are still using the old and invalid. The
solution proposed through database literature is to reduce redundancy in the organizational knowledge
base. On the other hand one can argue in favor of redundancy since it leads to more people being
actively involved in the improvement of the knowledge (Nonaka and Takeuchi 1995). The question
here is how to avoid differences in knowledge that lead to organizational differentiation, which can
become political in nature, and how members of the organization can improve their learning capacities
by solving knowledge and information conflicts (Senge 1990). These are just some of possible several
different administrative issues that must be resolved and that remains to be explored in future research.
The purpose of this work was to investigate emerging knowledge distribution means and to
characterize these in a descriptive study. The results, especially the framework but also the
characterizations at the different interpretive levels, may be used for developing hypothesis and raise
questions about usage and thus serves as a basis for empirical studies and research. As such, this study
yields the academic benefit of serving as a foundation for theory constructions related to behavioral
aspects of knowledge distribution means. Then, one may ask questions such as why do individuals use
these distribution means, are there any social aspects, such as sub-cultures, underlying the usage of
them and required for them to be used successfully, and under what conditions is the knowledge
distribution enabled? Such questions are not irrelevant and reveal underlying motivation for using the
knowledge distribution means and thus how viable they are. It is obvious that people use them because
its gives them an advantage to the traditional ways of acquiring or disseminating knowledge.
Factors that affect what these different distribution means are used for and how they are used may be
availability, perceived usefulness, personal style, and issues concerning social influences and critical
mass, as we have talked about in an earlier chapter in relation to issues of media choice and use. To
study usage demands that we raise questions concerning behavioral and social aspects surrounding
these distribution means. Our results may in this respect function as a catalyst for raising these
questions and launching an empirical study of behavioristic characteristics of knowledge distribution
means in a social setting. Here the theory adaptive structuration (Poole and DeSanctis 1990) may be a
fruitful frame to study the distribution means from when raising questions of usage since it aims to
explain how technology is adopted when individuals use it in pursuits of their own social practices.
59
60
References
Ackerman, M.S., and Halverson, C. (1998). Considering an Organization's Memory. In the
Proceedings of the ACM Conference on Computer-Supported Cooperative Work (CSCW'98).
Ackerman, M.S., and Malone, T.W. (1990). Answer garden: A tool for growing organizational
memory. In Proceedings of the ACM Conference on Office Automation Systems, pp. 31-39. Boston:
Massachusetts Institute of technology.
Alter, S. (1992). Why Persist with DSS when the Real Issue Is Improving Decision Making? In
DECISION SUPPORT SYSTEMS: Experiences and Expectation, North-Holland: Elsevier Science
Publishers B.V.
Alter, S. (1996). Information systems: a management perspective. Menlo Park, California: The
Benjamin Cummings Publishing Company Inc.
Angus, J., Patel, J., and Harty, J. (1998). Knowledge Management: Great concept but what is it?
Information Week, no. 16, March.
Argyris, C. (1993). On Organizational Learning. Cambridge, MA: Blackwell Publishers.
Argyris, C., and Schön, D.A. (1978). Organizational Learning. Reading, MA: Addison-Wesley.
Baskerville, R. (1998). Emancipatory IT and Knowledge Management. Organizations and Society in
Information Systems, no. 2, July.
Berger, P.L., and Luckmann, T. (1966). Social Construction of Reality. New York: Doubleday.
Brookes, B.C. (1980). The foundations of information science; Part I: Philosophical Aspect. Journal
of Information Science, vol. 2, no. 3-4, October.
Cavaye, A.L.M. (1996). Case study research: a multi-faceted research approach for IS. Information
Systems Journal, 6, pp. 227-242.
Chew, B.W., Leonard-Barton, D., and Bohn, R.E. (1991). Beating Murphy's law. Sloan Management
Review, vol. 32, no. 3, pp. 5-16.
CNET, URL: http://www.cnet.com/.
Conklin, J. (1987). Hypertext: An Introduction and Survey. Computer, September, pp. 17-41.
Culnan, M.J. (1987). Mapping the intellectual Structure of Management Information Systems, 19801985: A Co-citation Analysis. MIS Quarterly, 10, pp. 341-353.
Daft, R.L., and Huber, G.P. (1987). How organizations learn: A communication framework. Research
in the Sociology of Organizations, no 5, pp. 1-36.
Daft, R.L., and Lengel, R.H. (1986). Organizational information requirements, media richness and
structural design. Management Science, vol. 32, no. 5, May.
Daft, R.L., and Macintosh, N.B. (1981). A Tentative Exploration into the Amount and Equivocality of
Information Processing in Organizational Work Units. Administrative Science Quarterly, no. 26, pp.
207-224.
Davenport, T.H, and Prusak, L. (1998). Working Knowledge: How Organizations Manage What They
Know. Boston, Massachusetts: Harvard Business School Press.
61
Dretske, F. (1981). Knowledge and the Flow of Information. Cambridge, MA: MIT Press.
Drucker, P.F. (1988). The coming of the new organization. Harvard Business Review, JanuaryFebruary, pp. 45-53.
Drucker, P.F. (1994). The age of social transformation. Atlantic Monthly, 274, no. 5.
Eco, U. (1976). A theory of semiotics. Bloomington, Indiana: Indiana University Press.
Farhoomand, A.F. (1992). Scientific Progress of Management Information Systems. In Galliers, R.
(Ed.), Information systems research: issues, methods and practical guidelines, pp. 93-111, Oxford:
Blackwell Scientific Publications.
Fiol, C.M., and Lyles, M.A. (1985). Organizational Learning. Academy of Management Review, no.
10, pp. 803-813.
Fulk, J., and Boyd, B. (1991). Emerging Theories of Communication in Organizations. Journal of
Management, vol. 17, no. 2, pp. 407-446.
Fulk, J., Schmitz, J., and Steinfield, C.W. (1990). A social influence model of technology use. In Fulk,
J., and Steinfield, C.W. (Eds.), Organizations and Communication Technology, pp. 117-140, London:
Sage.
Galbraith, J. (1977). Organizational Design. Reading, MA: Addison-Wesley.
Galliers, R.D. (1992). Choosing Information Systems Research Approaches. In Galliers, R. (Ed.),
Information systems research: issues, methods and practical guidelines, pp. 144-162, Oxford:
Blackwell Scientific Publications.
Giddens, A. (1979). Central problems in social theory. Berkley: University of California Press.
Giddens, A. (1984). The constitution of society: Outline of the theory of structure. Berkley: University
of California Press.
Goffman, E. (1981). Forms of Talk. Philadelphia, Pennsylvania: University of Pennsylvania Press.
Goodman, P.S., and Darr, E.D. (1998). Computer-Aided Systems and Mechanisms for Organizational
Learning in Distributed Environments. MIS Quarterly, December.
Hansen, M.T., Nohria, N., and Tierney, T. (1999). What's your strategy for managing knowledge?
Harvard Business Review, March-April.
Hoffer, J.A., and Valacich, J.S. (1993). Group Memory in Group Support Systems: A Foundation for
Design. In Jessup L.M., and Valacich, J.S. (Eds.), Group Support Systems: New Perspectives, pp. 214229, New York: Macmillan Publishing Company.
Huber, G.P. (1984). The Nature and Design of Post-Industrial Organizations. Management Science,
vol. 30, no. 8, August.
Huber, G.P. (1990). A theory of the effects of advanced information technologies on organizational
design, intelligence and decision making. Academy of Management Review, vol. 15, no. 1.
Huber, G.P. (1991). Organizational Learning: The Contributing Processes and the Literatures.
Organization Science, 2:1, February, pp. 88-115.
62
Huber, G.P., and Daft, R.L. (1987). The Information Environment of Organizations. In Jablin, F.,
Putnam, L.L., Roberts, K.H., and Porter, L.W. (Eds.), Handbook of Organizational Communication,
pp. 130-164, Beverly Hill, CA: Sage.
Innis, H. (1972). Empire and communications. Toronto: University of Toronto Press.
Java Developer Connection, URL: http://java.sun.com/.
Keen, P. (1991). Relevance and Rigor in information Systems Research: Improving Quality,
Confidence, Cohesion and Impact. In Information Systems Research: Contemporary Approaches and
Emergent Traditions, pp. 27-49, North-Holland: Elsevier Science Publishers.
Kolb, D.A., (1984). Experiential Learning: Experiences as the Source of Learning and Development.
Englewood Cliffs: Prentice Halls.
Kuhn, T.S. (1970). The Structure of Scientific Revolution. Chicago, IL: University of Chicago Press.
Langefors, B. (1973). Theoretical Analysis of Information Systems. Lund, Sweden: Studentlitteratur.
Laudon, K.C and Laudon, J.P. (1996). Essentials of management information systems. New Jersey:
Prentice hall.
Lee, A.S. (1994). Electronic mail as a medium for rich communication: an empirical investigation
using hermeneutic interpretation. MIS Quarterly, June, pp. 145-157.
Leonard-Barton, D. (1995). Wellsprings of knowledge: building and sustaining the sources of
innovation. Boston: Harvard Business School Press.
Levitt, B., and March, J.G. (1988). Organizational Learning. Annual Review of Sociology, vol. 14,
pp. 319-340.
Luft, J. (1969). Of Human Interaction. Palo Alto, CA: National Press.
Lundberg B.G. (1999). Knowledge Management - toward a systematic approach. In the Proceedings
of Conference on Systems Integration, Prague, Czech Republic.
Machlup, F. (1983). Semantic Quirks in Studies of Information. In Machlup, F., and Mansfield, U.
(Eds.), The study of Information, New York: John Wiley.
Markus, M.L. (1990). Towards a 'critical mass' theory of interactive media. In Fulk, J., and Steinfield,
C.W. (Eds.), Organizations and Communication Technology, pp. 194-218, London: Sage.
Markus, M.L. (1994). Electronic mail as the medium of managerial choice. Organizational Science
5(4), pp. 502-727.
Microsoft ®, URL: http://www.microsoft.com/.
Miller, C.R. (1984). Genre as social action. Quarterly Journal of Speech, 70, pp. 151-167.
Nonaka, I. (1994). A Dynamic Theory of Organizational Knowledge Creation. Organization Science,
vol. 5, no. 1, February.
Nonaka, I., and Takeuchi, H. (1995). The knowledge creating company: how Japanese companies
create the dynamics of innovation. New York: Oxford University Press.
63
Ong, W. (1982). Orality and literacy: The technologizing of the word. London: Methuen, Inc.
Orlikowski, W.J. (1990). In Turner, J.A., Relevance versus rigor in information systems research: an
issue of quality. In Nissen, H-E., Klein, H.K., and Hirschheim, R. (Eds.), Information systems
research: contemporary approaches and emergent traditions, pp. 715-745, Amsterdam: NorthHolland.
Palme, J. (1989). Elektronisk post. Lund, Sweden: Studentlitteratur.
Patel R., and Davidson B. (1991). Forskningsmetodikens grunder: Att planera, genomföra och
rapportera en undersökning. Lund, Sweden: Studentlitteratur.
Polyani, M. (1958). Personal Knowledge. Chicago: University of Chicago Press.
Polyani, M. (1966). The Tacit Dimension. London: Routledge & Kegan Paul.
Poole, M.S., and DeSanctis, G. (1990). Understanding the use of group support systems: The theory of
adaptive structuration. In Steinfield, C., and Fulk, J. (Eds.), Organizations and communication
technology, pp. 173-193, Beverly Hills: Sage Publications.
Poole, M.S., and Jackson, M.H. (1993). Communication Theory and Group Support Systems. In
Jessup L.M., and Valacich, J.S. (Eds.), Group Support Systems: New Perspectives, pp. 281-293, New
York: Macmillan Publishing Company.
Reder, S., and Schwab, R.G. (1988). The communicative economy of the workgroup: multi-channel
centres of communication. Presented at the Conference of Computer-Supported Cooperative Work
(CSCW'88), Portland, Oregon, September, pp. 354-368, New York: ACM Press.
Rice, R.E. (1984). The new media: Communication, research, and technology. Beverly Hills: Sage
Publications.
Rice, R.E. (1987). Computer-mediated communication systems and organizational innovation. Journal
of Communication, 37, 4, pp. 65-94.
Rogers, E.M, and Agarwala-Rogers, R. (1976). Communication in Organizations. New York: The
Free Press.
Rudström, Å. (1995). Applications of Machine Learning. Licentiate thesis, Department of Computer
and Systems Sciences, Royal Institute of Technology/Stockholm University, Sweden.
Rudy I.A. (1996). A critical review of research on electronic mail. European Journal of Information
Systems, 4, pp. 198-213.
Schmitz, J., and Fulk, J. (1991). Organizational colleagues, media richness, and electronic mail: a test
of the social influence model of technology. Communication Research, 18, pp. 487-523.
Senge, P.M. (1990). The fifth discipline: The art and practice of the learning organization. New York:
Doubleday Currency.
Shannon, C.E., and Weaver, W. (1949). The mathematical theory of communications. Urbana, IL:
University of Illinois Press.
Short, J., Williams, E., and Christie, B. (1976). The Social Psychology of Telecommunications.
London: Wiley.
64
Sneiders, E. (1999). Automated FAQ Answering on WWW Using Shallow Language Understanding.
Licentiate thesis, Department of Computer and Systems Sciences, Royal Institute of
Technology/Stockholm University, Sweden.
Stein, E.W., and Zwass, V. (1995). Actualizing Organizational Memory with Information Systems.
Information Systems Research, 6:2.
Steinfield, C.W. (1992). Computer-mediated communications in organizational settings: emerging
conceptual foundations and directions for research. Management Communication Quarterly, 5, pp.
348-365.
Stohl, C., and Redding, W.C. (1987). Messages and message exchange processes. In Putnam, L.L.,
Roberts, K.H., and Porter, L.W. (Eds.), Handbook of organizational communications: An
interdisciplinary perspective, pp. 451-502, Newbury Park, CA: Sage.
Swan, J., Scarbrough, H., and Preston, J. (1999). Knowledge Management - The Next Fad to Forget
People? In Proceedings of the Seventh Annual European Conference on Information (ECIS'99), pp.
668-678. Copenhagen, Denmark: Copenhagen Business School.
Tom's Hardware Guide, URL: http://sysdoc.pair.com/.
Trevino, L.K., Daft, R.L., and Lengel, R.H. (1990). Understanding managers' media choice in
organizations: A symbolic interactionist perspective. In Steinfield, C., and Fulk, J. (Eds.),
Organizations and communication technology, pp. 71-94, Newbury Park, CA: Sage Publications
Trevino, L.K., Lengel, R.H., and Daft, R.L. (1987). Media symbolism, media richness and media
choice in organizations: A symbolic interactionist perspective. Communication Research, vol. 14, no.
5, pp. 553-575.
Tubbs, S.L., and Moss, S. (1994). Human communication, New York: MacGraw-Hill.
Vickery, B., and Vickery, A. (1992). Information Science: in theory and practice. London: BowkerSaur.
Visala, S. (1991). Broadening the Empirical Framework of Information Systems Research. In
Information Systems Research: Contemporary Approaches and Emergent Traditions, pp. 347-364,
North-Holland: Elsevier Science Publishers.
Walker, W.H., and Kintsch, W. (1985). Automatic and Strategic Aspects of Knowledge Retrieval.
Cognitive Science, 9, pp. 261-283.
Walsh, J.P., and Ungson, G.R. (1991). Organizational Memory. Academy of Management Review,
vol. 16, no. 1, pp. 57-91.
Weick, K.E. (1979). The psychology of organizing, 2nd ed. Reading, MA: Addison-Wesley.
Wiig, K.M. (1996). On the management of knowledge. The Knowledge management forum, June 20th.
Wijnhoven, F. (1998). Designing Organizational Memories: Concept and Method. Journal of
Organizational Computing and Electronic Commerce, 8(1), pp. 29-55.
Yates, J., and Orlikowski, W.J. (1992). Genres of organizational communication: a structural
approach to studying communication and media. Academy of Management Review, vol. 17, no. 2, pp.
299-326.
65
Zack, M.H. (1993). Interactivity and communication mode choice in ongoing management groups.
Information Systems Research, 4(3), pp. 207-239.
Zwass, V. (1998). Foundations of Information Systems. USA: Irving/McGraw-Hill.
66