Communications of the Association for Information Systems

Volume 8 Article 18

March 2002

An Extended Model of Knowledge-Flow

Dynamics
Mark E. Nissen
Naval Postgraduate School, [email protected]

Follow this and additional works at: https://aisel.aisnet.org/cais

Recommended Citation
Nissen, Mark E. (2002) "An Extended Model of Knowledge-Flow Dynamics," Communications of the Association for Information
Systems: Vol. 8 , Article 18.
DOI: 10.17705/1CAIS.00818
Available at: https://aisel.aisnet.org/cais/vol8/iss1/18

This material is brought to you by the AIS Journals at AIS Electronic Library (AISeL). It has been accepted for inclusion in Communications of the
Association for Information Systems by an authorized administrator of AIS Electronic Library (AISeL). For more information, please contact
[email protected].
Communications of the Association for Information Systems (Volume 8, 2002) 251-266 251

AN EXTENDED MODEL OF KNOWLEDGE-FLOW

DYNAMICS

Mark E. Nissen
Graduate School of Business and Public Policy
Naval Postgraduate School
[email protected]

ABSTRACT
The modern enterprise depends upon timely and effective flows of knowledge through its
organizations for success. But knowledge is not evenly distributed through the enterprise, and a
dearth of information systems is available to enable such timely and effective flows. Further, the
few theoretical knowledge-flow models available have not yet been developed to a point where
they can effectively inform the design of information systems and business processes to support
knowledge flow in the enterprise. A survey of current practice shows that such system and
process design is accomplished principally by trial and error, one of the least effective
approaches known. The research described in this article builds upon and extends current theory
about knowledge flow. It focuses in particular on investigating flow dynamics to inform the design
of information systems and business processes to enhance the flow of knowledge through the
enterprise. Leveraging the good understanding of flows in other domains, we strive to extend
theory that can lead to "devices" of considerable utility in the enterprise knowledge domain. The
result is a four-dimensional, dynamic model that can be used to classify and visualize a diversity
of knowledge-flow patterns through the enterprise. These patterns can, in turn, be analyzed to
inform the design of useful information systems and business processes. The implications of this
dynamic model are explored and a number of hypotheses are generated to motivate and guide
future research into the phenomenology of knowledge flow.

KEYWORDS: business process re-engineering, knowledge flow, knowledge management,

knowledge transfer.

I. INTRODUCTION
Many scholars [e.g., Drucker 1995] assert that knowledge represents one of the very few
sustainable sources of comparative advantage, and the practice of knowledge management (KM)
takes the power of knowledge to the group, organization and even enterprise level [Davenport
and Prusak 1998]. Within KM, current survey work identifies knowledge transfer as a key area in
need of additional research, indicating that there are "large gaps in the body of knowledge in this
area" [Alavi and Leidner 2001, p. 126]. If we accept that knowledge is an entity that can be
transferred [cf. Brown and Duguid 1998], then familiarity with other transferable entities (e.g.,
electricity, fluids, manufactured items, cargo) leads us to conceptualize this phenomenon in terms
of flow. We leverage our understanding of well-defined flows from both physical and
organizational domains to help build theory to describe the phenomenology of knowledge flow.

An Extended Model of Knowledge Flow Dynamics by M. Nissen

252 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

The primary objective of knowledge flow is to enable the transfer of capability and
expertise from where it resides to where it is needed—across time, space and organizations as
necessary. The problem is, knowledge is not evenly distributed through the enterprise. The
larger, more geographically-dispersed and time-critical an enterprise, the more that it depends on
the timely and effective flow of knowledge through its organizations for success. Exacerbating this
problem is the dearth of information systems available to enable such timely and effective flows.
Notwithstanding the many contemporary information systems labeled "KM tools" (e.g.,
groupware, Web portals, search engines) that are available and being employed in hopes of
enhancing the flow of knowledge through many enterprises, few such tools even address
knowledge as the focus or object of flow. Rather, nearly all contemporary information systems
focus instead of the transfer of information and data, which are qualitatively different across
numerous dimensions [cf. Davenport et al. 1998, Teece 1998]. Further, the few theoretical
knowledge-flow models available [e.g., Dixon 2000, Nonaka 1994] have not yet been developed
to a point where they can effectively inform the design of information systems and business
processes to enable, automate and support knowledge flow in the enterprise. A survey of current
practice [Nissen et al. 2000] shows that such system and process design is accomplished
principally by trial and error, one of the least-effective design approaches known.
The research described in this article builds upon and extends current theory pertaining
to knowledge flow and focuses, in particular, on investigating its dynamics to inform the design of
information systems and business processes. Leveraging the good understanding of flows in
physical (e.g., electronics, aerospace) and organizational (e.g., manufacturing, logistics) domains,
in which many flow-enhancing devices (e.g., amplifiers, engines, assembly lines, distribution
hubs) were developed and demonstrated, we extend theory that can lead to "devices" (e.g.,
knowledge amplifiers and engines) of comparable utility in the enterprise knowledge domain.
This article begins with background on knowledge flow (Section II). Section III describes
our knowledge-flow model development through extension to existing theory, upon which we
reflect to identify important research implications and propose a number of research hypotheses.
The contribution of this work is summarized along with key conclusions to close the article.

II. BACKGROUND
This section draws heavily from Nissen et al. [2000] to summarize key background work
pertaining to knowledge flow. The section begins with an overview of important concepts from the
emerging knowledge management literature. Research to integrate re-engineering with
knowledge management is then covered, after which we outline the theoretical underpinnings
used to model knowledge-flow dynamics in this study.

KNOWLEDGE MANAGEMENT LITERATURE

For purposes of this article, four important concepts from the KM literature are
summarized:

1. knowledge hierarchy,
2. information technology,
3. knowledge-based systems, and
4. knowledge management life cycle.

The corresponding discussion helps frame current thinking and activity in KM. It is specifically
focused on concepts employed in this research.

KNOWLEDGE HIERARCHY
Many scholars [cf. Davenport and Prusak 1998, Nissen et al. 2000, von Krough et al.
2000] conceptualize a hierarchy of knowledge, information, and data. As illustrated in Figure 1,
each level of the hierarchy builds on the one below. For example, data are required to produce
information, but information involves more than just data (e.g., need to have the data in context).

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 253

Similarly, information is required to produce knowledge, but knowledge involves more than just
information (e.g., it enables action). We operationalize the triangular shape of

Actionability

Knowledge

Information

Data

Abundance
Figure 1. Knowledge Hierarchy

this hierarchy using two dimensions—abundance and actionability—to differentiate among the
three constructs.
Briefly, data lie at the bottom level, with information in the middle and knowledge at the
top. The broad base of the triangle reflects the abundance of data, with exponentially less
information available than data and even fewer chunks of knowledge in any particular domain.
Thus, the width of the triangle at each level reflects decreasing abundance in the progress from
data to knowledge. The height of the triangle at each level reflects actionability (i.e., the ability to
take appropriate action, such as a good decision or effective behavior). Converse to their
abundance, data are not particularly powerful for supporting action, and information is more
powerful than data. But knowledge supports action directly, hence its position near the top of the
triangle.
This, notional view of the hierarchy is shared by many scholars, but certainly not all. For
example, Tuomi [2000] argues for an inverted hierarchy, in which hierarchical relationships such
as those outlined above are inverted to reflect data on the “top” and knowledge on the “bottom.”
His argument is that knowledge is required to establish a semantic structure to represent
information, which in turn represents a prerequisite for creating data.
Perhaps this apparent contradiction can be resolved by introducing the concept
directionality in terms of knowledge flow. As depicted in the context of knowledge transfer through
Figure 2, the transferor of knowledge could indeed view the hierarchy as conceptualized by
Tuomi—where knowledge is necessary to produce information, which in turn is necessary for
creating data that is conveyed (e.g., via paper, network, speech, observable action). However,
the receiver of knowledge would view the hierarchy in the opposite perspective outlined above—
where data are placed into context to become information, and information that enables action
becomes knowledge.
The dynamic aspect of knowledge associated with directional flow also represents an
important part of Spiegler’s [2000] alternative conceptualization, which focuses on
transformations (e.g., data to information, information to knowledge). This alternative
conceptualization also supports a double hierarchy, in which such transformations convert data to
knowledge and vice versa. With the flow-directionality concept from above, these alternative
models are not contradictory. The latter model also discusses a level “above” knowledge in the
hierarchy termed wisdom, which also receives speculation in the trade press [cf. Angus 1998,
Mullins 1999]. The present article does not attempt to address “wisdom management.”

An Extended Model of Knowledge Flow Dynamics by M. Nissen

254 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

Data
Information

?
Knowledge

Knowledge
Information
?

Data

Figure 2. Knowledge Flow Directionality

INFORMATION TECHNOLOGY
Current information technology used to support KM is limited primarily to conventional
database management systems (DBMS), data warehouses and data mining tools (DW/DM),
intranets/extranets, portals and groupware [O’Leary 1998]. Arguably, just looking at the word
"data" in the names of many "knowledge management tools" (e.g., DBMS, DW/DM), we are not
even working at the level of information, much less knowledge. Although (esp. Web-based)
Internet tools applied within and between organizations provide a common, machine-independent
medium for the distribution and linkage of multimedia documents, current intranet and extranet
applications focus principally on the management and distribution of information, not knowledge
per se.
Along these same lines, groupware offers infrastructure support for knowledge work and
enhances the environment in which knowledge artifacts are created and managed, but the flow of
knowledge itself remains indirect. For example, groupware is widely noted as helpful in:

• the virtual office environment (e.g., when geographically-dispersed knowledge

workers must collaborate remotely),
• provides networked tools such as shared, indexed and replicated document
databases and discussion threads (e.g., Lotus Notes/Domino applications),
• shared "white boards,"
• joint document editing capabilities, and
• full-duplex, multimedia communication features.

These tools serve to mitigate collaboration losses that can arise when rich, face-to-face joint work
is not practical or feasible. But supporting (even rich and remote) communication is not sufficient
to guarantee knowledge flow.

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 255

KNOWLEDGE-BASED SYSTEMS
Construction and use of knowledge-based systems (KBS) can make knowledge explicit
and its application direct. Key KBS technologies include applications such as:

• expert systems and intelligent agents,

• infrastructure and support tools such as ontologies, knowledgebases, inference
engines, search algorithms, list and logic programming languages, and
• a variety of representational formalisms (e.g., rules, frames, scripts, cases, models,
semantic networks).

Much deeper than just their names' sake, KBS are predicated on the capture, formalization and
application of strong domain knowledge. The use of KBS for knowledge organization and
distribution is well known, widespread, and now the subject of textbook application [cf. Russell
and Norvig 1995, Turban and Aronson 2001].
Unlike the extant IT tools noted above, the substance of KBS is knowledge itself—not just
information or data—and KBS are designed to interpret and apply represented knowledge
directly. These capabilities and features make KBS distinct from most classes of IT applications
presently employed for KM [cf. Smith and Farquhar 2000]. However, expert system
development—through classic knowledge engineering—requires explicit capture and
formalization of tacit knowledge possessed by experts. This is just the kind of tacit knowledge that
researchers [e.g., Leonard and Sensiper 1998, p. 112] stress "underlies many competitive
capabilities." However, such knowledge has long been known as being "hard to capture."

KNOWLEDGE MANAGEMENT LIFE CYCLE

Nissen et al. [2000] observe a sense of process flow or a life cycle associated with
knowledge management. Integrating their survey of the literature [e.g., Despres and Chauvel
1999, Gartner Group 1999, Davenport and Prusak 1998, Nissen 1999], they synthesize an
amalgamated KM life cycle model as outlined in Table 1. Notice that most of the four life cycle
models begin with a "create" or "generate" phase; only the Nissen model begins with knowledge
capture, an activity appearing in the third phase of the Gartner Group model. The second phase
pertains to the organization, mapping or bundling of knowledge; Davenport and Prusak do not
make specific reference to this organization phase from their model, but it is implied by their
codify phase and appears very prominently in all the others.

Table 1. Knowledge Management Life Cycle Models

[Adapted from Nissen et al. 2000]

Model Phase 1 Phase 2 Phase 3 Phase 4 Phase 5 Phase 6

Despres and Create Map/ Store Share/ Reuse Evolve

Chauvel bundle transfer
Gartner Group Create Organize Capture Access Use
Davenport & Generate Codify Transfer
Prusak
Nissen Capture Organize Formalize Distribute Apply
Amalgamated Create Organize Formalize Distribute Apply Evolve

Phase three uses different terms across the models, but they all address some
mechanism for making knowledge formal or explicit. Likewise, the fourth phase uses different
terms but addresses the ability to share or distribute knowledge in the enterprise. Three of the
four models include a fifth phase for application or (re)use of knowledge for problem solving or
decision making in the organization, but such application and (re)use is implied as an objective in

An Extended Model of Knowledge Flow Dynamics by M. Nissen

256 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

all. Only the Despres and Chauvel model explicitly includes a sixth phase for knowledge
evolution.
The Amalgamated Model integrates the key concepts and terms from the four life cycle
models. Compare the steps proposed by Nissen (1999), for example, with the Amalgamated
Model. Notice from Table 1 the Amalgamated life cycle model makes a distinction between
knowledge creation (as proposed by Despres and Chauvel and Gartner Group) and its capture or
formalization (i.e., Phase 3). Whereas knowledge creation involves discovery and the
development of new knowledge, knowledge capture requires only that the knowledge be new to a
particular individual or organization, and formalization involves the conversion of existing
knowledge from tacit to explicit form. The Amalgamated Model therefore seems more complete
with its beginning at the creation step. Similarly, the Amalgamated Model also adopts the
evolution step from Despres and Chauvel.
Drawing further from this research on life cycle models, Nissen et al. note that coverage
of existing information systems and business practices across these life cycle phases is patchy.
For example, numerous systems and practices are identified from the literature, but they support
only three of the six life cycle phases: knowledge organization, formalization and distribution.
Alternatively, relatively few counterpart systems and practices are found to correspond with the
other three phases: knowledge application, evolution and creation. We thus observe a relative
abundance of systems and practices available to support three of the phases of the KM life cycle
and a dearth for the other three phases.

RE-ENGINEERING AND KNOWLEDGE MANAGEMENT INTEGRATION

Substantial integration of knowledge management with re-engineering is observed in
current practice, as companies realize the direct connection between KM and knowledge-work
process innovation [Davenport et al. 1998]. In their study of more than thirty KM efforts in
industry, Davenport et al. [1996] note the practice is "fundamentally change management
projects." Emerging theory of knowledge creation and management has a dynamic, distinctly
process-oriented flavor [see esp. Nonaka 1994]. Ruggles [1998] goes so far as to suggest a
primary objective of practice is to assess the impact of KM as a process, fundamentally a
proposition of re-engineering.
However, as learned through the painful, expensive and failure-prone "first wave" of re-
engineering [Cypress 1994], simply inserting IT into a process in no way guarantees performance
improvement. Indeed, many otherwise successful and effective firms experience process
degradation as the result of re-engineering [cf. Caron et al. 1994, Hammer and Champy, 1993].
This point is underscored by Hammer [1990], whom colorfully refers to such practice as
"automating the mess" (e.g., making a broken process simply operate—broken—faster).
Drawing on Leavitt [1965] and others [cf. Davenport 1993, Nissen 1998], new IT needs to
be integrated with the design of the process it supports. That is, the organization, people,
procedures, culture and other key factors need to be considered in addition to technology. Given
that many KM projects now revolve around IT implementation (e.g., intranets/extranets, Web
portals, groupware; [cf. Nissen et al. 2000]), re-engineering and knowledge management even
appear to be sharing some of the same mistakes.
Building upon this research, we begin to characterize a powerful interaction between the
flow of work (i.e., workflow; [cf. Georgakopoulos et al. 1995]) and the flow of knowledge (i.e.,
knowledge flow) in an enterprise. Following Oxendine and Nissen [2001], we refer to these flows
as horizontal processes and vertical processes, as conceptualized in Figure 3. Briefly, the two
horizontal directed graphs in the figure delineate separate examples of a work process (e.g.,
steps 1 – 6 as performed at different points in time, space, organization). The graph at the top of
Figure 3 represents one particular example (e.g., performed at a specific point in time, location,
organization) of this notional process, and the graph at the bottom represents a different example
(e.g., performed at a separate point in time, location, organization).

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 257

Knowledge-Flow Process
Time, Space, Org
Step 11 Step 21 Step 31 Step 41 Step 51 Step 61

Step 12 Step 22 Step 32 Step 42 Step 52 Step 62

Figure 3. Horizontal and Vertical Processes

Both horizontal graphs represent the flow of work through the enterprise. The vertical
graph represents a complementary set of processes responsible for the flow of knowledge. As
noted in Section I, knowledge is not evenly distributed through the enterprise, yet enterprise
performance is dependent upon consistency and effectiveness across various workflows. The
associated knowledge (e.g., process procedures, best practices, tool selection, and usage) flow
across time, space and organizations. Such cross-process activities are seen as driving the flow
of knowledge—as opposed to the flow of work—through the enterprise. Indeed, Nissen and
Espino [2000] identify seven vertical processes (e.g., training, personnel assignment, IT support)
that interact in a complex manner that is not reflected by the simple, linear flow depicted in the
figure. It is upon these vertical process flows that we concentrate in this research.

CURRENT KNOWLEDGE-FLOW THEORY

One of the best-known theoretical treatments of knowledge flow to date stems from
Nonaka [1994] in the context of organizational learning. This work outlines two dimensions for
knowledge:
• epistemological, and
• ontological.
The epistemological dimension depicts a binary contrast between explicit and tacit knowledge.
Explicit knowledge can be formalized through artifacts such as books, letters, manuals, standard
operating procedures, and instructions, whereas tacit knowledge pertains more to understanding
and expertise contained within people’s minds. The ontological dimension depicts knowledge that
is shared with others in groups or larger aggregations of people across the organization. Although
this aggregation of organizational units appears arbitrary, in the enterprise context, it could clearly
apply to small teams, work groups, formal departments, divisions, business units, firms and even
business alliances or networks.
As shown in Figure 4, Nonaka uses the interaction between these dimensions as the
principal means for describing knowledge flow. This flow is roughly characterized through four
steps.
First, Nonaka asserts that new knowledge is created only by individuals in the
organization and is necessarily tacit in nature. The first flow of knowledge is then theorized to
occur through a process termed socialization, which denotes members of a team sharing
experiences and perspectives, much as one anticipates through communities of practice.

An Extended Model of Knowledge Flow Dynamics by M. Nissen

258 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

Socialization flow is noted as vector 1 in Figure 4 and corresponds to tacit knowledge (i.e., along
the epistemological dimension) flowing from the individual to the group level (i.e., along the
ontological dimension).

Epistemological

Explicit
3. Combination

2. Externalization
4. Internalization
1. Socialization

Tacit

Individual Group Organization Inter-organization

Ontological

Figure 4. Nonaka Knowledge Flow Theory

[Adapted from Nonaka 1994]

The second flow of knowledge (vector 2) is theorized to occur through a process termed
externalization, which denotes the use of metaphors through dialog that leads to articulation of
tacit knowledge and its subsequent formalization to make it concrete and explicit.
The third flow of knowledge (vector 3) is theorized to occur through a process termed
combination. Combination denotes coordination between different groups in the organization—
along with documentation of existing knowledge—to combine new, intra-team concepts with
other, explicit knowledge in the organization.
The fourth flow of knowledge (vector 4) is theorized to occur through a process termed
internalization. Internalization denotes diverse members in the organization applying the
combined knowledge from above—often through trial and error—and in turn translating such
knowledge into tacit form at the organization level.

III. KNOWLEDGE-FLOW DYNAMICS

This section begins by building upon Nonaka’s theory to conceptualize an extended
model of knowledge-flow dynamics. This extended model is intended to help researchers to
understand better the phenomenology of knowledge flow, and as a theoretical contribution, it may
help to describe better and explain how knowledge flows through the enterprise.

BUILDING UPON CURRENT THEORY

The first step toward building on current knowledge-flow theory is to augment Nonaka’s
two-dimensional framework by incorporating a third dimension, the KM life cycle. We

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 259

operationalize the construct using the life-cycle stages from the Amalgamated Model that was
presented in Table 1. Further, because the concept of flow is inherently dynamic, we extend this
framework by incorporating time as a key, fourth dimension. Such augmented dimensionality
preserves—and indeed subsumes—Nonaka’s two-dimensional framework and provides the basis
for a richer model. This richer model may enhance our descriptive and explanatory power in
terms of understanding the knowledge-flow phenomenon.
The second step toward building on current knowledge-flow theory pertains to the
epistemological dimension (Section II) that includes only binary states (i.e., tacit, explicit). In
contrast, we propose that knowledge fills a continuum between the tacit and explicit endpoints.
Instead of a simple contrast between explicit and tacit knowledge, a continuum allows tracing
knowledge as it flows through a continuous range of explicitness. A continuous dimension makes
for a richer model than—and indeed subsumes—one with only two binary states.
This same rationale can be applied to the ontological dimension (Section II). It supports
only a few, granular states (e.g., individual, group, organization). In contrast, we propose that
knowledge may fill a continuum along the dimension characterized by how many people are
reached by the knowledge (e.g., at a particular level of explicitness, life cycle phase). Tracing
knowledge flows across a continuous dimension similarly makes for a richer model than—and
indeed subsumes—one with only a few discrete states.
A third step toward theory building stems from the differentiation between vertical
processes and their horizontal-process counterparts (Section II) in terms of enabling knowledge
flows versus work flows, respectively. Such differentiation is simply absent from current theory.
But it highlights an important, cross-process focus of knowledge flow, and it may help explain the
mechanics associated with prior theory (e.g., Nonaka’s concepts of socialization, externalization,
combination). Moving from description to explanation represents an important aspect of theory
building [Bacharach 1989].

AN EXTENDED MODEL OF KNOWLEDGE FLOW

The theory building in the previous subsection provides the basis from which to develop
an extended model of the knowledge-flow phenomenon. In Figure 51, we note a few, notional,
knowledge-flow vectors for illustrating and classifying various dynamic patterns of knowledge as it
flows through the enterprise. For example, the simple, linear flow labeled “Policies and
Procedures” depicts the manner in which most enterprises inform and train employees through
the use of policies and procedures: explicit documents and guidelines that individuals in the
organization are expected to memorize, refer to and observe. As another example, the cyclical
flow of knowledge described by the amalgamated KM life cycle model (Table 1), shown in the
figure, reflects a more-complex dynamic than its simple, linear counterpart. This flow describes a
cycle of knowledge creation, distribution and evolution within a workgroup, for example.
Further, Nonaka’s dynamic theory of knowledge flow can also be described in this space
by the curvilinear vector sequence corresponding to the processes labeled “create,” “socialize,”
“externalize,” “combine” and “internalize,” respectively. Thus, our model subsumes the one
proposed by Nonaka and shows a somewhat-complex dynamic as knowledge flows along the life
cycle. Moreover, examination of this space suggests also including the refine vector, which is not
part of Nonaka’s theory but represents a key element of the empirically-derived, Amalgamated
Model (e.g., the key to knowledge evolution). Clearly, a great many other flows and patterns can
be shown in this manner. Preliminary results from field work [cf. Nissen 2001 for research
agenda] suggest that this vector-space approach to depicting and visualizing knowledge flows
can be very useful for empirical investigation into the phenomenology of knowledge flow.
To complete our model development, we incorporate the time dimension into the model.
Because static displays such as the graph presented in Figure 5 are difficult to visualize in more
than three dimensions, we do not attempt to show all four dimensions at once. Rather, we

1
Because Nonaka’s terminology for the dimensions reflected in Figure 4 can lead to confusion (e.g., with
respect to use of the terms epistemological and ontological), we substitute the term explicitness for
epistemological and reach for ontological in Figure 5.

An Extended Model of Knowledge Flow Dynamics by M. Nissen

260 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

substitute the dimension time for its life cycle counterpart in Figure 6 to characterize the order-of-
magnitude differences in flow times associated with various kinds of knowledge.

E x plic itn e ss
Exp licit
P olic ie s & P ro ce d ure s

Int e rna lize

C om b ine

E x te rn aliz e

S oc ia lize
R ef ine

Tacit C rea t e
e
In divid ual Group Orga nization I nter-organ iza tion
eat

R ea c h
Cr
ze
ani
rg
O
e
iz
al
rm
Fo
e

K M L ife Cy c le
ut
r ib
is t
D
y
pl
Ap
ve
ol
Ev

Lif e Cy c le

Figure 5. Extended Model with Knowledge Flows

E xplicitn ess

S /W in sta lla tio n

Do c c re a tio n

Ta c it
1 10 100
Reach
rs
ou
H
s
ay

Te a m bu il di ng
D
ks
ee
W
s
th
on
M
s
ar
Ye
es

S /W de ve lo p m en t
ad
ec
D

Flo w T ime

Figure 6. Knowledge Flows with Time Dimension

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 261

For example, if we take some highly explicit knowledge—say a printed document

describing how to install a major software application on a desktop personal computer—then this
flow of knowledge can conceivably be completed in a matter of hours. A computer literate person
needs only to read the instructions before being able to install the software effectively. We plot
this first example of knowledge flow (i.e., “S/W installation”) in Figure 6, which delineates its
three-dimensional classification as:
• highly-explicit knowledge,
• involving only one person in terms of reach,
• with flow time on the order of hours.
This example distinguishes between the flow of work (i.e., installing software) and the
complementary flow of knowledge (i.e., understanding installation instructions) that enables such
work to be done. Thus, in diagrams such as Figure 6, we explicitly define the kinds of vertical
processes (i.e., knowledge flows) that drive knowledge required to perform horizontal processes
(i.e., workflows).
As another instance, take this same document and consider the flow associated with its
creation. Presumably, the authors of a software-installation document would be knowledgeable
about the corresponding software application, as well as how to write effective installation
instructions. Depending on how extensive and complex an application is, understanding its
installation idiosyncrasies could take several months, even though writing the instructions
themselves could probably be accomplished in a matter of weeks. Here again, we differentiate
between the flow of work (i.e., writing installation instructions; requiring weeks) and the
corresponding flow of knowledge (i.e., understanding software-installation idiosyncrasies;
requiring months) that enables such work to be performed effectively.
Notice also, the flow time associated with knowledge required to develop the installation
document (e.g., months) is one or more orders of magnitude longer than the complementary flow
time associated with an individual understanding how to install the software (e.g., hours). We plot
this second knowledge flow (i.e., “Doc creation”) in the figure, as it represents moderately-explicit
knowledge (e.g., some tacit knowledge is required to understand the software), involving a group
of people (e.g., 10) in terms of reach (e.g., assuming that people from several different
organizations are required to develop the instructions), and requiring months for the knowledge to
complete its flow.
As a third instance, consider development of the software application itself. Again
depending upon the extensiveness and complexity, a comparatively-long period of time could be
required for people to acquire the levels of software-engineering knowledge and experience
necessary for its development. Consider that the software architects and engineers must
complete several years of college and acquire numerous years of software experience before
developing the knowledge and experience required to develop a major software product. In some
cases, a decade or more may be required for the requisite knowledge to complete its flow and
enable someone to develop the software application. We plot this third instance (i.e., “S/W
development”) in Figure 6.The knowledge required to develop software is relatively tacit
compared to the knowledge required to write the installation manual. We show this knowledge
flow at the individual level.
To develop a major software application such as discussed in this example, a large
number of individual software architects and engineers must learn to work effectively in groups
and organizations. This team-building knowledge flow is depicted (i.e., “Team building”) in the
figure, concerning relatively-tacit knowledge, involving many people (e.g., 100) in a relatively-
large organization in terms of reach, and conceivably requiring years to complete its flow. Once
again, we distinguish between the flow of work to develop a software application and the
complementary flow of knowledge (e.g., software engineering, team-building) that drives it.
Further, if we trace the flow of knowledge through the life cycle shown in Figure 6, we
develop a composite illustration of its dynamics. For example, the knowledge begins with
education and experience qualifying a person to develop software (i.e., “S/W development”),
which is highly tacit and requires a decade or more to flow for a given individual. The next flow
involves building an effective team to develop the software (i.e., “Team building”), which we

An Extended Model of Knowledge Flow Dynamics by M. Nissen

262 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

delineate as tacit and involving many people over years. The knowledge required to write and test
the installation instructions is more explicit, requiring fewer people and less time to acquire (i.e.,
“Doc creation”), and the flow of knowledge associated with reading the instruction and installing
the software is labeled “S/W installation” (i.e., highly-explicit, individual, hours). We use the broad
arrows in Figure 7 to show the composite knowledge-flow vector as it crosses these four life-cycle
phases.

E xplicitn ess

S /W in sta lla tio n

Do c c re a tio n

Ta c it
1 10 100
Reach
rs
ou
H
s
ay

Te a m bu il di ng
D
ks
ee
W
s
th
on
M
s
ar
Ye
es

S /W de ve lo p m en t
ad
ec
D

Flo w T ime

Figure 7. Composite Knowledge Flow

In this section we built on existing theory to extend a model of knowledge-flow dynamics.

We illustrated the use of this model by classifying and describing several knowledge-flow vectors
associated with the development and installation of a major software application. On the surface,
this extended model appears to be considerably richer and more-explanatory than the prior
theoretical counterparts upon which it builds, and it may prove to be useful for understanding the
phenomenology of knowledge flow better. Such better understanding, we argue, is important for
informing the design of useful information systems and business processes to enhance the flow
of knowledge in the enterprise. But, as a theoretical proposition, this assertion remains to be
examined and verified empirically.

IV. RESEARCH IMPLICATIONS AND HYPOTHESES

So what are the implications of this extended model of knowledge-flow dynamics, and
what new research hypotheses can we develop to help motivate and guide future research? In
this section, we discuss three such implications and propose corresponding hypotheses.

FIRST IMPLICATION
The first implication is associated with the increased dimensionality of the extended
knowledge-flow model. This extension provides a richer model and subsumes the prior work of
Nonaka and others. It also offers a multidimensional framework for attempting to classify various
knowledge flows. We can further use these knowledge flows to help characterize, visualize, and
compare the diversity of flows expected to exist in modern enterprises. Further, these dimensions
may prove useful to identify various knowledge-flow patterns that can be observed or otherwise

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 263

inferred through empirical work. With such patterns identified, we may be able to then organize
and correlate them, perhaps developing a taxonomy. Three research hypotheses relate directly to
this increased dimensionality.

Hypothesis 1. The dimensions associated with the Extended Knowledge-Flow Model can be
used to classify and plot a variety of knowledge flows.

Hypothesis 2. The Extended Knowledge-Flow Model can be used to identify and discriminate
between distinct knowledge-flow patterns.

Hypothesis 3. Distinct knowledge-flow patterns can be matched with the IT most appropriate for
their automation and support.

SECOND IMPLICATION
The second implication is associated with the vertical processes posited to drive the flow of
knowledge. This insight into the mechanics of knowledge flow is useful for contrasting the flow of
work through an enterprise with the complementary flow of knowledge. Further, we can consider
that the time frames associated with some knowledge flows may be orders of magnitude longer
than those associated with the corresponding workflows, as well as other flows such as
information and data through the enterprise. Indeed, even flow times associated with alternative
knowledge flows may differ by one or more orders of magnitude. Three research hypotheses
relate directly to the vertical processes.

Hypothesis 4. Vertical processes can be identified and related to complementary workflows.

Hypothesis 5. Specific vertical processes drive distinct knowledge-flow patterns.

Hypothesis 6. Distinct knowledge-flow patterns can be matched with their corresponding vertical
processes in terms of problem diagnosis and redesign.

THIRD IMPLICATION
The third implication is associated more generally with the phenomenology of knowledge
flow. In terms of theory development, the knowledge flow concept may prove useful to describe
the dynamics of knowledge management, and models to describe and explain the mechanics of
how knowledge flows in the enterprise may be developed to expand our understanding of this
phenomenon. If we are ultimately interested in developing useful “devices” to automate and
support enterprise knowledge flows, then one can argue that we will first need to understand the
associated mechanics. Otherwise, we must continue to rely on the kinds of trial-and-error design
approaches being employed today. In domains such as electronics, aerospace, manufacturing
and logistics, an understanding of the underlying flow mechanics represents a prerequisite to
developing useful devices (e.g., amplifiers, engines, assembly lines, distribution hubs) to enhance
and optimize such flows. We can foresee no difference in terms of the enhancement and
optimization of knowledge flow.

V. CONCLUSION
The modern enterprise depends upon timely and effective flows of knowledge through its
organizations for success. But knowledge is not evenly distributed through the enterprise, and a
dearth of information systems is available to enable timely and effective flows. Further, the few
theoretical knowledge-flow models available have not yet been developed to a point where they
can effectively inform the design of information systems and business processes to support
knowledge flow in the enterprise. A survey of current practice shows that such system and
process design is accomplished principally by trial and error, one of the least effective
approaches known.
The research described in this article builds upon and extends current theory pertaining
to knowledge flow. It focuses in particular on investigating its dynamics to inform the design of

An Extended Model of Knowledge Flow Dynamics by M. Nissen

264 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

information systems and business processes to enhance the flow of knowledge through the
enterprise. Leveraging the good understanding of flows in other domains, we extend the theory
that can lead to "devices" of considerable utility in the enterprise knowledge domain. The result is
a four-dimensional, dynamic model that can be used to classify and visualize a diversity of
knowledge-flow patterns through the enterprise. The patterns can, in turn, be analyzed to inform
the design of useful information systems and business processes.
This extended, dynamic model makes a theoretical contribution by enriching our
descriptive capability through increased dimensionality, and it increases our explanatory
capability by delineating some mechanics associated with the flow of knowledge. Moreover, by
differentiating between flows of knowledge and their complementary flows of work through the
enterprise, we identify an important dynamic in terms of organizational capability. Further, we
explored the implications of this dynamic model and generated a number of hypotheses to help
motivate and guide future research into the phenomenology of knowledge flow. We look forward
to contributing to such future phenomenological research.

ACKNOWLEDGEMENT
This project is sponsored by the Office of Naval Research, Young Investigator Program:
N0001401WR20304.

Editor’s Note: This article was received on January 23, 2002 and was published on March 4, 2002.

REFERENCES
Editor’s Note: The following reference list contains hyperlinks to World Wide Web pages. Readers who have
the ability to access the Web directly can gain direct access to these linked references. Readers are warned,
however, that:
1. These links existed as of the date of publication but are not guaranteed to be working thereafter.
2. The contents of Web pages may change over time. Where version information is provided in the
References, different versions may not contain the information or the conclusions referenced.
3. The author(s) of the Web pages, not AIS, is (are) responsible for the accuracy of their content.
4. The author of this article, not AIS, is responsible for the accuracy of the URL and version information.

Alavi, M. and D.E. Leidner (2001) “Review: Knowledge Management and Knowledge
Management Systems: Conceptual Foundations and Research Issues,” MIS Quarterly (25)1, pp.
107-136.
Angus, J. (1998) “Executive Report: The Organization and Knowledge,” Informationweek
Online (14 September).
Bacharach, S.B. (1989) "Organizational Theories: Some Criteria for Evaluation," Academy of
Management Review (14)4, pp. 496-515.
Brown, J.S. and P. Duguid (1998)"Organizing Knowledge," California Management Review
(40)3, pp. 90-111.
Caron, J.R., S.L., Jarvenpaa and D.B. Stoddard (1994) "Business re-engineering at CIGNA
corporation: experiences and lessons learned from the first five years," MIS Quarterly (18)3, pp.
233-250.
Cypress, H.L. (1994) "Re-engineering - MS/OR imperative: make second generation of
business process improvement mode work," OR/MS Today(21)1, pp. 18-29.
Davenport, T.H. (1993) Process Innovation: Re-engineering Work through Information
Technology Boston, MA: Harvard University Press.
Davenport, T.H., S.L. Jarvenpaa, and M.C. Beers (1996) "Improving Knowledge Work
Processes," Sloan Management Review (37)4, pp. 53-65.
Davenport, T.H., D.W. De Long, and M.C. Beers (1998) "Successful Knowledge Management
Projects," Sloan Management Review (39)2, pp. 43-57.
Davenport, T.H. and L. Prusak (1998)Working Knowledge: How Organizations Manage what
they Know. Boston, MA: Harvard Business School Press.
Despres, C. and D. Chauvel (1999)"Mastering Information Management: Part Six – Knowledge
Management," Financial Times (8 March), pp. 4-6.

An Extended Model of Knowledge Flow Dynamics by M. Nissen

Communications of the Association for Information Systems (Volume 8, 2002) 251-266 265

Dixon, N.M. (2000) Common Knowledge Boston, MA: Harvard Business School Press.
Drucker, P.F. (1995) Managing in a Time of Great Change. New York, NY: Truman Talley.
Gartner Group (1998) "Knowledge Management Scenario," conference presentation,
Stamford, CN, presentation label SYM8KnowMan1098Kharris.
Georgakopoulos, D., M. Hornick, and A. Sheth (1995) "An Overview of Workflow
Management: From Process Modeling to Workflow Automation Infrastructure," Distributed and
Parallel Databases (3)2, pp. 119-153.
Hammer, M. (1990) "Re-engineering Work: Don't Automate, Obliterate," Harvard Business
Review, pp. 104-112.
Hammer, M. and J. Champy (1993) Re-engineering the Corporation: A Manifesto for
Business Revolution New York, NY: Harper Business Press.
Leavitt, H.J. (1965) "Applying organizational change in industry: structural, technological and
humanistic approaches," in J. March (Ed.), Handbook of Organizations Chicago, IL: Rand
McNally.
Leonard, D. and S. Sensiper (Spring 1998) "The Role of Tacit Knowledge in Group
Innovation," California Management Review (40)3, pp. 112-132.
Mullins, C.S. (1999) “What Is Knowledge and Can It Be Managed,” The Data and
Administration Newsletter 8.0, March www.tdan.com/i008fe03.htm.
Nissen, M.E. (1998) "Redesigning Re-engineering through Measurement-Driven Inference,"
MIS Quarterly (22)4, pp. 509-534.
Nissen, M.E. (1999) "Knowledge-Based Knowledge Management in the Re-engineering
Domain," Decision Support Systems (27), pp. 47-65.
Nissen, M.E. and J.P. Espino (2000) "Knowledge Process and System Design for the Coast
Guard," Knowledge and Process Management Journal (7)3, pp. 165-176.
Nissen, M.E., M.N. Kamel, and K.C. Sengupta (2000) "Integrated Analysis and Design of
Knowledge Systems and Processes," Information Resources Management Journal (13)1, pp. 24-
43.
Nissen, M.E. (2001) "Toward a Program of Research on Knowledge Flow in the Very-Large
Enterprise," Monterey, CA; Naval Postgraduate School NPS Technical Report NPS-GSBPP-01-
003.
Nonaka, I. (1994) "A Dynamic Theory of Organizational Knowledge Creation," Organization
Science (5)1, pp. 14-37.
O'Leary, D.E. (1998) "Enterprise Knowledge Management," Computer (31)3, pp. 54-61.
Oxendine, E. and M.E. Nissen (2001) "Knowledge Process and System Design for the Naval
Battlegroup," Journal of the KMCI (1)3, pp. 89-109.
Russell, S.J. and P. Norvig, (1995) Artificial Intelligence : A Modern Approach. Englewood
Cliffs, NJ: Prentice Hall.
Smith, R.G. and A. Farquhar (2000) “The Road Ahead for Knowledge Management,” AI
Magazine (24)1, pp. 17-40.
Spiegler, I. (2000) “Knowledge Management: a New Idea or a Recycled Concept?,”
Communications of the Association for Information Systems (3)14, pp. 1-24.
Teece, D.J. (Spring 1998) "Research Directions for Knowledge Management, California
Management Review (40)3, pp. 289-292.
Tuomi, I. (1999) “Data is More than Knowledge: Implications of the Reversed Knowledge
Hierarchy for Knowledge Management and Organizational Memory,” Journal of Management
Information Systems (16)3, pp. 103-117.
Turban, E. and J. Aronson (2001) Decision Support Systems and Intelligent Systems (Fifth
Edition), Upper Saddle River, NJ: Prentice-Hall.
von Krough, G., K. Ichijo, and I. Nonaka (2000) Enabling Knowledge Creation: How to Unlock
the Mystery of Tacit Knowledge and Release the Power of Innovation New York, NY: Oxford
University Press.

An Extended Model of Knowledge Flow Dynamics by M. Nissen

266 Communications of the Association for Information Systems (Volume 8, 2002) 251-266

ABOUT THE AUTHOR

Mark E. Nissen (http://web.nps.navy.mil/~menissen/) is Assistant Professor of
Information Systems at the Naval Postgraduate School (NPS) and Young Investigator for the
Office of Naval Research. His research focuses on the investigation of knowledge systems for
enabling and managing change in areas such as process innovation, electronic business, and
knowledge flow. He has been investigating knowledge systems to innovate processes in various
domains for over a decade, and developing and experimenting with multi-agent systems for
supply chain dis/re-intermediation for several years. His current research concentrates on the
phenomenology of knowledge flow. Mark’s publications span the information systems and related
fields, and he published his first book in 2000. In 2000, he also received the Menneken Faculty
Award for Excellence in Scientific Research, the top research award available to faculty at NPS.
His Ph.D. in Information Systems is from the University of Southern California.

Copyright © 2002 by the Association for Information Systems. Permission to make digital
or hard copies of all or part of this work for personal or classroom use is granted without fee
provided that copies are not made or distributed for profit or commercial advantage and that
copies bear this notice and full citation on the first page. Copyright for components of this work
owned by others than the Association for Information Systems must be honored. Abstracting with
credit is permitted. To copy otherwise, to republish, to post on servers, or to redistribute to lists
requires prior specific permission and/or fee. Request permission to publish from: AIS
Administrative Office, P.O. Box 2712 Atlanta, GA, 30301-2712 Attn: Reprints or via e-mail from
[email protected] .

An Extended Model of Knowledge Flow Dynamics by M. Nissen

 ISSN: 1529-3181
EDITOR-IN-CHIEF
Paul Gray
Claremont Graduate University
AIS SENIOR EDITORIAL BOARD
Rudy Hirschheim Paul Gray Phillip Ein-Dor
VP Publications AIS Editor, CAIS Editor, JAIS
University of Houston Claremont Graduate University Tel-Aviv University
Edward A. Stohr Blake Ives Reagan Ramsower
Editor-at-Large Editor, Electronic Publications Editor, ISWorld Net
Stevens Inst. of Technology University of Houston Baylor University

CAIS ADVISORY BOARD

Gordon Davis Ken Kraemer Richard Mason
University of Minnesota University of California at Irvine Southern Methodist University
Jay Nunamaker Henk Sol Ralph Sprague
University of Arizona Delft University University of Hawaii

CAIS EDITORIAL BOARD

Steve Alter Tung Bui H. Michael Chung Donna Dufner
University of San University of Hawaii California State University of Nebraska -
Francisco University Omaha
Omar El Sawy Ali Farhoomand Jane Fedorowicz Brent Gallupe
University of Southern The University of Hong Bentley College Queens University, Canada
California Kong, China
Robert L. Glass Sy Goodman Joze Gricar Ruth Guthrie
Computing Trends Georgia Institute of University of Maribor California State University
Technology Slovenia
Chris Holland Juhani Iivari Jaak Jurison Jerry Luftman
Manchester Business University of Oulu Fordham University Stevens Institute of
School, UK Finland Technology
Munir Mandviwalla M.Lynne Markus Don McCubbrey Michael Myers
Temple University City University of Hong University of Denver University of Auckland,
Kong, China New Zealand
Seev Neumann Hung Kook Park Dan Power Maung Sein
Tel Aviv University, Israel Sangmyung University, University of Northern Agder University College,
Korea Iowa Norway
Peter Seddon Doug Vogel Hugh Watson Rolf Wigand
University of Melbourne City University of Hong University of Georgia Syracuse University
Australia Kong, China

ADMINISTRATIVE PERSONNEL
Eph McLean Samantha Spears Reagan Ramsower
AIS, Executive Director Subscriptions Manager Publisher, CAIS
Georgia State University Georgia State University Baylor University