Knowledge Sharing in an Emerging Network of Practice

European Management Journal Vol. 23, No. 3, pp. 300–314, 2005
Ó 2005 Elsevier Ltd. All rights reserved.
Printed in Great Britain
doi:10.1016/j.emj.2005.04.008 0263-2373 $30.00

Knowledge Sharing in
an Emerging Network
of Practice:
The Role of a Knowledge
Portal
PETER VAN BAALEN, Erasmus University, Rotterdam
JACQUELINE BLOEMHOF-RUWAARD, Erasmus University, Rotterdam
ERIC VAN HECK, Erasmus University, Rotterdam

This article addresses the emergence of networks of and on the emergence of a network of practice.
practice and the role of knowledge sharing via The results show that pre-conditions for the emer-
knowledge portals. Its focus is on factors that stim- gence of a network of practice are a sense of urgency
ulate the successful emergence of networks of prac- and fragmented awareness. These results also indi-
tice. Literature on knowledge management and cate the important role of a knowledge broker. The
communities of practice suggest the pre-existence developed knowledge portal seems to lead to over-
of shared knowledge or a shared belief system as coming structural holes and a closer cognitive dis-
a condition sine qua non for the networks of practice tance among the projects. However, we did not
to emerge. We challenge this assumption and argue ﬁnd a direct effect of the knowledge portal on shar-
and demonstrate that common knowledge and ing tacit knowledge. In the initial phase of a
belief systems are rather a result of knowledge shar- network of practice the knowledge exchange seems
ing rather than a pre-condition. The central ques- to focus on general, non-project speciﬁc and explicit
tion is how a knowledge portal facilitates the knowledge. There was also no direct effect of the
diffusion of knowledge among rather loosely cou- knowledge portal on the reciprocity of knowledge
pled and often disconnected innovation projects. exchange among the projects. However, knowledge
Research is carried out in the agricultural industry was shared between the project level and the plat-
in The Netherlands. In this industry there is a need form and public level. Conclusions and directions
to change from a product-oriented to a problem-ori- for future research are formulated.
ented innovation structure. The set up of a platform Ó 2005 Elsevier Ltd. All rights reserved.
and knowledge portal around agro-logistics – cross-
ing different product-oriented production clusters – Keywords: Agro-logistics, Innovation projects,
was therefore a logical result. It gave the opportunity Knowledge portal, Knowledge sharing, Network
to analyze what the impact of a knowledge portal is of practice, Social networks
in a situation where people and projects come from
different organizations and do not know each other.
Do they start to share knowledge and what are the
conditions? With regard to the case study of the Introduction
knowledge portal in the agricultural industry we
conclude that a knowledge portal will have an The diffusion of innovative knowledge is considered
impact on how projects are sharing knowledge to be one of the main challenges in the emerging

300 European Management Journal Vol. 23, No. 3, pp. 300–314, June 2005

KNOWLEDGE SHARING IN AN EMERGING NETWORK OF PRACTICE

knowledge society. As this innovative knowledge is Research was carried out in the agricultural industry
distributed and fragmented, Internet-based informa- in The Netherlands in particular the set up of innova-
tion and communication technologies can help to tion projects around themes related to agro-logistics,
leverage the knowledge diffusion. These technolo- see Ministries of LNV and V&W (2001). Agro-logis-
gies can easily connect distributed and loosely cou- tics deals with the logistics e.g. transportation, stor-
pled ‘pockets of innovation’ and diffuse relevant age, and distribution of agricultural products. The
information at high speed and at relatively low costs, answer to the above question was sought in a case
see Tuomi (2002). study approach. The case study provides a basis
upon which theoretical propositions are formulated
For this reason a platform of representatives of gov- and generalized (so called analytic generalization),
ernment, industry, and knowledge institutes in The see Yin (2003). The choice of the case setting made
Netherlands, the so-called Platform Agro-logistics, ini- it possible to analyze how a network emerges and
tiated the setting up of a knowledge portal in order how people and groups – that did not know each
to facilitate and speed up the diffusion of innovative other – started to share knowledge. The case study
knowledge in the agricultural industry. The set-up of let us closely track the design and use of a knowledge
this knowledge portal in the Dutch agricultural portal that could facilitate knowledge sharing among
industry should be considered as an innovation it- different innovation projects.
self. For many years this industry was characterized
by a closed and hierarchical knowledge infrastruc- This article is divided into three main sections. First,
ture in which the government dictated the research a literature review of knowledge sharing in networks
themes to the agricultural knowledge institutes. The and the role of knowledge portals is developed into a
research results were disseminated and communi- conceptual framework, complemented with six prop-
cated to the agricultural companies who were ex- ositions. Second, the empirical setting in the agricul-
pected to apply this new knowledge in practice. tural industry with research method and data will be
But recent disasters such as the outbreak of animal explained. Third, an empirical analysis of the case of
diseases such as BSE showed the limits of this ap- the knowledge portal in the agricultural industry
proach and new ways of innovations were explored. will be presented. Lessons learned, conclusions,
and suggestions for further research are formulated.
In this paper we consider the diffusion of innovative
knowledge as a form of collective action that requires
social (collective) organization. It implies that the
knowledge diffusion is viewed as an interactive pro- Literature Review and Conceptual
cess including the involvement of different collective Framework
actors.
Knowledge Sharing
The research question we address here is how a
knowledge portal facilitates the diffusion of knowl- The diffusion of innovative knowledge has become
edge among rather loosely coupled and often discon- one of the major research interests in management
nected innovation projects. Although the knowledge science and economics. A huge body of literature fo-
portal can easily connect these disconnected projects cuses on innovation as a ‘‘thing’’ about which infor-
and thereby facilitate knowledge diffusion we will mation needs to be provided to potential adopters
argue that a minimal social organization is needed and users in order to implement this innovation suc-
to initiate this diffusion process. Literature on knowl- cessfully (Swan et al., 1999: 262). As knowledge has
edge management and communities of practice sug- become to be seen as an innovation in itself new, crit-
gest the pre-existence of shared knowledge or a ical questions arise how to define knowledge and
shared believe system as a condition sine qua non how innovative knowledge can be diffused. Since
for the networks of practice to emerge, see for exam- the former question has been discussed extensively
ple Cohen and Levinthal (1990), Nonaka (1994), and in the recent management literatures it suffices to
Grant (1996). We challenge this assumption and discuss it briefly here. Since the publication of Non-
argue and demonstrate that common knowledge aka’s seminal paper ‘‘A dynamic theory of organiza-
and common belief systems are rather the result of tional knowledge creation’’ the complex distinction
knowledge sharing instead of a pre-condition. The between explicit and tacit knowledge has been
aim of this article is twofold. The first objective is widely accepted (Nonaka, 1994). The issue is not if
to conceptually describe the emergence of a network there exists such a distinction but how to understand
of people and groups that do not share knowledge the complex relationship between explicit and tacit
and beliefs at the initial situation. The second objec- knowledge. Roughly, two different views can be dis-
tive is to empirically show how this network emerges tinguished in this debate: the ‘near tangible view’
and evolves and what factors contribute to the suc- and the distributed view on knowledge (Tsoukas,
cessful emergence. It implies that we do not assume 2003). In the former view it is assumed that explicit
the existence of a particular form of a social network and tacit knowledge can be converted to each other
(e.g. community of practice) in advance, but will (Nonaka, 1994; Nonaka and Takeuchi, 1995). This view
view this as the outcome of network evolution. suggests that knowledge, by means of articulation,

European Management Journal Vol. 23, No. 3, pp. 300–314, June 2005 301


can be called upon for use in reasoning and can Krogh, 2003: 373). This mutuality in the knowledge
be translated into language and other media (Wino- transfer suggests that the process can be construed
grad and Flores, 1986: 73). In the distributed view it as a sequence of collective action in which the source
is believed that tacit knowledge is a component of and the recipient are involved (Von Krogh, 2003: 373).
all knowledge and as such cannot be converted into For this reason we will use the term knowledge sharing,
explicit knowledge. Tacit knowledge is not interna- instead of diffusion and transfer, as it succinctly re-
lised explicit knowledge, nor is explicit knowledge fers to the social processes that are involved. Sharing
externalised tacit knowledge. In viewing ideas as ob- knowledge is not giving a full representative account
jects that can be extracted from people and transmit- of what is known by the source about a particular
ted to others over a conduit, Nonaka and Takeuchi practice to the recipient. Because of the tacit compo-
reduce practical knowledge to technical knowledge. nent, knowledge contains an ineffable element; it is
According to Tsoukas (2003) tacit and explicit knowl- based on an act of personal insight that is essentially
edge are complementary, in the sense that explicit inarticulable. Tsoukas argues that this does not mean
knowledge is always grounded on a tacit component that we cannot share knowledge about a practice, but
and vice versa. Tsoukas further criticizes the notion of it should be viewed as re-punctuation of distinctions
knowledge as a given or something that is to be dis- underlying the practice, as drawing attention to
covered. The organization is a distributed knowledge unnoticed aspects and as making people aware of
system and cannot be surveyed as a whole; it is lack- new connections (Tsoukas, 2003). The stickiness of
ing an ‘‘overseeing mind’’. Similarly, Winograd and knowledge sharing does not only refer to the episte-
Flores argue that articulation of the unspoken is a mological but also to the relational problems. Accord-
never-ending process, as we must do it in a language ing to Szulanski (2003) people on the source side may
and a background that itself reflects a pre-under- be reluctant to share their knowledge with others for
standing. ‘‘Knowledge’’, as they put it, ‘‘is always a fear of losing ownership, a position of privilege, supe-
result of the interpreter, which depends on the entire riority, for the lack of insufficient rewards, for lacking
previous situation and on its position in a tradition time to communicate about an innovative practice.
(1986: 75). Thus knowledge has an important tacit Another reason can be that people are unaware of
component, which resides in individual skills, under- the fact that their knowledge might be of interest to
standing, collaborative social arrangements, but also others. On the recipient side important factors like
in tools, documents, and processes that embody as- the reluctance to accept new knowledge from an
pects of knowledge (Wenger et al., 2002: 11). As these external source (‘not invented here’-syndrome), the
skills and social arrangements are related to work inability to exploit outside sources of knowledge
activities we will call them practices (Szulanski, (absorptive capacity), an inability to retain the newly
2003). This view contrasts the ‘near tangible view’ acquired knowledge in the organization, increase the
as it suggests that any form of explicit knowledge as- stickiness of knowledge sharing.
sumes the existence of tacit knowledge that cannot be
articulated. As a consequence, the transfer of innova-
tive knowledge from one practice to another will be- Emergence of Networks of Practice
come problematic. Disembedding knowledge from
one practice and re-embedding this knowledge into The sharing of knowledge requires social organiza-
another practice does not go without any costs. tion and governance. Traditional organizational
Von Hippel has coined the concept of ‘‘stickiness’’ forms (markets and hierarchies) show serious deficits
of knowledge to refer to the incremental costs to in organizing the complex nature of knowledge
transfer knowledge from one practice to another (Jones et al., 1997). For this reason new organizational
(Von Hippel, 1994: 430, see also Szulanski, 2003). forms are introduced to deal effectively with the
When transfer costs are low, knowledge stickiness sharing of explicit and implicit knowledge. The com-
is low; when it is high, knowledge stickiness is high. munity of practice concept, introduced by Lave and
Both Von Hippel (1994) and Szulanski (2003) point to Wenger (1991) and transferred to the management
the fact that the stickiness of knowledge involves not domain by Brown and Duguid (2000), represents
only the complex epistemology of knowledge, but probably one of the potentially most useful and
also attributes of the knowledge source, the knowl- enduring concepts in this respect. Most definitions
edge recipient, and of the context. When the knowl- of communities of practice (CoP) stress the impor-
edge source and the knowledge recipient share the tance of shared practice, repertoire, interests, knowl-
same context and are engaged in the same practice, edge, on informality, and on the self-organizing
the stickiness will be relatively low, whereas the character of the community. Recently, Brown and
transfer cost will increase when the knowledge Duguid (2000) have distinguished two types of net-
source and the knowledge recipient operate in differ- works, networks of practice (NoPs) and communities
ent contexts and are engaged in different practices. of practice (CoPs). In the former, people have prac-
tice and knowledge in common but are mostly un-
Knowledge transfer within and between organisa- known to each other. The links between the
tions is not a one-way activity, but a process of trial networks are mostly indirect (e.g. databases, newslet-
and error, feedback, and mutual adjustment of both ters, info bulletins) and members coordinate and
the source and the recipient of knowledge (Von communicate normally explicit. NoPs can have an



enormous reach. There is relatively little reciprocity these networks emerge and evolve. In their study
across NoPs as the members do not interact directly. of CoPs Wenger et al. (2002) made a first attempt to
NoPs are loosely coupled systems that hardly initiate sketch the evolution of CoPs by identifying five
collective action and produce little knowledge. stages of community development. According to
the authors, CoPs typically start as loose networks
CoPs on the other hand represent relatively tight-knit that hold the potential of becoming more connected
groups of people who know each other well and and develop towards a tightly-knit community.
work together directly. Online communication is However, loose connectedness presumes the exis-
often supported by face-to-face interactions, which tence of particular ties between the members of a po-
enable them to coordinate and communicate to a tential network. This might make sense within the
high degree on implicit knowledge. Due to these context of one organization or CoPs where homoge-
face-to-face relationships the communication reach neity of interests and knowledge can be presumed.
is bounded. CoPs are characterized by strong reci- Our question, however, focuses on the emergence
procity norms which help to sustain the community. of those initial ties between actors that come from
different organizations and who do not or hardly
Although the distinction between CoPs and NoPs know each other. Many authors state that the coordi-
seems to be clear at the surface level, it is hard to nation and sharing of knowledge cannot take place
determine precisely in advance if the social collective without assuming a vast amount of mutual knowl-
should be conceived as a CoP or a NoP. We suggest edge, mutual beliefs, and mutual assumptions. This
that both, CoP and NoP, are particular forms and is what is called common ground. Similarly, Grant
therefore suggest taking the social network as the (1996: 115) argues that common knowledge (lan-
starting point for our analysis and conceive CoPs guage, other forms of symbolic communications,
and NoPs as particular forms of social networks. A shared meanings, commonality of specialized knowl-
social network can be defined as a patterned organi- edge etc.), defined as the intersection of individual
zation of a collection of actors and their relationships knowledge sets, should be conceived as a precondi-
(Jones et al., 1997). It is impor- tion for the knowledge inte-
tant to note that in this minimal gration. Cohen and Levinthal
definition no specifications are The collection of actors (1990) point to the importance
given about the nature of the of overlapping knowledge in
actors and their relationships. should contain more than organizations in order to
According to Wellman and assimilate external knowledge,
Gulia (1999) this implies that two... to be defined as a whereas Nonaka (1994) views
even when people are only redundancy of knowledge as
connected through a computer network a necessary precondition for
network, they should be con- knowledge creation and the
ceived as a social network. building of trust. Nooteboom
We don’t agree with this minimal definition because (2000) has coined the concept of cognitive distance
if no interaction takes place one cannot speak of a so- and cognitive proximity to refer to cognitive close-
cial network. The collection of actors should contain ness and similarity between people. It does not only
more than two actors to be defined as a network. Tri- refer to the cognitive variety but also to the differ-
adic relationships differ fundamentally from dyadic ences in abilities of perception, interpretation, and
relationships because in the former 1) individuality different views on the world that develop out active
is reduced; 2) the individual power is reduced; 3) interaction with the physical and social environment
and conflicts are moderated by the presence of a third (Nooteboom, 2000). Cognitive distance yields both a
party. We can add to this definition two other charac- problem and an opportunity (Nooteboom and
teristics (Podolny and Page, 1998). The first is that the Bogenrieder, 2003). When the cognitive distance is
collection of actors pursue repeated, enduring ex- too short or is absent for people to share the same
change relations with one another. If exchanges are knowledge, there is no incentive to share knowledge.
not enduring but episodic - engaging in an incidental This might be the case when people interact fre-
transfer of goods, services or information - there is no quently and consequently establish strong networks
social network but a market situation. The second is like CoPs. However when cognitive distances are too
that social networks lack a legitimate organizational great, the more difficult it becomes to cross the dis-
authority to arbitrate and resolve disputes that may tance, i.e. to mutually understand the actions and
arise during the exchange (as is the case in hierar- expression in the network. The notion of cognitive
chies). Based on these characteristics a network can distance is relevant here as it succinctly points to
be viewed as a social exchange structure with its the potential for a network to emerge. The question
own governance structure and patterns of interaction however is still how much cognitive similarity is
in which flows of resources between independent needed to initiate knowledge sharing. In order
units (or individuals) take place. to find out what is minimally needed to initiate
enduring interaction we briefly discuss Weick’s
Most research on social networks focuses on existing (1979) theory on the emergence of collective struc-
social structures. Less attention is paid to the way ture. Weick (1979) argues that people initially don’t



have to agree on goals to act collectively. In any po- mies of online cooperation Kollock (1999) points to
tential collective, people have different interests, the limitations of online cooperation and collective
preferences etc. and want to accomplish different action. Although it is quite easy to produce and share
things. In order to achieve these ends they have to digital information, it requires coordinated activities
initiate action towards others by which they create from the beginning. Another weakness is that if ac-
mutual commitment (interlocked behaviours) to col- tive knowledge producers withdraw from the online
lectively pursue diverse ends through common network, the network will cease to exist. The basic
means. Once people are engaged in mutual commit- features for an online network are: ongoing interac-
ments a subtle shift takes place from diverse to com- tion, identity persistence, and knowledge of the pre-
mon ends. As Weick argues, diverse ends remain, vious interactions. (Kollock, 1999: 235). The notion of
but they become subordinated to an emerging set online generalized exchange demonstrates how frag-
of shared ends. This part of Weick’s evolutionary ile the minimal social situation of emergent social
theory contrasts conventional thinking about the pre- network is.
conditions for the emergence of collective structures.
The second part of Weick’s theory addresses the
question of how coordination can take place even Design of Knowledge Portals
though ties between people are minimal. To this
end he discusses the notion of a mutual equivalence The question then is what the design and the man-
structure (MES). The MES is like an implicit contract agement of an interactive information and communi-
between people that can be built and sustained cation system should look like for the online
without knowing the motives of another, and with- (generalized) exchange of knowledge in a minimal
out people having to share goals. Weick points to social situation. The agro-logistic projects are geo-
three preconditions for an MES to emerge. The first graphically dispersed and the participants hardly
is that a person must perceive that his ability to per- know each other and have rarely communicated. In
form his consummatory act depends on the instru- the literature three dominant perspectives on the role
mental act of the other. The second is that a person of information and communication technologies on
must perceive that his own instrumental act serves knowledge sharing are distinguished: deterministic
to elicit the instrumental act of the other. Third, a view on technology, medium choice theory, emer-
MES only emerges when a person repeats his instru- gent process perspective. We believe that the three
mental act. A fourth precondition can be added, that perspectives are not mutually exclusive. It is now
is that a person must have some knowledge (expecta- widely accepted that user involvement is essential
tions) about another person that can fulfil his instru- in the process of design and implementation. How-
mental acts. From this perspective we may conclude ever the user is just one node in the potential
that the common ground, needed to initiate a MES, network of knowledge sharing. It is not only impor-
does not refer to common knowledge or a common tant to know if user A prefers to use a particular
belief system but from a mutual expectation application of the intranet for the transfer of a partic-
structure. ular type of knowledge but also if persons B and C
have similar preferences and expectations. The impli-
The issue of reciprocity in online networks is widely cation of our argument here is that the design of an
debated among researchers. Especially the motiva- interactive information system should not reflect
tion for people to contribute to an online connected the needs of individual users but the social structure
group of people who do not or hardly know each (triadic relationships) of the emerging network. We
other have been subject to extensive research. Well- therefore call for a relational and rich information sys-
man and Gulia (1999) point to different types of tem design. That design will have the following three
explanations. The first refers to the fact that online characteristics:
contributions are a means of expressing one’s iden-
tity. Helping others might increase self-esteem, repu- v It should be relational as it should not only
tation, respect from others etc. The second one is address the needs of individual users but also
generalized reciprocity and organizational citizen- the triadic expectations of a potential social net-
ship. In their recent research on three online commu- work. It means that in the initial stage of the
nities McLure Wasko and Faraj (2000) concluded that development of the knowledge portal the poten-
sharing knowledge and helping others is ‘the right tial network actors should be informed about
thing to do’ and that people also have a desire to ad- these expectations (e.g. by organizing meetings,
vance the community as a whole. Participants did providing information about the projects, adver-
not expect to be reciprocated by the same person tising, see Damsgaard, 2002).
with whom they shared knowledge (direct exchange) v It should also be rich as it is impossible to predict
but did expect to receive future help from someone how the actors in the network will communicate.
in the network. Wellman and Gulia (1999) argue that Social networks are complex social systems that
the logistic and social costs involved in online contri- cannot be simply founded. They develop and
bution are relatively low. The easy access to online transform over time (Wenger et al., 2002). To
social networks allows and enables people to contrib- allow the online social network to take different
ute at low participation costs. Discussing the econo- shapes and to evolve in different directions the



design of the knowledge portal should dynami- Conceptual Framework
cally match different social profiles of the net-
work. It implies that the knowledge portal The following conceptual framework depicts the role
should provide different spaces of knowledge and impact of knowledge portals and how networks
sharing, synchronous and asynchronous commu- of practice emerge, see Figure 1. Based on the litera-
nications media, document storage and retrieval ture review we formulate the following propositions:
etc.
v Perhaps the most important requirement for the As discussed by Brown and Duguid (2000), in net-
design in the minimal social situation is that works of practice people have practice and knowl-
people converge on the means, in our case the edge in common but are mostly unknown to each
knowledge portal. Following Weick’s theory on other. The links between the networks are mostly
the emergence of collective structures we con- indirect (e.g. databases, newsletters, info bulletins)
sider the knowledge portal as a means to facilitate and members coordinate and communicate and
the sharing of knowledge between different these are normally explicit. There are two factors that
groups. seem to be a pre-condition for the emergence of net-
works of practice: sense of urgency and fragmented
Markus et al. (2002) argue that traditional informa- awareness.
tion system design theories are badly equipped to
deal with emergent knowledge processes (EKPs). Proposition 1. A higher sense of urgency to tackle
EKPs are defined as organizational activity patterns specific problems of practice will lead to the emer-
that exhibit three characteristics in combination: gence of a network of practice.
deliberations with no best structure or sequence;
highly unpredictable potential users and work con- People are tackling specific problems of practice and
texts; and information requirements that include one way to do that is to coordinate and communi-
general, specific, and tacit knowledge distributed cate. However, there has to be a high sense of ur-
across experts and non-experts. We believe that the gency that people will coordinate and communicate
set up of a knowledge portal in agro-logistics in with people they hardly know (and also from other
order to facilitate the diffusion of knowledge be- organizations). Without that level of urgency ‘‘out
tween the distributed projects fits pretty well the sit- of the box’’ thinking seems a strategy people will
uation as described by Markus et al. (2002). However, not follow.
the portal is only the ‘front door’ of an intranet or an
extranet (Chaffey and Wood, 2004). Intranets and Proposition 2. Fragmented awareness in a dispersed
extranets are called decentralized, general purpose- industry will lead to the emergence of a network of
and open-ended technologies which mean they can practice.
be designed for different purposes and can poten-
tially be constructed and modified by those who A second pre-condition is that there has to be a frag-
are involved in the design and use of these informa- mented awareness in a dispersed industry. People
tion systems (Damsgaard, 2002). need to have the expectation that somewhere out

Sense of Fragmented
Urgency Awareness

Active P3B P1 P2
Knowledge
P3A
Broker

Knowledge P4A Overcoming P4B Emergence of
Portal Structural Holes Network of Practice

P5A
P5B

Type of
Knowledge
P6A P6B

Type of
Knowledge Sharing

Figure 1 The Conceptual Framework



there solutions are available. They know what they As Kollock (1999) argues, the generalized exchange
don’t know and they know that somebody else might system of sharing is both more generous and riskier.
know (about similar problems and potential solu- It is more generous because the person who gives
tions). Without that fragmented awareness there provides the network with a benefit without the
seems to be no logical reason to strengthen ties expectation of immediate return. However general-
among people. ized exchange is also more risky because actors are
easily temped to a free ride (taking without contrib-
Next to these two basic factors we think there are uting). However, the basic features for an online net-
more specific factors that stimulate or hamper the work are: ongoing interaction, identity persistence,
emergence of Networks of Practice (NoPs). The fol- and knowledge of the previous interactions. (Kol-
lowing factors will be taken into account: action by lock, 1999: 235). Therefore a knowledge portal has
broker, role structural holes, type of knowledge, to lead to a certain level of reciprocity in knowledge
and type of knowledge sharing. sharing to sustain the emergence of a network of
practice.
Proposition 3. An active knowledge broker will lead
to the development of a knowledge portal and the
emergence of a network of practice.
Research Methods and Data
The links in a network are mostly indirect. Therefore
in the initial phase there has to be an active broker Case Study Background
bringing people together who did not know each
other before. Previous research has shown that an ac- The role and impact of knowledge portals for the
tive broker (in a coordinated or spontaneous way) emergence of networks of practice is illustrated
helps to create the indirect linkages among members here by a case study of a knowledge portal for
of an emerging network. The role of a knowledge agro-logistic innovation projects in The Nether-
broker is identified by, for example, Davenport and lands. For a detailed description of case study re-
Prusak (1998). search, see Yin (2003). Agro-logistics deals with
the transport, storage, and distribution of the agri-
cultural flows of food and non-food goods in the
Proposition 4. A knowledge portal will bridge struc-
entire supply chain. Agro-logistics is an important
tural holes and contribute to the emergence of a
sector. In The Netherlands, more than 20% of good
network of practice.
transportation (including import and export) in-
cludes agro products. The agribusiness has recently
Potential knowledge portals have the ability to create
dealt with a number of bottlenecks such as animal
direct linkages (between the portal and the
diseases leading to trade embargos, congestion on
knowledge sender/receiver) in such a way that direct
the Dutch highways, international competition,
linkages between the sender and receiver are not nec-
and stronger legislation regarding food safety and
essary. In such a case structural holes are overcome
animal well-being. Recently, a number of
(Burt, 1992). As we have seen there is a paradox in
developments in society have taken place, influenc-
the sense that overcoming structural holes will lead
ing the management of agro-logistic flows. These
to effective knowledge exchange because as Noote-
developments are: higher consumer awareness, pull
boom and Bogenrieder (2003) indicated cognitive dis-
strategy (market) instead of push strategy
tance yields both a problem and an opportunity. It is
(producers), fragmentation, scaling-up in retail
still unclear how much cognitive similarity is needed
and agro-distribution, globalisation and liberalisa-
to initiate knowledge sharing.
tion, sustainable entrepreneurship, sharpened legis-
lation, and more attention to tracing and food
Proposition 5. A knowledge portal will lead to the safety.
exchange of project-domain knowledge and there-
fore contribute to the emergence of a network of The agricultural community has a product-related
practice. cluster structure. This can be illustrated by the
names of the Product Boards (regulatory organiza-
A knowledge portal will make it easier and less costly tions for businesses in the agricultural supply
to transfer and exchange knowledge. However, as we chains): Animal Feed, Beer and Wine, Cattle, Meat
have seen, related to the stickiness of knowledge both and Eggs, Dairy, Farming, Grains and Seeds, Horti-
Von Hippel (1994) and Szulanski (2003) indicate the culture. These sectors are highly independent of
transfer cost will increase when the knowledge each other with weak ties between each other. They
source and the knowledge recipient operate in differ- often are called the Pillars of Agriculture. Within
ent contexts and are engaged in different practices. these pillars, knowledge is available and people
have regular contacts with each other. Between
Proposition 6. A knowledge portal will lead to reci- the pillars, the information sharing and communi-
procity in knowledge sharing and therefore contrib- cation is quite low. Recent developments and bot-
ute to the emergence of a network of practice. tlenecks encouraged the community to change



from a product-related structure towards a prob- these interviews was to answer the following
lem-related one. These problems occur in the area questions:
of spatial planning, EU legislation, high scale infra-
structure, and optimizing logistic networks. In or- v Do the selected innovative projects have a need
der to develop a vision on the sustainable for a knowledge portal?
coherent future of the agro-logistics sector, a plat- v Is there a need for specific knowledge and infor-
form of representatives of government, industry, mation (both in content as in type of knowledge/
and knowledge institutes was set up, the so-called information)?
Platform Agro-logistics (Ministries of LNV and v Is there a willingness to share?
V&W, 2001). The Vision Agro-logistics aims to reach v If there is a need, what is the main design of the
a sustainable, innovating and transport-efficient sec- portal, and what are critical success factors for
tor and is based on three keywords, i.e. Clustering, the design, building, and implementation of the
Binding, and Directing. The national government, knowledge portal?
cooperating with the Platform, invited the sector
to propose innovative projects in the area of agro- The interview results showed there was a high sense of
logistics to improve sustainable development. The urgency to tackle specific problems of practice exists.
innovative character can be related to Clustering About 80% of the project leaders stated a need to ex-
(realisation of large scale agri-business areas), Bind- change knowledge by means of a knowledge portal.
ing (innovative logistical concepts) and Directing The knowledge portal was also seen as an effective
(towards virtual livestock markets or worldwide medium to reach the public in general, creating a basis
plant cultivation networks). The Platform Agro- for innovative projects, making projects known to the
logistics focuses on coordinating between parties, public, and finding new partners. The interviews also
tuning with governmental organizations, and creat- show that knowledge and information need is very
ing support. The goals of the platform are (i) to ad- diverse between projects. It is unusual that these
vise, cooperate and coach pilot projects to succeed groups are not divided by the central Platform themes
in system innovations and (ii) to attract attention Clustering, Binding and Directing, neither by the pil-
and share information on threats and opportunities lars of the Agro Sector. The Projects can be divided
in the agro-logistics sector. In 2003, 20 innovative as follows:
pilot projects were selected by the Dutch
government to be supported in their development v Entrepreneurs who want to share knowledge on
by the Platform Agro-logistics. The project propos- legislation, best practices, subsidiaries, and
als came from almost all pillars in the agricultural lobbying;
industry like vegetable products, chicken, plant v Knowledge institutes with no need for a social
cultivation, cheese, and pig farming. The innova- knowledge network, interested in European
tions are not essentially product based, but had to subsidiaries;
fit into the themes of Clustering, Binding, and v Umbrella projects with a need to share knowledge
Directing. in the field of project management and regional
scaling-up.

Knowledge Portal: Stages and Data All interviewees showed willingness to share infor-
mation with each other and with the public space.
It is important that knowledge on how to innovate With respect to the design of the knowledge portal,
and the innovation itself are exchanged in an efficient a layered structure was suggested in such a way
and effective way among the projects and potential that it reflected the current community structure,
new projects. The Platform Agro-logistics suggested i.e. a project level, a platform level, and a public le-
setting up a (virtual) place to meet each other, share vel. Each level gives entrance to specific types of
information and knowledge, deal with governmental information. Being present in the public space
and policy issues, and seek financial resources, in was one of the priorities of the project leaders
other words, to be a network of practice. A knowl- (visibility).
edge portal, a platform based on Internet-technology,
can support the forming of a network or community.
The knowledge portal should open the door to inno- In the second stage (July 2003–August 2003) the
vative knowledge in the various pilot projects, knowledge portal was designed, built, and tested. It
regardless of time, place, and existing relations of was decided to structure the knowledge portal in
knowledge exchange. three levels. The first level deals with the innovation
projects, the second level with the platform, the third
The development of a knowledge portal took place level with the public space. At the project level
in three stages. In the first stage (March 2003–June knowledge sharing among the members of the pro-
2003) objectives, requirements, and design rules ject is facilitated. The members of a project share a
were determined. Structured interviews were held common practice in which knowledge primarily re-
with all project leaders of the different innovation lated to the project is developed and shared. Rela-
projects, see Van Baalen et al. (2003). The aim of tions are direct and tightly coupled and implicit



600 571

500 465 475
number of visitors 448 440

390
400
331 341
312
296
300 277

218

200

100

0
sep oct nov dec jan feb mar apr may jun jul aug
2003 2003 2003 2003 2004 2004 2004 2004 2004 2004 2004 2004

months

Figure 2 Overview of the Number of Visitors Since the Start of the Knowledge Portal

knowledge sharing (learning by doing) possible. At Analysis of Empirical Data
the platform level knowledge can be shared among
the different projects and the members of the Plat- Knowledge Portal Statistics
form Agro-logistics. The platform level is only acces-
sible by the members of the different projects and the The knowledge portal (www.agrologistiek.nl) went
platform. The knowledge exchange at this level is live in September 2003. For one year, we analyzed
worthwhile for the community as a whole, as it gives the knowledge portal statistics. In one year the total
a base for sharing experiences and best practices number of hits was about 275,000. The results indi-
among the sectors and therefore from moving from cate the knowledge portal had between 15,000 and
a product-related innovation structure towards a 20,000 hits per month in the period December
problem-related structure. Finally, at the public level 2003–May 2004. In the summer of 2004 the number
knowledge can be exchanged between the innova- of hits increased. Figure 2 shows the monthly num-
tion projects and the actors outside (public, innova- ber of unique visitors to the knowledge portal.
tion projects outside agro-logistics, other
industries). Here information is available for every- The data of the number of visitors show there was a
body and free of charge. steady increase in visitor numbers due to the fact that
the public was more aware of the existence of the
In the third stage (September 2003–until now) the knowledge portal. The total number of visitors to
knowledge portal was used. A web master was tak- public level varied from 218 in the first month to
ing care of the functioning of the portal and of the 571 one year later. Figure 2 shows that the number
instruction of users. In that period we were able to of visitors grew steadily over the year, with two
monitor the use of the knowledge portal and there- exceptions: the second month (October) had a rela-
fore could analyze who was using it and how it tive high number of visitors, caused by the novelty
was used. The use of the knowledge portal in the first of the site and the month of May had a relative low
year (September 2003–August 2004) will be pre- number of visitors, probably due to the Spring holi-
sented in this article. Detailed statistics were avail- days. The number of visits suggest there is a need
able on the profile of visitors, the amount of hits for information sharing within and outside the agri-
and page views, and details about visitor sessions. cultural community.
A visit is defined as a hit originating from the same
IP-address with a maximum time between the hits In one year more than 7,500 documents were down-
of 20 minutes. A monitoring tool was developed loaded. Table 1 presents the top 10 downloaded files.
and linked to a social network analysis software pro- Original titles of the documents were in Dutch.
gram called UCINET 5 (Borgatti et al., 2004). With the
help of this program a general analysis of relation- From this list, we conclude that the need for informa-
ships among the projects in the emerging network tion focuses on the existence of the Platform Agro-
of practice could be identified. For a more thorough logistics and its vision (1, 6, 8, 9), and less on the
explanation of the research methods and techniques content and urgency of the innovative projects. Only
used, we refer to Van Baalen et al. (2003). one of the themes of the Platform Agro-logistics is in



Table 1 Top 10 of Downloaded Files from Septem- receivers. Providers and receivers are innovation
ber 2003 to September 2004 projects. In total there were 25 innovation projects
indicated by P01, P02, . . ., P23, P26, and P27. P24
Document Number of
and P25 were projects for general and project man-
downloads
agement purposes. Not all projects come back in
1 Brochure_Platform_Agro-logistics.pdf 2786 Table 2 due to the fact that these projects were
2 Agro-Logistic Invitation Letter.pdf 1252 stopped or merged with others. The values in the
3 Final report_Agro-chains 960 matrix represent the number of times that, for exam-
and Clusters.pdf ple, project 01 downloads information from project
4 Agro-logistics_ Examples.MPG 697 02 in the Platform or Project space. These spaces
5 Pre-announcement_Agro-logistics.pdf 600 were restricted to members only and can be visited
6 Vision_Agro-Logistics.pdf (Platform) 514 through a login name and a password.
7 Bundling of Agro-streams.pdf 476
(Clustering) Table 2 shows mixed results: some projects were
8 Conference_registration_Aro-logistics.pdf 465 active providers of documents, some projects were
(Platform) active consumers of documents, and some pro-
9 Letter_MinistersLNVandVenWto 316 jects were not very active at all in sharing documents.
Second Chamber.pdf (Platform) A more detailed analysis of knowledge sharing
10 Examples Agri-parcs.pdf (Clustering) 278
among projects will be discussed in the next section.

the top 10, namely Clustering logistics flows in Agri- Lessons Learned
business centers (3, 7, 10).
Based on interviews of the project leaders, the design
Apart from a statistical analysis at the Public level, it and use of the knowledge portal, the knowledge por-
was possible to monitor the communication between tal statistics, and monitoring of the document ex-
projects themselves, at the Platform and Project level. change among the projects the proposition – as
Some functionalities were hardly or never used. For defined in Section 2 – were validated. A network of
example, the discussion forum was hardly ever used. practice will emerge if there is a sense of urgency
Also, the use of the bulletin board was negligible. (Proposition 1) and fragmented awareness in a dis-
These features were pointed out as potentially useful persed community (Proposition 2). Proposition 3
features in the interviews. The feature of sharing states that active action of a broker will lead to the
information by downloading documents was fre- development of a knowledge portal. The broker in
quently used, as indicated by Table 1. For each this case is the Platform Agro-logistics. For testing
downloaded document, data was available with re- Propositions 4–6, we carried out interviews with all
spect to the supplier of the document (providing innovative pilot projects (Van Baalen et al., 2003) and
information) and the client (receiving information). analyzed for one year (September 2003–August
Table 2 provides information on who shared docu- 2004) the actual use of the knowledge portal. The pur-
ments with whom. In Table 2 the providers are rep- pose of the interviews is to find data on the network
resented in the rows (between brackets the number situation before the start of the knowledge portal
of posted documents), the columns represent the (Proposition 4, Structural Holes), to investigate the

Table 2 Document Sharing among the Projects in the Knowledge Portal

Supply Demand

P01 P02 P03 P04 P11 P12 P13 P15 P17 P20 P21 P23 P25 P26 P27

P01(12) 127 0 0 0 0 0 0 0 0 0 0 0 0 0 0
P02(5) 0 0 0 0 2 0 0 0 0 1 0 0 1 0 0
P03(9) 1 0 12 0 0 0 0 0 0 2 0 0 1 0 0
P04(7) 1 0 0 1 0 0 0 0 0 6 0 0 3 0 0
P11(14) 9 0 0 0 12 0 0 0 0 0 0 0 4 0 0
P12(10) 4 0 4 0 4 0 0 0 0 15 0 0 18 0 0
P13(1) 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
P15(9) 0 0 0 0 0 0 0 0 0 3 0 0 1 0 0
P17(11) 0 0 0 0 0 0 0 0 1 6 0 0 2 0 0
P20(45) 0 0 3 0 0 0 0 0 0 559 0 0 2 0 0
P21(9) 2 0 0 0 0 0 0 0 0 3 1 0 2 0 0
P23(1) 0 0 0 0 0 0 0 0 0 8 0 0 0 0 0
P25(6) 1 0 0 0 0 0 0 0 0 16 0 0 141 0 0
P26(5) 0 0 0 0 0 0 0 0 0 13 0 0 21 0 0
P27(5) 0 0 0 0 0 0 0 0 0 6 0 0 0 0 0



need for knowledge and the type of knowledge Proposition 3. An active knowledge broker will lead
needed (Proposition 5, Types of knowledge), and to the development of a knowledge portal and the
the willingness to share knowledge (Proposition 6, emergence of a network of practice.
Reciprocity in knowledge sharing). The analysis pro-
vides the following lessons learned. The set up of the Platform Agro-logistics including
representatives of different stakeholders in the differ-
Proposition 1. A higher sense of urgency to tackle ent product-related communities together with repre-
specific problems of practice will lead to the emer- sentatives of local and national authorities encouraged
gence of a network of practice. a broking role in an emerging network of practice. In
particular, the chairman and secretary of the platform
In the former, people have practice and knowledge in acted as active brokers – they took the initiative to
common but are mostly unknown to each other. The develop a knowledge portal. In the initial phase of
links between the networks are mostly indirect (e.g. the knowledge portal most documents and initiatives
databases, newsletters, info bulletins) and members were posted by the web master. Proposition 3 is
coordinate and communicate in a normally explicit accepted.
fashion. NoPs can have an enormous reach. There is
relatively little reciprocity across NoPs as the mem- Proposition 4. A knowledge portal will bridge struc-
bers do not interact directly to one another. NoPs tural holes and contribute to the emergence of a
are loosely coupled systems that barely initiate collec- network of practice.
tive action and produce little knowledge.
To look at the impact of the knowledge portal in over-
The Agro-logistics case shows that there was on the coming structural holes and decreasing cognitive dis-
one hand a high sense of urgency in the agricultural tance among projects we examined which projects
sector to start the Platform Agro-logistics. Several were known to each other before the platform and
reasons were mentioned during the initial phase of knowledge portal were implemented. The different
the set up of the platform. These reasons were re- project leaders were interviewed and asked if they
lated to the outbreak of animal diseases, transporta- knew the other projects (Van Baalen et al., 2003).
tion problems, and more strict legislation for food Based on these interview results relationships among
safety. Also during the interviews with the project the different projects were measured and drawn
leaders at the start it turned out that most of them with the help of UCINET software (Borgatti et al.,
indicated an urgent need for collective action with 2004). On the left side of Figure 3 the initial network
regard to transport and distribution problems re- is sketched. As one can see, there are seven projects
lated to agro-products in The Netherlands. On the that have no relationship with other projects and some
other hand in analyzing objectives, incentives, and projects have very weak ties with others. None of the
the lack of direct subsidies for projects to participate projects regularly exchanged information and knowl-
in there was not a high level of urgency. Projects edge. After one year of using the knowledge portal
could not be pushed to deliver results in a fast we examined which documents were exchanged
way, subsidies were not directly given to projects, among the different projects. The right side of Figure
the platform was installed to facilitate the different 3 presents the network after the introduction and use
innovation projects. Overall, there was a shared le- of the knowledge portal. As Figure 3 indicates, pro-
vel of urgency to innovate to keep The Netherlands jects are exchanging documents and therefore could
competitive in the field of agro-logistics. Proposition learn from each other. One can see a network of prac-
1 is accepted. tice is emerging. The knowledge portal overcomes
some structural holes and there are indications that
Proposition 2. Fragmented awareness in a dispersed the cognitive distances between the actors are not
industry will lead to the emergence of a network of too short e.g. that there is no incentive to share knowl-
practice. edge. Proposition 4 is accepted.

The Agro-logistics case shows there was a frag- Proposition 5. A knowledge portal will lead to the
mented awareness in a dispersed agricultural indus- exchange of project-domain knowledge and there-
try. Traditionally, the agricultural community is fore contribute to the emergence of a network of
structured in a product-oriented way (meat, milk & practice.
cheese, flowers, fruit & vegetables). Agro-logistical
problems and solutions are endemic in these We distinguish three types of knowledge:
different product units. Therefore there was a need
for agro-logistical experts to learn from innovations v Type I – This type of knowledge is project-
in different product-oriented communities. In these domain knowledge and developed by one of
different communities there was awareness that the the innovation projects.
agro-logistics community is highly dispersed and a v Type II – This type of knowledge is platform-
lack of coordinated action was hampering solutions domain knowledge and developed by one of
to agro-logistical problems. Proposition 2 is accepted. the innovation projects.



Figure 3 Information Exchange Among Projects Before the Introduction (left side) and One Year After the
Introduction (right side) of the Knowledge Portal

v Type III – This type of knowledge is public- However, the empirical results indicate a revised
domain knowledge and developed by one of Proposition 5: In the initial phase of a network of
the innovation projects. practice knowledge exchange will be focused on gen-
eral, non-project specific and explicit knowledge. It is
As discussed, the knowledge portal was developed expected that after this phase the exchange will be
with three levels or spaces: project, platform and directed to project-specific and tacit knowledge.
public. By analyzing the use of the knowledge portal
(documents, web pages, bulletin board) we conclude Proposition 6. A knowledge portal will lead to
that information exchange among the different pro- reciprocity in knowledge sharing and therefore
jects (Type I) was rather limited. The analysis also contribute to the emergence of a network of
shows that some documents were exchanged at the practice.
platform level (Type II) and most documents were
exchanged at the public level (Type III). There seem Reciprocity in a network means that projects are post-
to be two potential explanations. The first one relates ing and demanding knowledge from other projects.
to the stickiness of knowledge – see Von Hippel This is different from market relationships where a
(1994) and Szulanski (2003). The different innovation specific activity (an indication to buy something) will
projects did not exchange because transfer costs were automatically lead to an offer by the other party. In a
too high due to the fact that the knowledge source network of practice one can post knowledge but one
and the knowledge recipient operate in different con- does not automatically and directly get something
texts and are engaged in different practices. The de- in return. However, in the longer term one expects
creased transfer costs of the knowledge portal did that if one posts a question to the network – or rather,
not overcome the high level of transfer costs related to a member, (in our case projects) – that it will react
to the stickiness of the knowledge. The second one with an offer. We analyzed the knowledge exchange
relates to the concept of cognitive distance – see among the projects and distinguished four type of
Nooteboom and Bogenrieder (2003). The cognitive projects. These four types are labeled:
distance between the innovation projects seems to
be too high and therefore it is more difficult to cross v Individualistic Projects: These projects do not
the distance between the projects. It seems to be log- post or demand information and knowledge.
ical that at the emergence of a network of practice v Altruistic Projects: These types of project post a
knowledge exchange will start with knowledge with lot, but make no demand.
low transaction costs and a low cognitive distance v Free rider Projects: These projects show no post-
(such as general project knowledge). The analysis ing, but demand a lot.
also indicates there is information exchange within v Reciprocity Projects: These project post and make
some of the projects. This can be considered as a demands.
new type of knowledge (knowledge exchange within
project). In particular, large projects (with around 50 It is interesting to analyze how projects developed
project members) were eager to exchange informa- during the use of the knowledge portal. Figure 4
tion within the project. Proposition 5 is not accepted. identifies the typology of the projects after one year



is also more risky because actors are easily tempted
P20
P11
to free ride (taking without contributing). For this
reason the generalized exchange has the structure
P01
of a social dilemma in which individually reasonable
P17
behavior might lead to collective disaster (Kollock,
P12
1999). In our case there was no free-rider behavior
P15,P21
P03 among the projects, but also no balanced knowledge
Altruïst Reciprocal supply and demand. In the longer term there is the
supply potential risk that the knowledge portal runs dry
P04 Individualist Free rider
P25 and that the network of practice will dissolve.
P02, P26, P27

Conclusions
P13, P23
Research Problem
demand
The central research question of this article is how a
Figure 4 Reciprocity in Knowledge Sharing Among knowledge portal facilitates the diffusion of knowl-
Projects After One Year of Using the Knowledge Portal edge among rather loosely coupled and often discon-
nected innovation projects. With regard to the case
study of the knowledge portal in the agricultural
of using the knowledge portal based on document industry we conclude that a knowledge portal will
exchanges. have an impact on how projects share knowledge
and on the emergence of a network of practice. The
Figure 4 indicates there is not much reciprocity with results show that pre-conditions for the emergence
regard to explicit innovative knowledge sharing of a network of practice are a sense of urgency and
among the projects. The empirical analysis shows a fragmented awareness. The results also indicate
there is high reciprocity related to two projects the important role of a knowledge broker. The devel-
(P01, P20). There is one project (P25) – which is not oped knowledge portal seems to lead to overcoming
one of the core innovation projects but a cooperation structural holes and a closer cognitive distance
project among project leaders of the innovation pro- among the projects. However, we did not find a direct
jects that has a free rider characteristic. In general, effect of the knowledge portal on sharing tacit knowl-
there was no free rider behavior among the innova- edge. In the initial phase of a network of practice the
tion projects. The question remains if – given the knowledge exchange seems to focus on general, non-
objectives of the knowledge portal e.g. exchanging project specific and explicit knowledge. There was
innovative knowledge – the knowledge portal stimu- also no direct effect of the knowledge portal on the
lated the exchange of knowledge. Given the empiri- reciprocity of knowledge exchange among the pro-
cal results the answer is that there was no direct jects. However, knowledge was shared between the
and strong relationship between the impact of the project level and the platform and public level.
knowledge portal on reciprocity of knowledge shar-
ing. Proposition 6 is not accepted. Several explana- This paper makes three key contributions to the liter-
tions could be given for this result. The first one is ature of knowledge management, networks of prac-
that the innovation projects started with a long cogni- tice, and innovation policy. First, it identifies critical
tive distance among the projects. Therefore it is not factors in explaining how networks of practice
easy to develop mutual understanding and trust. emerge. It focuses on the situation where people
The second explanation is that reciprocity is not exe- and projects in different organisations are previously
cuted via the knowledge portal but via other chan- unaware of each other and start to share knowledge
nels (direct contact, telephone). A third potential and use a knowledge portal. It shows that even in
explanation is that the level of urgency – although this type of situation networks of practice can
indicated as high at the start of the platform – to emerge. Second, it provides a conceptual framework
share knowledge projects was not that high. It seems that explains critical factors for the development of
there was a high level of urgency within the projects common knowledge and the emergence of networks
to execute them (deliver to deadlines, secure finan- of practice. We think that the typology of knowledge
cial resources, link to relevant partners), but a lower related to project, platform, and public will be useful
level of urgency to directly help other projects. The in the design of future knowledge portals. Also, the
empirical results indicate that, as argued by Kollock typology of projects in terms of supply and demand
(1999) the generalized exchange system of sharing of knowledge is a useful tool to analyze (potential)
is both more generous and riskier. It is more gener- reciprocity in knowledge exchange relationships.
ous because the person who gives provides the net- Third, it provides a detailed analysis of the emer-
work with a benefit without the expectation of gence of a network of practice around agro-logistical
immediate return. However, generalized exchange innovation projects in The Netherlands.



This study has some limitations and the results need Grant, R.M. (1996) Toward a knowledge-based theory of
to be interpreted with care. These limitations are: the ﬁrm. Strategic Management Journal 17 (winter
special issue), 109–122.
Hippel, von E. (1994) Sticky Information and the locus of
v The use of the knowledge portal could not be iso- problem solving: implications for innovation. Manage-
lated from other knowledge exchanges among ment Science 40(4), 429–439.
the projects. Telephone contact and physical Jones, C., Hesterly, W.S. and Borgatti, S.P. (1997) A general
meetings also stimulated the sharing of informa- theory of network governance: exchange conditions
and social mechanisms. Academy of Management Review
tion and knowledge and this impact is not taken 22(4), 911–945.
into account. Kollock, P. (1999) The economies of online cooperation:
v There was no detailed analysis done on the gifts and public goods in cyberspace. In Communities
impact of shared documents on the practice of in Cyberspace, eds M.A. Smit and P. Kollock, pp.
the receiving project. Did it lead to active use of 220–239. Routledge, London etc.
Krogh, von G. (2003) Knowledge Sharing and the Com-
the knowledge gathered? munal Resource. In Handbook of Organizational Learning
v The knowledge portal is analyzed before and and Knowledge Management, eds M. Easterby-Smith and
after one year of use. The period might be too M.A. Lyles, pp. 372–392. Blackwell Publishing, Mal-
short to see a sustainable effect and impact. den, Oxford, Melbourne, Berlin.
Lave, J. and Wenger, E. (1991) Situated Learning: Legitimate
Peripheral Participation. Cambridge University Press,
Cambridge, New York, Melbourne.
Suggestions for Further Research Markus, M.L., Majchrzak, A. and Gasser, L. (2002) A design
theory for systems that support emergent knowledge
There are at least two directions for further research processes. MIS Quarterly 26(3), 179–210.
on the impact of knowledge portal on the emergence McLure Wasko, M. and Faraj, S. (2000) ‘‘It is what one
of networks of practice. does’’: why people participate and help others in
electronic communities of practice. Journal of Strategic
Information Systems 9, 155–173.
Our research propositions need to be tested by large- Ministries of LNV and V&W. (2001). Visie Agrologistiek:
scale statistical inquiry. Clusteren, Verbinden, Regisseren (in Dutch), Rapport aan
de Tweede Kamer, Den Haag, 30 p., 14 November
We concentrated on document exchange as a ﬁrst 2001 (www.agrologistiek.nl).
indicator of knowledge exchange. A broader analysis Nonaka, I. (1994) A dynamic theory of organizational
knowledge creation. Organization Science 5(1), 14–37.
is required. Nonaka, I. and Takeuchi, H. (1995) The Knowledge-Creating
Company. Oxford University Press, Oxford etc.
Nooteboom, B. (2000) Learning by interaction, absorptive
Acknowledgement capacity, cognitive distance, and governance. Journal of
Management and Governance 4, 69–92.
Nooteboom, B. and I. Bogenrieder. 2003. Change of Routines:
The research in this article was carried out in the a Multi-Level Analysis, Rotterdam: ERIM report
KLICT project Kennisportal Agrologistiek HR-185. (www.erim.nl).
We thank the Platform Agrologistiek for their sup- Podolny, J.M. and Page, K.L. (1998) Network forms of
organization. Annual Review Sociology 24, 57–76.
port during the project and thank Mathijs van der Swan, J., Newell, S., Scarbrough, H. and Hislop, D. (1999)
Vlis for his assistance in analyzing the data. Knowledge management and innovation: networks
and networking. Journal of Knowledge Management 3(4),
262–275.
References Szulanski, G. (2003) Sticky Knowledge. Barriers to Knowing in
the Firm. Sage Publications, London, Thousand Oaks,
Baalen, van P., J. Bloemhof, E. van Heck, P. van de Veire. New Delhi.
(2003), Analyse kennisportal Agrologistiek (in Dutch), Tuomi, I. (2002) Networks of Innovation: Change and Meaning
report Erasmus University Rotterdam, 40 pp. in the Age of the Internet. Oxford University Press,
Borgatti, S., M. Everett and L. Freeman. (2004). UCINET 5 Oxford.
for Windows: User’s Guide. Harvard: Analytic Technol- Tsoukas, H. (2003) Do we really understand tacit knowl-
ogies, 7 June 2004. edge? In Handbook of Organizational Learning and
Brown, J.S. and Duguid, P. (2000) The Social Life of Knowledge Management, eds M. Easterby-Smith and
Information. Harvard Business School Press, Boston. M. Lyles, pp. 410–427. Blackwell Publishing, Malden
Burt, R.S. (1992) Structural Holes: The Social Structure of etc.
Competition. Harvard University Press, Cambridge, Weick, K. (1979) The Social Psychology of Organizing. (second
London. ed.), New York.
Chaffey, D. and Wood, S. (2004) Business Information Wellman, B. and Gulia, M. (1999) Virtual communities as
Management: Improving Performance Using Information communities: Net surfers don’t ride alone. In Commu-
Systems. Prentice-Hall, Essex. nities in Cyberspace, eds M.A. Smith and P. Kollock, pp.
Cohen, W.M. and Levinthal, D.A. (1990) Absorptive 167–194. Routlegde, London and New York.
capacity: a new perspective on learning and innova- Wenger, E., McDermott, R. and Snyder, W.M. (2002)
tion. Administrative Science Quarterly 35, 128–152. Cultivating Communities of Practice. Harvard Business
Damsgaard, J. (2002) Managing an Internet Portal. Com- School Press, Boston.
munications of the Association for Information Systems 9, Winograd, F. and Flores, F. (1986) Understanding Computers
408–420. and Cognition: A new Foundation for Design. Ablex
Davenport, T.H. and Prusak, L. (1998) Working Knowledge: Publishing Corporation, New Jersey.
How Organizations Manage What They Know. Harvard Yin, R.K. (2003) Case Study Research: Design and Methods.
Business School Press, Boston MA. (Third ed.). Sage, Thousand Oaks.



PETER VAN ERIC VAN HECK,
BAALEN, RSM Eras- RSM Erasmus Univer-
mus University, Depart- sity, Department of
ment of Decision and Decision and Information
Information Sciences, P.O. Sciences, P.O. Box 1738,
Box 1738, 3000 DR Rot- 3000 DR Rotterdam,
terdam, The Netherlands. The Netherlands. E-mail:
E-mail: pbaalen@rsm.nl evanheck@rsm.nl

Peter van Baalen is Eric van Heck is Profes-
Associate Professor in the sor of Electronic Markets
Department of Decision at Erasmus University’s
and Information Sciences Rotterdam School of
at RSM Erasmus University. His main ﬁelds of Management. His research concentrates on electronic
research interests are knowledge management, markets and IT-enabled business networks. His recent
knowledge networks, e-communities, e-learning, ICT book Smart Business Networks – co-edited with
adoption and diffusion, and management education. Peter Vervest, Kenneth Preiss, and Louis-Francois¸
Pau – was published in January 2005 (Springer).
JACQUELINE
BLOEMHOF-
RUWAARD, RSM
Erasmus University,
Department of RSM
Decision and Information
Sciences, P.O. Box 1738,
3000 DR Rotterdam,
The Netherlands. E-mail:
jbloemhof@rsm.nl

Jacqueline Bloemhof-
Ruwaard is Assistant
Professor in the Department of Decision and Infor-
mation Sciences at the RSM Erasmus University
Rotterdam. Her main ﬁelds of research interests are
supply chain management, closed loop supply chains,
logistics networks, agricultural distribution networks,
and sustainability.


Knowledge Sharing in an Emerging Network of Practice

More Related Content

Similar to Knowledge Sharing in an Emerging Network of Practice (20)

More from Eric van Heck (20)

Recently uploaded (20)

Knowledge Sharing in an Emerging Network of Practice