Knowledge Sharing in an Emerging Network of Practice

doi:10.1016/j.emj.2005.04.008

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

� 2005 Elsevier Ltd. All rights reserved.

Printed in Great Britain

0263-2373 $30.00

Knowledge Sharing inan Emerging Networkof Practice:The Role of a KnowledgePortal

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 ofpractice and the role of knowledge sharing viaknowledge portals. Its focus is on factors that stim-ulate the successful emergence of networks of prac-tice. Literature on knowledge management andcommunities of practice suggest the pre-existenceof shared knowledge or a shared belief system asa condition sine qua non for the networks of practiceto emerge. We challenge this assumption and argueand demonstrate that common knowledge andbelief systems are rather a result of knowledge shar-ing rather than a pre-condition. The central ques-tion is how a knowledge portal facilitates thediffusion of knowledge among rather loosely cou-pled and often disconnected innovation projects.Research is carried out in the agricultural industryin The Netherlands. In this industry there is a needto change from a product-oriented to a problem-ori-ented innovation structure. The set up of a platformand knowledge portal around agro-logistics – cross-ing different product-oriented production clusters –was therefore a logical result. It gave the opportunityto analyze what the impact of a knowledge portal isin a situation where people and projects come fromdifferent organizations and do not know each other.Do they start to share knowledge and what are theconditions? With regard to the case study of theknowledge portal in the agricultural industry weconclude that a knowledge portal will have animpact on how projects are sharing knowledge

300 E

and on the emergence of a network of practice.The results show that pre-conditions for the emer-gence of a network of practice are a sense of urgencyand fragmented awareness. These results also indi-cate the important role of a knowledge broker. Thedeveloped knowledge portal seems to lead to over-coming structural holes and a closer cognitive dis-tance among the projects. However, we did notfind a direct effect of the knowledge portal on shar-ing tacit knowledge. In the initial phase of anetwork of practice the knowledge exchange seemsto focus on general, non-project specific and explicitknowledge. There was also no direct effect of theknowledge portal on the reciprocity of knowledgeexchange among the projects. However, knowledgewas shared between the project level and the plat-form and public level. Conclusions and directionsfor future research are formulated.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Agro-logistics, Innovation projects,Knowledge portal, Knowledge sharing, Networkof practice, Social networks

Introduction

The diffusion of innovative knowledge is consideredto be one of the main challenges in the emerging

uropean 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 isdistributed and fragmented, Internet-based informa-tion and communication technologies can help toleverage the knowledge diffusion. These technolo-gies can easily connect distributed and loosely cou-pled ‘pockets of innovation’ and diffuse relevantinformation at high speed and at relatively low costs,see Tuomi (2002).

For this reason a platform of representatives of gov-ernment, industry, and knowledge institutes in TheNetherlands, the so-called Platform Agro-logistics, ini-tiated the setting up of a knowledge portal in orderto facilitate and speed up the diffusion of innovativeknowledge in the agricultural industry. The set-up ofthis knowledge portal in the Dutch agriculturalindustry should be considered as an innovation it-self. For many years this industry was characterizedby a closed and hierarchical knowledge infrastruc-ture in which the government dictated the researchthemes to the agricultural knowledge institutes. Theresearch results were disseminated and communi-cated to the agricultural companies who were ex-pected to apply this new knowledge in practice.But recent disasters such as the outbreak of animaldiseases such as BSE showed the limits of this ap-proach and new ways of innovations were explored.

In this paper we consider the diffusion of innovativeknowledge as a form of collective action that requiressocial (collective) organization. It implies that theknowledge diffusion is viewed as an interactive pro-cess including the involvement of different collectiveactors.

The research question we address here is how aknowledge portal facilitates the diffusion of knowl-edge among rather loosely coupled and often discon-nected innovation projects. Although the knowledgeportal can easily connect these disconnected projectsand thereby facilitate knowledge diffusion we willargue that a minimal social organization is neededto initiate this diffusion process. Literature on knowl-edge management and communities of practice sug-gest the pre-existence of shared knowledge or ashared believe system as a condition sine qua nonfor the networks of practice to emerge, see for exam-ple Cohen and Levinthal (1990), Nonaka (1994), andGrant (1996). We challenge this assumption andargue and demonstrate that common knowledgeand common belief systems are rather the result ofknowledge sharing instead of a pre-condition. Theaim of this article is twofold. The first objective isto conceptually describe the emergence of a networkof people and groups that do not share knowledgeand beliefs at the initial situation. The second objec-tive is to empirically show how this network emergesand evolves and what factors contribute to the suc-cessful emergence. It implies that we do not assumethe existence of a particular form of a social network(e.g. community of practice) in advance, but willview this as the outcome of network evolution.

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

Research was carried out in the agricultural industryin The Netherlands in particular the set up of innova-tion projects around themes related to agro-logistics,see Ministries of LNV and V&W (2001). Agro-logis-tics deals with the logistics e.g. transportation, stor-age, and distribution of agricultural products. Theanswer to the above question was sought in a casestudy approach. The case study provides a basisupon which theoretical propositions are formulatedand generalized (so called analytic generalization),see Yin (2003). The choice of the case setting madeit possible to analyze how a network emerges andhow people and groups – that did not know eachother – started to share knowledge. The case studylet us closely track the design and use of a knowledgeportal that could facilitate knowledge sharing amongdifferent innovation projects.

This article is divided into three main sections. First,a literature review of knowledge sharing in networksand the role of knowledge portals is developed into aconceptual framework, complemented with six prop-ositions. Second, the empirical setting in the agricul-tural industry with research method and data will beexplained. Third, an empirical analysis of the case ofthe knowledge portal in the agricultural industrywill be presented. Lessons learned, conclusions,and suggestions for further research are formulated.

Literature Review and ConceptualFramework

Knowledge Sharing

The diffusion of innovative knowledge has becomeone of the major research interests in managementscience and economics. A huge body of literature fo-cuses on innovation as a ‘‘thing’’ about which infor-mation needs to be provided to potential adoptersand users in order to implement this innovation suc-cessfully (Swan et al., 1999: 262). As knowledge hasbecome to be seen as an innovation in itself new, crit-ical questions arise how to define knowledge andhow innovative knowledge can be diffused. Sincethe former question has been discussed extensivelyin the recent management literatures it suffices todiscuss it briefly here. Since the publication of Non-aka’s seminal paper ‘‘A dynamic theory of organiza-tional knowledge creation’’ the complex distinctionbetween explicit and tacit knowledge has beenwidely accepted (Nonaka, 1994). The issue is not ifthere exists such a distinction but how to understandthe complex relationship between explicit and tacitknowledge. Roughly, two different views can be dis-tinguished in this debate: the ‘near tangible view’and the distributed view on knowledge (Tsoukas,2003). In the former view it is assumed that explicitand tacit knowledge can be converted to each other(Nonaka, 1994; Nonaka and Takeuchi, 1995). This viewsuggests that knowledge, by means of articulation,

301


can be called upon for use in reasoning and canbe translated into language and other media (Wino-grad and Flores, 1986: 73). In the distributed view itis believed that tacit knowledge is a component ofall knowledge and as such cannot be converted intoexplicit knowledge. Tacit knowledge is not interna-lised explicit knowledge, nor is explicit knowledgeexternalised tacit knowledge. In viewing ideas as ob-jects that can be extracted from people and transmit-ted to others over a conduit, Nonaka and Takeuchireduce practical knowledge to technical knowledge.According to Tsoukas (2003) tacit and explicit knowl-edge are complementary, in the sense that explicitknowledge is always grounded on a tacit componentand vice versa. Tsoukas further criticizes the notion ofknowledge as a given or something that is to be dis-covered. The organization is a distributed knowledgesystem and cannot be surveyed as a whole; it is lack-ing an ‘‘overseeing mind’’. Similarly, Winograd andFlores argue that articulation of the unspoken is anever-ending process, as we must do it in a languageand a background that itself reflects a pre-under-standing. ‘‘Knowledge’’, as they put it, ‘‘is always aresult of the interpreter, which depends on the entireprevious situation and on its position in a tradition(1986: 75). Thus knowledge has an important tacitcomponent, which resides in individual skills, under-standing, collaborative social arrangements, but alsoin tools, documents, and processes that embody as-pects of knowledge (Wenger et al., 2002: 11). As theseskills and social arrangements are related to workactivities we will call them practices (Szulanski,2003). This view contrasts the ‘near tangible view’as it suggests that any form of explicit knowledge as-sumes the existence of tacit knowledge that cannot bearticulated. As a consequence, the transfer of innova-tive knowledge from one practice to another will be-come problematic. Disembedding knowledge fromone practice and re-embedding this knowledge intoanother practice does not go without any costs.Von Hippel has coined the concept of ‘‘stickiness’’of knowledge to refer to the incremental costs totransfer knowledge from one practice to another(Von Hippel, 1994: 430, see also Szulanski, 2003).When transfer costs are low, knowledge stickinessis low; when it is high, knowledge stickiness is high.Both Von Hippel (1994) and Szulanski (2003) point tothe fact that the stickiness of knowledge involves notonly the complex epistemology of knowledge, butalso attributes of the knowledge source, the knowl-edge recipient, and of the context. When the knowl-edge source and the knowledge recipient share thesame context and are engaged in the same practice,the stickiness will be relatively low, whereas thetransfer cost will increase when the knowledgesource and the knowledge recipient operate in differ-ent contexts and are engaged in different practices.

Knowledge transfer within and between organisa-tions is not a one-way activity, but a process of trialand error, feedback, and mutual adjustment of boththe source and the recipient of knowledge (Von

302 E

Krogh, 2003: 373). This mutuality in the knowledgetransfer suggests that the process can be construedas a sequence of collective action in which the sourceand the recipient are involved (Von Krogh, 2003: 373).For this reason we will use the term knowledge sharing,instead of diffusion and transfer, as it succinctly re-fers to the social processes that are involved. Sharingknowledge is not giving a full representative accountof what is known by the source about a particularpractice to the recipient. Because of the tacit compo-nent, knowledge contains an ineffable element; it isbased on an act of personal insight that is essentiallyinarticulable. Tsoukas argues that this does not meanthat we cannot share knowledge about a practice, butit should be viewed as re-punctuation of distinctionsunderlying the practice, as drawing attention tounnoticed aspects and as making people aware ofnew connections (Tsoukas, 2003). The stickiness ofknowledge sharing does not only refer to the episte-mological but also to the relational problems. Accord-ing to Szulanski (2003) people on the source side maybe reluctant to share their knowledge with others forfear of losing ownership, a position of privilege, supe-riority, for the lack of insufficient rewards, for lackingtime to communicate about an innovative practice.Another reason can be that people are unaware ofthe fact that their knowledge might be of interest toothers. On the recipient side important factors likethe reluctance to accept new knowledge from anexternal source (‘not invented here’-syndrome), theinability to exploit outside sources of knowledge(absorptive capacity), an inability to retain the newlyacquired knowledge in the organization, increase thestickiness of knowledge sharing.

Emergence of Networks of Practice

The sharing of knowledge requires social organiza-tion and governance. Traditional organizationalforms (markets and hierarchies) show serious deficitsin organizing the complex nature of knowledge(Jones et al., 1997). For this reason new organizationalforms are introduced to deal effectively with thesharing of explicit and implicit knowledge. The com-munity of practice concept, introduced by Lave andWenger (1991) and transferred to the managementdomain by Brown and Duguid (2000), representsprobably one of the potentially most useful andenduring concepts in this respect. Most definitionsof communities of practice (CoP) stress the impor-tance of shared practice, repertoire, interests, knowl-edge, on informality, and on the self-organizingcharacter of the community. Recently, Brown andDuguid (2000) have distinguished two types of net-works, networks of practice (NoPs) and communitiesof practice (CoPs). In the former, people have prac-tice and knowledge in common but are mostly un-known to each other. The links between thenetworks are mostly indirect (e.g. databases, newslet-ters, info bulletins) and members coordinate andcommunicate normally explicit. NoPs can have an



enormous reach. There is relatively little reciprocityacross NoPs as the members do not interact directly.NoPs are loosely coupled systems that hardly initiatecollective action and produce little knowledge.

CoPs on the other hand represent relatively tight-knitgroups of people who know each other well andwork together directly. Online communication isoften supported by face-to-face interactions, whichenable them to coordinate and communicate to ahigh degree on implicit knowledge. Due to theseface-to-face relationships the communication reachis bounded. CoPs are characterized by strong reci-procity norms which help to sustain the community.

Although the distinction between CoPs and NoPsseems to be clear at the surface level, it is hard todetermine precisely in advance if the social collectiveshould be conceived as a CoP or a NoP. We suggestthat both, CoP and NoP, are particular forms andtherefore suggest taking the social network as thestarting point for our analysis and conceive CoPsand NoPs as particular forms of social networks. Asocial network can be defined as a patterned organi-zation of a collection of actors and their relationships

The collection of actors

should contain more than

two... to be defined as a

network

(Jones et al., 1997). It is impor-tant to note that in this minimaldefinition no specifications aregiven about the nature of theactors and their relationships.According to Wellman andGulia (1999) this implies thateven when people are onlyconnected through a computernetwork, they should be con-ceived as a social network.
We don’t agree with this minimal definition becauseif no interaction takes place one cannot speak of a so-cial network. The collection of actors should containmore than two actors to be defined as a network. Tri-adic relationships differ fundamentally from dyadicrelationships because in the former 1) individualityis reduced; 2) the individual power is reduced; 3)and conflicts are moderated by the presence of a thirdparty. We can add to this definition two other charac-teristics (Podolny and Page, 1998). The first is that thecollection of actors pursue repeated, enduring ex-change relations with one another. If exchanges arenot enduring but episodic - engaging in an incidentaltransfer of goods, services or information - there is nosocial network but a market situation. The second isthat social networks lack a legitimate organizationalauthority to arbitrate and resolve disputes that mayarise during the exchange (as is the case in hierar-chies). Based on these characteristics a network canbe viewed as a social exchange structure with itsown governance structure and patterns of interactionin which flows of resources between independentunits (or individuals) take place.
Most research on social networks focuses on existingsocial structures. Less attention is paid to the way


these networks emerge and evolve. In their studyof CoPs Wenger et al. (2002) made a first attempt tosketch the evolution of CoPs by identifying fivestages of community development. According tothe authors, CoPs typically start as loose networksthat hold the potential of becoming more connectedand develop towards a tightly-knit community.However, loose connectedness presumes the exis-tence of particular ties between the members of a po-tential network. This might make sense within thecontext of one organization or CoPs where homoge-neity of interests and knowledge can be presumed.Our question, however, focuses on the emergenceof those initial ties between actors that come fromdifferent organizations and who do not or hardlyknow each other. Many authors state that the coordi-nation and sharing of knowledge cannot take placewithout assuming a vast amount of mutual knowl-edge, mutual beliefs, and mutual assumptions. Thisis what is called common ground. Similarly, Grant(1996: 115) argues that common knowledge (lan-guage, other forms of symbolic communications,shared meanings, commonality of specialized knowl-edge etc.), defined as the intersection of individualknowledge sets, should be conceived as a precondi-

tion for the knowledge inte-gration. Cohen and Levinthal(1990) point to the importanceof overlapping knowledge inorganizations in order toassimilate external knowledge,whereas Nonaka (1994) viewsredundancy of knowledge asa necessary precondition forknowledge creation and thebuilding of trust. Nooteboom

(2000) has coined the concept of cognitive distanceand cognitive proximity to refer to cognitive close-ness and similarity between people. It does not onlyrefer to the cognitive variety but also to the differ-ences in abilities of perception, interpretation, anddifferent views on the world that develop out activeinteraction with the physical and social environment(Nooteboom, 2000). Cognitive distance yields both aproblem and an opportunity (Nooteboom andBogenrieder, 2003). When the cognitive distance istoo short or is absent for people to share the sameknowledge, there is no incentive to share knowledge.This might be the case when people interact fre-quently and consequently establish strong networkslike CoPs. However when cognitive distances are toogreat, the more difficult it becomes to cross the dis-tance, i.e. to mutually understand the actions andexpression in the network. The notion of cognitivedistance is relevant here as it succinctly points tothe potential for a network to emerge. The questionhowever is still how much cognitive similarity isneeded to initiate knowledge sharing. In orderto find out what is minimally needed to initiateenduring interaction we briefly discuss Weick’s(1979) theory on the emergence of collective struc-ture. Weick (1979) argues that people initially don’t

303


have to agree on goals to act collectively. In any po-tential collective, people have different interests,preferences etc. and want to accomplish differentthings. In order to achieve these ends they have toinitiate action towards others by which they createmutual commitment (interlocked behaviours) to col-lectively pursue diverse ends through commonmeans. Once people are engaged in mutual commit-ments a subtle shift takes place from diverse to com-mon ends. As Weick argues, diverse ends remain,but they become subordinated to an emerging setof shared ends. This part of Weick’s evolutionarytheory contrasts conventional thinking about the pre-conditions for the emergence of collective structures.The second part of Weick’s theory addresses thequestion of how coordination can take place eventhough ties between people are minimal. To thisend he discusses the notion of a mutual equivalencestructure (MES). The MES is like an implicit contractbetween people that can be built and sustainedwithout knowing the motives of another, and with-out people having to share goals. Weick points tothree preconditions for an MES to emerge. The firstis that a person must perceive that his ability to per-form his consummatory act depends on the instru-mental act of the other. The second is that a personmust perceive that his own instrumental act servesto elicit the instrumental act of the other. Third, aMES only emerges when a person repeats his instru-mental act. A fourth precondition can be added, thatis that a person must have some knowledge (expecta-tions) about another person that can fulfil his instru-mental acts. From this perspective we may concludethat the common ground, needed to initiate a MES,does not refer to common knowledge or a commonbelief system but from a mutual expectationstructure.

The issue of reciprocity in online networks is widelydebated among researchers. Especially the motiva-tion for people to contribute to an online connectedgroup of people who do not or hardly know eachother have been subject to extensive research. Well-man and Gulia (1999) point to different types ofexplanations. The first refers to the fact that onlinecontributions are a means of expressing one’s iden-tity. Helping others might increase self-esteem, repu-tation, respect from others etc. The second one isgeneralized reciprocity and organizational citizen-ship. In their recent research on three online commu-nities McLure Wasko and Faraj (2000) concluded thatsharing knowledge and helping others is ‘the rightthing to do’ and that people also have a desire to ad-vance the community as a whole. Participants didnot expect to be reciprocated by the same personwith whom they shared knowledge (direct exchange)but did expect to receive future help from someonein the network. Wellman and Gulia (1999) argue thatthe logistic and social costs involved in online contri-bution are relatively low. The easy access to onlinesocial networks allows and enables people to contrib-ute at low participation costs. Discussing the econo-

304 E

mies of online cooperation Kollock (1999) points tothe limitations of online cooperation and collectiveaction. Although it is quite easy to produce and sharedigital information, it requires coordinated activitiesfrom the beginning. Another weakness is that if ac-tive knowledge producers withdraw from the onlinenetwork, the network will cease to exist. The basicfeatures for an online network are: ongoing interac-tion, identity persistence, and knowledge of the pre-vious interactions. (Kollock, 1999: 235). The notion ofonline generalized exchange demonstrates how frag-ile the minimal social situation of emergent socialnetwork is.

Design of Knowledge Portals

The question then is what the design and the man-agement of an interactive information and communi-cation system should look like for the online(generalized) exchange of knowledge in a minimalsocial situation. The agro-logistic projects are geo-graphically dispersed and the participants hardlyknow each other and have rarely communicated. Inthe literature three dominant perspectives on the roleof information and communication technologies onknowledge sharing are distinguished: deterministicview on technology, medium choice theory, emer-gent process perspective. We believe that the threeperspectives are not mutually exclusive. It is nowwidely accepted that user involvement is essentialin the process of design and implementation. How-ever the user is just one node in the potentialnetwork of knowledge sharing. It is not only impor-tant to know if user A prefers to use a particularapplication of the intranet for the transfer of a partic-ular type of knowledge but also if persons B and Chave similar preferences and expectations. The impli-cation of our argument here is that the design of aninteractive information system should not reflectthe needs of individual users but the social structure(triadic relationships) of the emerging network. Wetherefore call for a relational and rich information sys-tem design. That design will have the following threecharacteristics:

v It should be relational as it should not onlyaddress the needs of individual users but alsothe triadic expectations of a potential social net-work. It means that in the initial stage of thedevelopment of the knowledge portal the poten-tial network actors should be informed aboutthese expectations (e.g. by organizing meetings,providing information about the projects, adver-tising, see Damsgaard, 2002).

v It should also be rich as it is impossible to predicthow the actors in the network will communicate.Social networks are complex social systems thatcannot be simply founded. They develop andtransform over time (Wenger et al., 2002). Toallow the online social network to take differentshapes and to evolve in different directions the



design of the knowledge portal should dynami-cally match different social profiles of the net-work. It implies that the knowledge portalshould provide different spaces of knowledgesharing, synchronous and asynchronous commu-nications media, document storage and retrievaletc.

v Perhaps the most important requirement for thedesign in the minimal social situation is thatpeople converge on the means, in our case theknowledge portal. Following Weick’s theory onthe emergence of collective structures we con-sider the knowledge portal as a means to facilitatethe sharing of knowledge between differentgroups.

Markus et al. (2002) argue that traditional informa-tion system design theories are badly equipped todeal with emergent knowledge processes (EKPs).EKPs are defined as organizational activity patternsthat exhibit three characteristics in combination:deliberations with no best structure or sequence;highly unpredictable potential users and work con-texts; and information requirements that includegeneral, specific, and tacit knowledge distributedacross experts and non-experts. We believe that theset up of a knowledge portal in agro-logistics inorder to facilitate the diffusion of knowledge be-tween the distributed projects fits pretty well the sit-uation as described by Markus et al. (2002). However,the portal is only the ‘front door’ of an intranet or anextranet (Chaffey and Wood, 2004). Intranets andextranets are called decentralized, general purpose-and open-ended technologies which mean they canbe designed for different purposes and can poten-tially be constructed and modified by those whoare involved in the design and use of these informa-tion systems (Damsgaard, 2002).

Knowledge Sharing

ActiveKnowledge

Broker

ActiveKnowledge

Broker

OvercomingStructural Holes

OvercomingStructural Holes

Type ofKnowledge

Type ofKnowledge

Type of

KnowledgePortal

KnowledgePortal

P3AP3A

P4AP4A

P5AP5A

P6AP6A

PP

Type ofKnowledge Sharing

Figure 1 The Conceptual Framework


Conceptual Framework

The following conceptual framework depicts the roleand impact of knowledge portals and how networksof practice emerge, see Figure 1. Based on the litera-ture review we formulate the following propositions:

As discussed by Brown and Duguid (2000), in net-works of practice people have practice and knowl-edge in common but are mostly unknown to eachother. The links between the networks are mostlyindirect (e.g. databases, newsletters, info bulletins)and members coordinate and communicate andthese are normally explicit. There are two factors thatseem to be a pre-condition for the emergence of net-works of practice: sense of urgency and fragmentedawareness.

Proposition 1. A higher sense of urgency to tacklespecific problems of practice will lead to the emer-gence of a network of practice.

People are tackling specific problems of practice andone way to do that is to coordinate and communi-cate. However, there has to be a high sense of ur-gency that people will coordinate and communicatewith people they hardly know (and also from otherorganizations). Without that level of urgency ‘‘outof the box’’ thinking seems a strategy people willnot follow.

Proposition 2. Fragmented awareness in a dispersedindustry will lead to the emergence of a network ofpractice.

A second pre-condition is that there has to be a frag-mented awareness in a dispersed industry. Peopleneed to have the expectation that somewhere out

Emergence ofNetwork of Practice

Emergence ofNetwork of Practice

P3BP3B

4B4B

P5BP5B

P6BP6B

Sense ofUrgencySense ofUrgency

FragmentedAwarenessFragmentedAwareness

P1P1 P2P2

305


there solutions are available. They know what theydon’t know and they know that somebody else mightknow (about similar problems and potential solu-tions). Without that fragmented awareness thereseems to be no logical reason to strengthen tiesamong people.

Next to these two basic factors we think there aremore specific factors that stimulate or hamper theemergence of Networks of Practice (NoPs). The fol-lowing factors will be taken into account: action bybroker, role structural holes, type of knowledge,and type of knowledge sharing.

Proposition 3. An active knowledge broker will leadto the development of a knowledge portal and theemergence of a network of practice.

The links in a network are mostly indirect. Thereforein the initial phase there has to be an active brokerbringing people together who did not know eachother before. Previous research has shown that an ac-tive broker (in a coordinated or spontaneous way)helps to create the indirect linkages among membersof an emerging network. The role of a knowledgebroker is identified by, for example, Davenport andPrusak (1998).

Proposition 4. A knowledge portal will bridge struc-tural holes and contribute to the emergence of anetwork of practice.

Potential knowledge portals have the ability to createdirect linkages (between the portal and theknowledge sender/receiver) in such a way that directlinkages between the sender and receiver are not nec-essary. In such a case structural holes are overcome(Burt, 1992). As we have seen there is a paradox inthe sense that overcoming structural holes will leadto effective knowledge exchange because as Noote-boom and Bogenrieder (2003) indicated cognitive dis-tance yields both a problem and an opportunity. It isstill unclear how much cognitive similarity is neededto initiate knowledge sharing.

Proposition 5. A knowledge portal will lead to theexchange of project-domain knowledge and there-fore contribute to the emergence of a network ofpractice.

A knowledge portal will make it easier and less costlyto transfer and exchange knowledge. However, as wehave seen, related to the stickiness of knowledge bothVon Hippel (1994) and Szulanski (2003) indicate thetransfer cost will increase when the knowledgesource and the knowledge recipient operate in differ-ent contexts and are engaged in different practices.

Proposition 6. A knowledge portal will lead to reci-procity in knowledge sharing and therefore contrib-ute to the emergence of a network of practice.

306 E

As Kollock (1999) argues, the generalized exchangesystem of sharing is both more generous and riskier.It is more generous because the person who givesprovides the network with a benefit without theexpectation of immediate return. However general-ized exchange is also more risky because actors areeasily temped to a free ride (taking without contrib-uting). However, the basic features for an online net-work are: ongoing interaction, identity persistence,and knowledge of the previous interactions. (Kol-lock, 1999: 235). Therefore a knowledge portal hasto lead to a certain level of reciprocity in knowledgesharing to sustain the emergence of a network ofpractice.

Research Methods and Data

Case Study Background

The role and impact of knowledge portals for theemergence of networks of practice is illustratedhere by a case study of a knowledge portal foragro-logistic innovation projects in The Nether-lands. For a detailed description of case study re-search, see Yin (2003). Agro-logistics deals withthe transport, storage, and distribution of the agri-cultural flows of food and non-food goods in theentire supply chain. Agro-logistics is an importantsector. In The Netherlands, more than 20% of goodtransportation (including import and export) in-cludes agro products. The agribusiness has recentlydealt with a number of bottlenecks such as animaldiseases leading to trade embargos, congestion onthe Dutch highways, international competition,and stronger legislation regarding food safety andanimal well-being. Recently, a number ofdevelopments in society have taken place, influenc-ing the management of agro-logistic flows. Thesedevelopments are: higher consumer awareness, pullstrategy (market) instead of push strategy(producers), fragmentation, scaling-up in retailand agro-distribution, globalisation and liberalisa-tion, sustainable entrepreneurship, sharpened legis-lation, and more attention to tracing and foodsafety.

The agricultural community has a product-relatedcluster structure. This can be illustrated by thenames of the Product Boards (regulatory organiza-tions for businesses in the agricultural supplychains): Animal Feed, Beer and Wine, Cattle, Meatand Eggs, Dairy, Farming, Grains and Seeds, Horti-culture. These sectors are highly independent ofeach other with weak ties between each other. Theyoften are called the Pillars of Agriculture. Withinthese pillars, knowledge is available and peoplehave regular contacts with each other. Betweenthe pillars, the information sharing and communi-cation is quite low. Recent developments and bot-tlenecks encouraged the community to change



from a product-related structure towards a prob-lem-related one. These problems occur in the areaof spatial planning, EU legislation, high scale infra-structure, and optimizing logistic networks. In or-der to develop a vision on the sustainablecoherent future of the agro-logistics sector, a plat-form of representatives of government, industry,and knowledge institutes was set up, the so-calledPlatform Agro-logistics (Ministries of LNV andV&W, 2001). The Vision Agro-logistics aims to reacha sustainable, innovating and transport-efficient sec-tor and is based on three keywords, i.e. Clustering,Binding, and Directing. The national government,cooperating with the Platform, invited the sectorto propose innovative projects in the area of agro-logistics to improve sustainable development. Theinnovative character can be related to Clustering(realisation of large scale agri-business areas), Bind-ing (innovative logistical concepts) and Directing(towards virtual livestock markets or worldwideplant cultivation networks). The Platform Agro-logistics focuses on coordinating between parties,tuning with governmental organizations, and creat-ing support. The goals of the platform are (i) to ad-vise, cooperate and coach pilot projects to succeedin system innovations and (ii) to attract attentionand share information on threats and opportunitiesin the agro-logistics sector. In 2003, 20 innovativepilot projects were selected by the Dutchgovernment to be supported in their developmentby the Platform Agro-logistics. The project propos-als came from almost all pillars in the agriculturalindustry like vegetable products, chicken, plantcultivation, cheese, and pig farming. The innova-tions are not essentially product based, but had tofit into the themes of Clustering, Binding, andDirecting.

Knowledge Portal: Stages and Data

It is important that knowledge on how to innovateand the innovation itself are exchanged in an efficientand effective way among the projects and potentialnew projects. The Platform Agro-logistics suggestedsetting up a (virtual) place to meet each other, shareinformation and knowledge, deal with governmentaland policy issues, and seek financial resources, inother words, to be a network of practice. A knowl-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-vative knowledge in the various pilot projects,regardless of time, place, and existing relations ofknowledge exchange.

The development of a knowledge portal took placein three stages. In the first stage (March 2003–June2003) objectives, requirements, and design ruleswere determined. Structured interviews were heldwith all project leaders of the different innovationprojects, see Van Baalen et al. (2003). The aim of


these interviews was to answer the followingquestions:

v Do the selected innovative projects have a needfor a knowledge portal?

v Is there a need for specific knowledge and infor-mation (both in content as in type of knowledge/information)?

v Is there a willingness to share?v If there is a need, what is the main design of the

portal, and what are critical success factors forthe design, building, and implementation of theknowledge portal?

The interview results showed there was a high sense ofurgency to tackle specific problems of practice exists.About 80% of the project leaders stated a need to ex-change knowledge by means of a knowledge portal.The knowledge portal was also seen as an effectivemedium to reach the public in general, creating a basisfor innovative projects, making projects known to thepublic, and finding new partners. The interviews alsoshow that knowledge and information need is verydiverse between projects. It is unusual that thesegroups are not divided by the central Platform themesClustering, Binding and Directing, neither by the pil-lars of the Agro Sector. The Projects can be dividedas follows:

v Entrepreneurs who want to share knowledge onlegislation, best practices, subsidiaries, andlobbying;

v Knowledge institutes with no need for a socialknowledge network, interested in Europeansubsidiaries;

v Umbrella projects with a need to share knowledgein the field of project management and regionalscaling-up.

All interviewees showed willingness to share infor-mation with each other and with the public space.With respect to the design of the knowledge portal,a layered structure was suggested in such a waythat it reflected the current community structure,i.e. a project level, a platform level, and a public le-vel. Each level gives entrance to specific types ofinformation. Being present in the public spacewas one of the priorities of the project leaders(visibility).

In the second stage (July 2003–August 2003) theknowledge portal was designed, built, and tested. Itwas decided to structure the knowledge portal inthree levels. The first level deals with the innovationprojects, the second level with the platform, the thirdlevel with the public space. At the project levelknowledge sharing among the members of the pro-ject is facilitated. The members of a project share acommon practice in which knowledge primarily re-lated to the project is developed and shared. Rela-tions are direct and tightly coupled and implicit

307

218

331

277296

312

390

448465

341

440

475

571

0

100

200

300

400

500

600

sep2003

oct2003

nov2003

dec2003

jan2004

feb2004

mar2004

apr2004

may2004

jun2004

jul 2004

aug2004

months

nu

mb

er o

f vi

sito

rs

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


knowledge sharing (learning by doing) possible. Atthe platform level knowledge can be shared amongthe different projects and the members of the Plat-form Agro-logistics. The platform level is only acces-sible by the members of the different projects and theplatform. The knowledge exchange at this level isworthwhile for the community as a whole, as it givesa base for sharing experiences and best practicesamong the sectors and therefore from moving froma product-related innovation structure towards aproblem-related structure. Finally, at the public levelknowledge can be exchanged between the innova-tion projects and the actors outside (public, innova-tion projects outside agro-logistics, otherindustries). Here information is available for every-body and free of charge.

In the third stage (September 2003–until now) theknowledge portal was used. A web master was tak-ing care of the functioning of the portal and of theinstruction of users. In that period we were able tomonitor the use of the knowledge portal and there-fore could analyze who was using it and how itwas used. The use of the knowledge portal in the firstyear (September 2003–August 2004) will be pre-sented in this article. Detailed statistics were avail-able on the profile of visitors, the amount of hitsand page views, and details about visitor sessions.A visit is defined as a hit originating from the sameIP-address with a maximum time between the hitsof 20 minutes. A monitoring tool was developedand linked to a social network analysis software pro-gram called UCINET 5 (Borgatti et al., 2004). With thehelp of this program a general analysis of relation-ships among the projects in the emerging networkof practice could be identified. For a more thoroughexplanation of the research methods and techniquesused, we refer to Van Baalen et al. (2003).

308 E

Analysis of Empirical Data

Knowledge Portal Statistics

The knowledge portal (www.agrologistiek.nl) wentlive in September 2003. For one year, we analyzedthe knowledge portal statistics. In one year the totalnumber of hits was about 275,000. The results indi-cate the knowledge portal had between 15,000 and20,000 hits per month in the period December2003–May 2004. In the summer of 2004 the numberof hits increased. Figure 2 shows the monthly num-ber of unique visitors to the knowledge portal.

The data of the number of visitors show there was asteady increase in visitor numbers due to the fact thatthe public was more aware of the existence of theknowledge portal. The total number of visitors topublic level varied from 218 in the first month to571 one year later. Figure 2 shows that the numberof visitors grew steadily over the year, with twoexceptions: the second month (October) had a rela-tive high number of visitors, caused by the noveltyof the site and the month of May had a relative lownumber of visitors, probably due to the Spring holi-days. The number of visits suggest there is a needfor information sharing within and outside the agri-cultural community.

In one year more than 7,500 documents were down-loaded. Table 1 presents the top 10 downloaded files.Original titles of the documents were in Dutch.

From this list, we conclude that the need for informa-tion focuses on the existence of the Platform Agro-logistics and its vision (1,6,8,9), and less on thecontent and urgency of the innovative projects. Onlyone of the themes of the Platform Agro-logistics is in


http://www.agrologistiek.nl

Table 1 Top 10 of Downloaded Files from Septem-ber 2003 to September 2004

Document Number of

downloads

1 Brochure_Platform_Agro-logistics.pdf 2786

2 Agro-Logistic Invitation Letter.pdf 1252

3 Final report_Agro-chains

and Clusters.pdf

960

4 Agro-logistics_ Examples.MPG 697

5 Pre-announcement_Agro-logistics.pdf 600

6 Vision_Agro-Logistics.pdf (Platform) 514

7 Bundling of Agro-streams.pdf

(Clustering)

476

8 Conference_registration_Aro-logistics.pdf

(Platform)

465

9 Letter_MinistersLNVandVenWto

Second Chamber.pdf (Platform)

316

10 Examples Agri-parcs.pdf (Clustering) 278


the top 10, namely Clustering logistics flows in Agri-business centers (3,7,10).

Apart from a statistical analysis at the Public level, itwas possible to monitor the communication betweenprojects themselves, at the Platform and Project level.Some functionalities were hardly or never used. Forexample, the discussion forum was hardly ever used.Also, the use of the bulletin board was negligible.These features were pointed out as potentially usefulfeatures in the interviews. The feature of sharinginformation by downloading documents was fre-quently used, as indicated by Table 1. For eachdownloaded document, data was available with re-spect to the supplier of the document (providinginformation) and the client (receiving information).Table 2 provides information on who shared docu-ments with whom. In Table 2 the providers are rep-resented in the rows (between brackets the numberof posted documents), the columns represent the

Table 2 Document Sharing among the Projects in the

Supply Demand

P01 P02 P03 P04 P11 P12 P13

P01(12) 127 0 0 0 0 0 0

P02(5) 0 0 0 0 2 0 0

P03(9) 1 0 12 0 0 0 0

P04(7) 1 0 0 1 0 0 0

P11(14) 9 0 0 0 12 0 0

P12(10) 4 0 4 0 4 0 0

P13(1) 0 0 0 0 0 0 1

P15(9) 0 0 0 0 0 0 0

P17(11) 0 0 0 0 0 0 0

P20(45) 0 0 3 0 0 0 0

P21(9) 2 0 0 0 0 0 0

P23(1) 0 0 0 0 0 0 0

P25(6) 1 0 0 0 0 0 0

P26(5) 0 0 0 0 0 0 0

P27(5) 0 0 0 0 0 0 0


receivers. Providers and receivers are innovationprojects. In total there were 25 innovation projectsindicated by P01, P02, . . ., P23, P26, and P27. P24and P25 were projects for general and project man-agement purposes. Not all projects come back inTable 2 due to the fact that these projects werestopped or merged with others. The values in thematrix represent the number of times that, for exam-ple, project 01 downloads information from project02 in the Platform or Project space. These spaceswere restricted to members only and can be visitedthrough a login name and a password.

Table 2 shows mixed results: some projects wereactive providers of documents, some projects wereactive consumers of documents, and some pro-jects were not very active at all in sharing documents.A more detailed analysis of knowledge sharingamong projects will be discussed in the next section.

Lessons Learned

Based on interviews of the project leaders, the designand use of the knowledge portal, the knowledge por-tal statistics, and monitoring of the document ex-change among the projects the proposition – asdefined in Section 2 – were validated. A network ofpractice will emerge if there is a sense of urgency(Proposition 1) and fragmented awareness in a dis-persed community (Proposition 2). Proposition 3states that active action of a broker will lead to thedevelopment of a knowledge portal. The broker inthis case is the Platform Agro-logistics. For testingPropositions 4–6, we carried out interviews with allinnovative pilot projects (Van Baalen et al., 2003) andanalyzed for one year (September 2003–August2004) the actual use of the knowledge portal. The pur-pose of the interviews is to find data on the networksituation before the start of the knowledge portal(Proposition 4, Structural Holes), to investigate the

Knowledge Portal

P15 P17 P20 P21 P23 P25 P26 P27

0 0 0 0 0 0 0 0

0 0 1 0 0 1 0 0

0 0 2 0 0 1 0 0

0 0 6 0 0 3 0 0

0 0 0 0 0 4 0 0

0 0 15 0 0 18 0 0

0 0 0 0 0 0 0 0

0 0 3 0 0 1 0 0

0 1 6 0 0 2 0 0

0 0 559 0 0 2 0 0

0 0 3 1 0 2 0 0

0 0 8 0 0 0 0 0

0 0 16 0 0 141 0 0

0 0 13 0 0 21 0 0

0 0 6 0 0 0 0 0

309


need for knowledge and the type of knowledgeneeded (Proposition 5, Types of knowledge), andthe willingness to share knowledge (Proposition 6,Reciprocity in knowledge sharing). The analysis pro-vides the following lessons learned.

Proposition 1. A higher sense of urgency to tacklespecific problems of practice will lead to the emer-gence of a network of practice.

In the former, people have practice and knowledge incommon but are mostly unknown to each other. Thelinks between the networks are mostly indirect (e.g.databases, newsletters, info bulletins) and memberscoordinate and communicate in a normally explicitfashion. NoPs can have an enormous reach. There isrelatively little reciprocity across NoPs as the mem-bers do not interact directly to one another. NoPsare loosely coupled systems that barely initiate collec-tive action and produce little knowledge.

The Agro-logistics case shows that there was on theone hand a high sense of urgency in the agriculturalsector to start the Platform Agro-logistics. Severalreasons were mentioned during the initial phase ofthe set up of the platform. These reasons were re-lated to the outbreak of animal diseases, transporta-tion problems, and more strict legislation for foodsafety. Also during the interviews with the projectleaders at the start it turned out that most of themindicated an urgent need for collective action withregard to transport and distribution problems re-lated to agro-products in The Netherlands. On theother hand in analyzing objectives, incentives, andthe lack of direct subsidies for projects to participatein there was not a high level of urgency. Projectscould not be pushed to deliver results in a fastway, subsidies were not directly given to projects,the platform was installed to facilitate the differentinnovation projects. Overall, there was a shared le-vel of urgency to innovate to keep The Netherlandscompetitive in the field of agro-logistics. Proposition1 is accepted.

Proposition 2. Fragmented awareness in a dispersedindustry will lead to the emergence of a network ofpractice.

The Agro-logistics case shows there was a frag-mented awareness in a dispersed agricultural indus-try. Traditionally, the agricultural community isstructured in a product-oriented way (meat, milk &cheese, flowers, fruit & vegetables). Agro-logisticalproblems and solutions are endemic in thesedifferent product units. Therefore there was a needfor agro-logistical experts to learn from innovationsin different product-oriented communities. In thesedifferent communities there was awareness that theagro-logistics community is highly dispersed and alack of coordinated action was hampering solutionsto agro-logistical problems. Proposition 2 is accepted.

310 E

Proposition 3. An active knowledge broker will leadto the development of a knowledge portal and theemergence of a network of practice.

The set up of the Platform Agro-logistics includingrepresentatives of different stakeholders in the differ-ent product-related communities together with repre-sentatives of local and national authorities encourageda broking role in an emerging network of practice. Inparticular, the chairman and secretary of the platformacted as active brokers – they took the initiative todevelop a knowledge portal. In the initial phase ofthe knowledge portal most documents and initiativeswere posted by the web master. Proposition 3 isaccepted.

Proposition 4. A knowledge portal will bridge struc-tural holes and contribute to the emergence of anetwork of practice.

To look at the impact of the knowledge portal in over-coming structural holes and decreasing cognitive dis-tance among projects we examined which projectswere known to each other before the platform andknowledge portal were implemented. The differentproject leaders were interviewed and asked if theyknew the other projects (Van Baalen et al., 2003).Based on these interview results relationships amongthe different projects were measured and drawnwith the help of UCINET software (Borgatti et al.,2004). On the left side of Figure 3 the initial networkis sketched. As one can see, there are seven projectsthat have no relationship with other projects and someprojects have very weak ties with others. None of theprojects regularly exchanged information and knowl-edge. After one year of using the knowledge portalwe examined which documents were exchangedamong the different projects. The right side of Figure3 presents the network after the introduction and useof the knowledge portal. As Figure 3 indicates, pro-jects are exchanging documents and therefore couldlearn from each other. One can see a network of prac-tice is emerging. The knowledge portal overcomessome structural holes and there are indications thatthe cognitive distances between the actors are nottoo short e.g. that there is no incentive to share knowl-edge. Proposition 4 is accepted.

Proposition 5. A knowledge portal will lead to theexchange of project-domain knowledge and there-fore contribute to the emergence of a network ofpractice.

We distinguish three types of knowledge:

v Type I – This type of knowledge is project-domain knowledge and developed by one ofthe innovation projects.

v Type II – This type of knowledge is platform-domain knowledge and developed by one ofthe innovation projects.


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


v Type III – This type of knowledge is public-domain knowledge and developed by one ofthe innovation projects.

As discussed, the knowledge portal was developedwith three levels or spaces: project, platform andpublic. By analyzing the use of the knowledge portal(documents, web pages, bulletin board) we concludethat information exchange among the different pro-jects (Type I) was rather limited. The analysis alsoshows that some documents were exchanged at theplatform level (Type II) and most documents wereexchanged at the public level (Type III). There seemto be two potential explanations. The first one relatesto the stickiness of knowledge – see Von Hippel(1994) and Szulanski (2003). The different innovationprojects did not exchange because transfer costs weretoo high due to the fact that the knowledge sourceand the knowledge recipient operate in different con-texts and are engaged in different practices. The de-creased transfer costs of the knowledge portal didnot overcome the high level of transfer costs relatedto the stickiness of the knowledge. The second onerelates to the concept of cognitive distance – seeNooteboom and Bogenrieder (2003). The cognitivedistance between the innovation projects seems tobe too high and therefore it is more difficult to crossthe distance between the projects. It seems to be log-ical that at the emergence of a network of practiceknowledge exchange will start with knowledge withlow transaction costs and a low cognitive distance(such as general project knowledge). The analysisalso indicates there is information exchange withinsome of the projects. This can be considered as anew type of knowledge (knowledge exchange withinproject). In particular, large projects (with around 50project members) were eager to exchange informa-tion within the project. Proposition 5 is not accepted.


However, the empirical results indicate a revisedProposition 5: In the initial phase of a network ofpractice knowledge exchange will be focused on gen-eral, non-project specific and explicit knowledge. It isexpected that after this phase the exchange will bedirected to project-specific and tacit knowledge.

Proposition 6. A knowledge portal will lead toreciprocity in knowledge sharing and thereforecontribute to the emergence of a network ofpractice.

Reciprocity in a network means that projects are post-ing and demanding knowledge from other projects.This is different from market relationships where aspecific activity (an indication to buy something) willautomatically lead to an offer by the other party. In anetwork of practice one can post knowledge but onedoes not automatically and directly get somethingin return. However, in the longer term one expectsthat if one posts a question to the network – or rather,to a member, (in our case projects) – that it will reactwith an offer. We analyzed the knowledge exchangeamong the projects and distinguished four type ofprojects. These four types are labeled:

v Individualistic Projects: These projects do notpost or demand information and knowledge.

v Altruistic Projects: These types of project post alot, but make no demand.

v Free rider Projects: These projects show no post-ing, but demand a lot.

v Reciprocity Projects: These project post and makedemands.

It is interesting to analyze how projects developedduring the use of the knowledge portal. Figure 4identifies the typology of the projects after one year

311

P25

P20

demand

P03

P17

P04

P13, P23

P12

P11

P15,P21

P02, P26, P27

P01

ReciprocalAltruïst

Free rider Individualistsupply

Figure 4 Reciprocity in Knowledge Sharing AmongProjects After One Year of Using the Knowledge Portal


of using the knowledge portal based on documentexchanges.

Figure 4 indicates there is not much reciprocity withregard to explicit innovative knowledge sharingamong the projects. The empirical analysis showsthere is high reciprocity related to two projects(P01, P20). There is one project (P25) – which is notone of the core innovation projects but a cooperationproject among project leaders of the innovation pro-jects that has a free rider characteristic. In general,there was no free rider behavior among the innova-tion projects. The question remains if – given theobjectives of the knowledge portal e.g. exchanginginnovative knowledge – the knowledge portal stimu-lated the exchange of knowledge. Given the empiri-cal results the answer is that there was no directand strong relationship between the impact of theknowledge portal on reciprocity of knowledge shar-ing. Proposition 6 is not accepted. Several explana-tions could be given for this result. The first one isthat the innovation projects started with a long cogni-tive distance among the projects. Therefore it is noteasy to develop mutual understanding and trust.The second explanation is that reciprocity is not exe-cuted via the knowledge portal but via other chan-nels (direct contact, telephone). A third potentialexplanation is that the level of urgency – althoughindicated as high at the start of the platform – toshare knowledge projects was not that high. It seemsthere was a high level of urgency within the projectsto execute them (deliver to deadlines, secure finan-cial resources, link to relevant partners), but a lowerlevel of urgency to directly help other projects. Theempirical results indicate that, as argued by Kollock(1999) the generalized exchange system of sharingis both more generous and riskier. It is more gener-ous because the person who gives provides the net-work with a benefit without the expectation ofimmediate return. However, generalized exchange

312 E

is also more risky because actors are easily temptedto free ride (taking without contributing). For thisreason the generalized exchange has the structureof a social dilemma in which individually reasonablebehavior might lead to collective disaster (Kollock,1999). In our case there was no free-rider behavioramong the projects, but also no balanced knowledgesupply and demand. In the longer term there is thepotential risk that the knowledge portal runs dryand that the network of practice will dissolve.

Conclusions

Research Problem

The central research question of this article is how aknowledge portal facilitates the diffusion of knowl-edge among rather loosely coupled and often discon-nected innovation projects. With regard to the casestudy of the knowledge portal in the agriculturalindustry we conclude that a knowledge portal willhave an impact on how projects share knowledgeand on the emergence of a network of practice. Theresults show that pre-conditions for the emergenceof a network of practice are a sense of urgency anda fragmented awareness. The results also indicatethe important role of a knowledge broker. The devel-oped knowledge portal seems to lead to overcomingstructural holes and a closer cognitive distanceamong the projects. However, we did not find a directeffect of the knowledge portal on sharing tacit knowl-edge. In the initial phase of a network of practice theknowledge exchange seems to focus on general, non-project specific and explicit knowledge. There wasalso no direct effect of the knowledge portal on thereciprocity of knowledge exchange among the pro-jects. However, knowledge was shared between theproject level and the platform and public level.

This paper makes three key contributions to the liter-ature of knowledge management, networks of prac-tice, and innovation policy. First, it identifies criticalfactors in explaining how networks of practiceemerge. It focuses on the situation where peopleand projects in different organisations are previouslyunaware of each other and start to share knowledgeand use a knowledge portal. It shows that even inthis type of situation networks of practice canemerge. Second, it provides a conceptual frameworkthat explains critical factors for the development ofcommon knowledge and the emergence of networksof practice. We think that the typology of knowledgerelated to project, platform, and public will be usefulin the design of future knowledge portals. Also, thetypology of projects in terms of supply and demandof knowledge is a useful tool to analyze (potential)reciprocity in knowledge exchange relationships.Third, it provides a detailed analysis of the emer-gence of a network of practice around agro-logisticalinnovation projects in The Netherlands.



This study has some limitations and the results needto be interpreted with care. These limitations are:

v The use of the knowledge portal could not be iso-lated from other knowledge exchanges amongthe projects. Telephone contact and physicalmeetings also stimulated the sharing of informa-tion and knowledge and this impact is not takeninto account.

v There was no detailed analysis done on theimpact of shared documents on the practice ofthe receiving project. Did it lead to active use ofthe knowledge gathered?

v The knowledge portal is analyzed before andafter one year of use. The period might be tooshort to see a sustainable effect and impact.

Suggestions for Further Research

There are at least two directions for further researchon the impact of knowledge portal on the emergenceof networks of practice.

Our research propositions need to be tested by large-scale statistical inquiry.

We concentrated on document exchange as a firstindicator of knowledge exchange. A broader analysisis required.

Acknowledgement

The research in this article was carried out in theKLICT project Kennisportal Agrologistiek HR-185.We thank the Platform Agrologistiek for their sup-port during the project and thank Mathijs van derVlis for his assistance in analyzing the data.

References

Baalen, van P., J. Bloemhof, E. van Heck, P. van de Veire.(2003), Analyse kennisportal Agrologistiek (in Dutch),report Erasmus University Rotterdam, 40 pp.

Borgatti, S., M. Everett and L. Freeman. (2004). UCINET 5for Windows: User’s Guide. Harvard: Analytic Technol-ogies, 7 June 2004.

Brown, J.S. and Duguid, P. (2000) The Social Life ofInformation. Harvard Business School Press, Boston.

Burt, R.S. (1992) Structural Holes: The Social Structure ofCompetition. Harvard University Press, Cambridge,London.

Chaffey, D. and Wood, S. (2004) Business InformationManagement: Improving Performance Using InformationSystems. Prentice-Hall, Essex.

Cohen, W.M. and Levinthal, D.A. (1990) Absorptivecapacity: a new perspective on learning and innova-tion. Administrative Science Quarterly 35, 128–152.

Damsgaard, J. (2002) Managing an Internet Portal. Com-munications of the Association for Information Systems 9,408–420.

Davenport, T.H. and Prusak, L. (1998) Working Knowledge:How Organizations Manage What They Know. HarvardBusiness School Press, Boston MA.


Grant, R.M. (1996) Toward a knowledge-based theory ofthe firm. Strategic Management Journal 17 (winterspecial issue), 109–122.

Hippel, von E. (1994) Sticky Information and the locus ofproblem solving: implications for innovation. Manage-ment Science 40(4), 429–439.

Jones, C., Hesterly, W.S. and Borgatti, S.P. (1997) A generaltheory of network governance: exchange conditionsand social mechanisms. Academy of Management Review22(4), 911–945.

Kollock, P. (1999) The economies of online cooperation:gifts and public goods in cyberspace. In Communitiesin Cyberspace, eds M.A. Smit and P. Kollock, pp.220–239. Routledge, London etc.

Krogh, von G. (2003) Knowledge Sharing and the Com-munal Resource. In Handbook of Organizational Learningand Knowledge Management, eds M. Easterby-Smith andM.A. Lyles, pp. 372–392. Blackwell Publishing, Mal-den, Oxford, Melbourne, Berlin.

Lave, J. and Wenger, E. (1991) Situated Learning: LegitimatePeripheral Participation. Cambridge University Press,Cambridge, New York, Melbourne.

Markus, M.L., Majchrzak, A. and Gasser, L. (2002) A designtheory for systems that support emergent knowledgeprocesses. MIS Quarterly 26(3), 179–210.

McLure Wasko, M. and Faraj, S. (2000) ‘‘It is what onedoes’’: why people participate and help others inelectronic communities of practice. Journal of StrategicInformation Systems 9, 155–173.

Ministries of LNV and V&W. (2001). Visie Agrologistiek:Clusteren, Verbinden, Regisseren (in Dutch), Rapport aande Tweede Kamer, Den Haag, 30 p., 14 November2001 (www.agrologistiek.nl).

Nonaka, I. (1994) A dynamic theory of organizationalknowledge creation. Organization Science 5(1), 14–37.

Nonaka, I. and Takeuchi, H. (1995) The Knowledge-CreatingCompany. Oxford University Press, Oxford etc.

Nooteboom, B. (2000) Learning by interaction, absorptivecapacity, cognitive distance, and governance. Journal ofManagement and Governance 4, 69–92.

Nooteboom, B. and I. Bogenrieder. 2003. Change of Routines:a Multi-Level Analysis, Rotterdam: ERIM report(www.erim.nl).

Podolny, J.M. and Page, K.L. (1998) Network forms oforganization. Annual Review Sociology 24, 57–76.

Swan, J., Newell, S., Scarbrough, H. and Hislop, D. (1999)Knowledge management and innovation: networksand networking. Journal of Knowledge Management 3(4),262–275.

Szulanski, G. (2003) Sticky Knowledge. Barriers to Knowing inthe Firm. Sage Publications, London, Thousand Oaks,New Delhi.

Tuomi, I. (2002) Networks of Innovation: Change and Meaningin the Age of the Internet. Oxford University Press,Oxford.

Tsoukas, H. (2003) Do we really understand tacit knowl-edge? In Handbook of Organizational Learning andKnowledge Management, eds M. Easterby-Smith andM. Lyles, pp. 410–427. Blackwell Publishing, Maldenetc.

Weick, K. (1979) The Social Psychology of Organizing. (seconded.), New York.

Wellman, B. and Gulia, M. (1999) Virtual communities ascommunities: Net surfers don’t ride alone. In Commu-nities in Cyberspace, eds M.A. Smith and P. Kollock, pp.167–194. Routlegde, London and New York.

Wenger, E., McDermott, R. and Snyder, W.M. (2002)Cultivating Communities of Practice. Harvard BusinessSchool Press, Boston.

Winograd, F. and Flores, F. (1986) Understanding Computersand Cognition: A new Foundation for Design. AblexPublishing Corporation, New Jersey.

Yin, R.K. (2003) Case Study Research: Design and Methods.(Third ed.). Sage, Thousand Oaks.

313

http://www.agrologistiek.nl

http://www.erim.nl

PETER VANBAALEN, RSM Eras-mus University, Depart-ment of Decision andInformation Sciences, P.O.Box 1738, 3000 DR Rot-terdam, The Netherlands.E-mail: [email protected]

Peter van Baalen isAssociate Professor in theDepartment of Decisionand Information Sciences

at RSM Erasmus University. His main fields ofresearch interests are knowledge management,knowledge networks, e-communities, e-learning, ICTadoption and diffusion, and management education.

JACQUELINEBLOEMHOF-RUWAARD, RSMErasmus University,Department of RSMDecision and InformationSciences, P.O. Box 1738,3000 DR Rotterdam,The Netherlands. E-mail:[email protected]

Jacqueline Bloemhof-Ruwaard is Assistant

Professor in the Department of Decision and Infor-mation Sciences at the RSM Erasmus UniversityRotterdam. Her main fields of research interests aresupply chain management, closed loop supply chains,logistics networks, agricultural distribution networks,and sustainability.

ERIC VAN HECK,RSM Erasmus Univer-sity, Department ofDecision and InformationSciences, P.O. Box 1738,3000 DR Rotterdam,The Netherlands. E-mail:[email protected]

Eric van Heck is Profes-sor of Electronic Marketsat Erasmus University’sRotterdam School of

Management. His research concentrates on electronicmarkets and IT-enabled business networks. His recentbook Smart Business Networks – co-edited withPeter Vervest, Kenneth Preiss, and Louis-FrancoisPau – was published in January 2005 (Springer).


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

Knowledge Sharing in an Emerging Network of Practice

Business

Transcript of Knowledge Sharing in an Emerging Network of Practice