Interc0mmunity,boundary-spanning

knowledgeprocesses

lntroduction

This chapter is premissedon the idea that knowledge processeswithin and between
communities of practice are quite different and distinctive. This is primarily because
rvhile membersof a community of practicehave much common knowledgeand a strong
sharedsenseof identity, people who are not membersof the samecommunity typically
Jo not. While Chapter 5 examined the characteristics of intracommunity knowledge
rrocesses,the focus of this chaptel is exclusively on intercommunity knowledge
processes.Further, this chapter will show that intercommunity knowledge processes
are typically more complex and difficult to make successful.Why this is the casewill be
rully exploredasthe chapterprogresses.
Intercommunity knowledge processesencapsulatean enormous variety of contexts
and can involve knowledge processeswhich span community, occupational, organiza-
tional, functional, national, or project boundaries.This chapterbuilds on issuesraisedin
Chapter 3, such as the nature of the organizationalknowledge base,and is primarily
iounded in a practice-based perspectiveon knowledge.This will becomeapparentas the
chapterprogresses, asvarious terms and conceptsare (re)introduced.
The chapter begins in the following section by considering why intercommunity
knowledgeprocesses are so important. After this, the main sectionof the chapter exam-
ines the characterof intercommunity knowledge processes,and presentsa number of
eramples to illustrate the points made. Finally, the chapter closesby considering the way
that intercommunity knowledgeplocessescan be facilitatedand managed.

The significanceof intercommunity knowledge processes

Considerthe following situations:

. A ioint technology development project involving closecollaboration between UK and

Japaneseelectronicscompanies(Lam 1'997).
@ S o C I A LA N D C U L T U R AI L
SSUES

r The consolidation of the knowledge base in some accounting and consultancy

companiesfollowing mergersand acquisitions(Empson2001).
o CollaborationbetweenindigenousMaori groupswith the New Zealandgovernmentin
negotiation over land treaties(pauleenand yoong 2001).
' Attempts by proiect-basedcompaniesinvolved in the design of complex products to
shareknowledgebetweenprojects(prencipeand Tell 2001).
' A large-scaleinteruniversityresearchprolect involving staff from three UK universities
whose disciplinary backgrounds encompass engineering, operation management,
organizationalbehaviour,and marketing (Newelland Swan2000).
o Interorganizationalproduct development efforts in the biotechnology sector (Powell
et al. 1996).
o Ctoss-occupationalcollaboration that occurs as part of the concurrent engineering
work at a semiconductorequipment factory in the USA (Bechky2003).

All these situations, while being diverse in character,have one thing in common:
they involve the sharing, or joint utilization and development of knowledge among
people who do not typically work together, and who have substantially different
knowledgebases.One of the reasonswhy examining the dyramics of intercommunity
knowledge processesis so important is that the type of working practicesoutlined in
theseexamplesis becomingmore and more common. Thus evidencesuggests that the use
of proiect-basedworking methods and the utilization of interpersonaland interorganiza-
tional networks has become widespread(for example, seeCastells1996; Cravenset al.
1996; Daviesand Brady 2000; Powell 1990).For example,all three of the organizational
contextsexaminedlater,in Chapters1,2-14,i.e. knowledge-intensivefirms, global multi-
nationals,and network/virtual organizations,involve the utilization of intercommuniw
knowledgeprocesses.
Another factor that signalsthe importance of intercommunity knowledgeprocesses is
the growing acknowledgementthat the knowledgebasesof all organizationsare to some
extent fragmented into separate,specializedknowledge communities. As outlined in
chapter 3, this led Brown and Duguid (1991, 53), to refer to organizationsas being
comparableto a'community-of-communities'. Thus, the knowledgebaseof all organiza-
tions can be consideredas being made up from a diversity of localized communities
which have some overlapping knowledge in common, but which also possessmuch
specializedand speciflcknowledge.As this perspectiveis closelyassociatedwith the practice-
based perspective on knowledge, the specialized and localized nature of much
organizationalknowledge is related to the particular tasks and activities that different
groupsof workersundertake.
From this perspective,one of the generaltasksof managementis to coordinate these
diverseinternal communities,integrating,diffusing,and combining fragmentedinternal
knowledge as necessary (Blackler et al. 2000; Brown and Duguid zool; Grant 1996;
Tsoukas1996).Thus, if the knowledgebaseof all organizationsis constitutedby a diverse
collection of specializedknowledgecommunities,managingintercommunity knowledge
processes will be a day-to-dayactivity for most organizations.
 KG
BOUNDARY-SPANNIN N O W L E D G EP R O C E S S E S

: :he levelof fragmentation in an organization'sknowledgebaselikelyto be proportronal

to
:-ganizational of managingsucha fragmentedknowledgebase
size?Further,if so, arethe difficulties
(ely to be greatestfor large,globalmultinationals?

Thus, the importance of intercommunity knowledge processesstemsboth from the fact

:hat the contempolaly Iestructudng of olganizations is placing a Sreatel emphasis on
tntercommunity and interorganizational working than has been traditional, and also
":ecauseintra-organizational coordination can be conceptualized as involving intercom-

:nunity interaction.

g intercommunity knowledge processes

Characterizin

-ntercommunity knowledge processesinvolve collaboration between individuals who are

-ikely to have a limited amount knowledge in common, and who may have a limited, or
-,seaksenseof shared identity. As will be seen, in terms of knowledge processes,the
.onsequencesof this are significant.
.trs illustrated by Figure 5.1 in the previous chapter, knowledge plocesses within
.ommunities of practice are facilitated by the high degreeof common knowledge, over-
-appingvalues,and sharedsenseof identity that community memberstypically possess.
Ihis is becausein such circumstancesit is likely that the tacit assumptionsunderpinning
reople's knowledge, which are key to effective knowledge-sharing, are likely to be well
':nderstood, or commonly shared. Also, the level of trust and mutual understanding
retween people in this context is also likely to be conducive to effective knowledge-
.haring. Hansen (L999), in the context of product innovation and development
frocesses,argues that effective knowledge-sharing requires two key elements to exist
Table6.1). First,people must be willing to sharetheir knowledge,and secondly,people
nust have the ability to shareknowledge. Both these elementstypically exist within com-
nunities of practice as due to the shared knowledge and values, there is enough mutual
-rnderstandingto make the sharingof knowledgepossible,while the senseof sharediden-
:itv and values makes it probable that people will be willing to sharetheir knowledge.
However, in intercommunity knowledge processesthe situation is somewhat different
seeTable 6.2). ln these circumstancespeople will have much less shared, common
.nowledge, they may only have a weak sense of shared identiry or may even have
distinctive and separateidentities, and, finally, may have fundamentally different value

Table 6.1. Factorsunderpinningeffectiveknowledge-sharing

(adaotedfrom Hansen1999)

Willingness
Ability(adequate
mutualunderstanding)
 S O C I A LA N D C U L T U B A LI S S U E S

Table6.2. Factorsmakingintercommunity
knowledge
processesdifficult

Limitedcommonknowledge
Weaksharedidentityor differentsenseof identity
Values/assumptions
potentially
different

systems.Thus, the social relations between people who are not members of the same
Sroup/community are much lessconducive to effective knowledge-sharing.For example,
Hansen (1999) found that when weak ties existed between people this was likely to
impede the transfer of complex knowledge (knowledge which was highly tacit, and
which had a high level of interdependencewith other knowledge).
The following two subsectionsconsider how the lack of a sharedidentity, andlor a
limited degreeof common knowledge can inhibit knowledge processes,illustrating the
issuesexaminedwith examples.

ldentity

Peoplefrom different groups or communities who work together may have either a weak
senseof common identity, or may have distinctive and separateidentities. For example,
consider the situation described by Lam (1997), outlined above, and elaborated more
fully later. In the electronicscorporation examined, the Japaneseand uK staff who
requiredto collaboratehad a weak senseof sharedidentity asbeing membersof the same
organization. Instead,their identity was more closely linked to the divisions they had
historically worked within. More negatively,Empson (2001)found post-mergerattempts
at consolidatingthe organizationalknowledgebasein one of the consulting companies
she examined to have been signiflcantly inhibited by the strength of identity that staff
retained for their pre-mergerorganizations,and the typically disdainful view that they
had regardingthe knowledgeand experienceof workersin the company they had been
mergedwith.
This potentially weak senseof common identity arguably complicates knowledge
processes through the potential for conflict this creates,aspeoplewith differing sensesof
identity may perceivedifferencesof interestto exist betweenthemselvesand others.The
issueof conflicting interests,and how this can inhibit knowledge-sharingwastouched on
earlierin Chapter 4 and is examinedagain more fully in Chapter 7.

Globalbank:conflicting identities inhibiting knowledge-sharing

G l o b a l b a n ki s a D u t c h b a n k t h a t g r e w a g g r e s s i v e l b
y y a c q u i s i t i o nB. y t h e l a t e ' l 9 9 0 s i t h a d
divisionsin over 70 countriesworldwide. At this point corporatemanagementdecided it was
necessaryto lmprove levels of coordinationand knowledge-sharing between divisions.A key
element of this strategy was the developmentof a global intranet,a project developedand
 BOUNDARY-SPAN
I NNG K N O W L E D G EP R O C E S S E S a
-anaged by corporatelT staff. However,Globalbankhad a strong historicalcultureof divisional

:.rtonomy,with divisionshavingtypicallyoperatedcompletelyindependentlyfrom each other.

-rus
each divisionhad controlledhow it was organized,with the consequencethat each division
-ad its own working practices,lT systems,etc. For example,each divisionhad its own intranet

: te, with its own specificstyle, levelof functionality,etc.

Staff thus typicallyhad a strong sense of identitywith their division,and possessedspecialist
.rrowledgerelatedto their division'sparticularcustomers,products,marketconditions,and inter-
^al ways of working. The globalintranetprojectexperiencedsignificantproblems however,as
-anagement staff f rom most divisionswere hostileto the idea,primarilybecausethey perceived
:.e objectivesof the projectto be incompatiblewith their desire to retaindivisionalautonomy.
-1us
one of the main obstaclesto the project'sprogresswas the strongersense of identitythat
.ey divisionalmanagementstaff typicallyhad for their specificdivisionratherthan the corporate
l'oup asa wnore.

,'Jhatcanbe doneto overcomethe narrowsenseof divisional identitythat staffhad,whichwas

:rting as a brakeon the progressof the globalintranetproject?

Knowledge

The difficulties of knowledge-sharing between communities are however related to more

than just the senseof identities that individuals possess.Another, equally important
factor complicating such processes,outlined above, is the nature of the knowledge pos-
sessedby people in these situations. Thesedifficulties stem from three interrelated factors
,Table6.3). Firstly,the degreeof common knowledgesharedby peoplemay be quite lim-
ited, with different people possessingspecialistknowledge related to the specificactivities
they each undertake.Secondly,the knowledgepossessed by people may also be 'sticky'
and difficult to share as it may be context-specific, tacit, and highly localized in nature
rBrown and Duguid 1998; Lam 1997).Thirdly, and finally, there may be significant epi-
stemologicaldifferencesin the knowledgepeoplepossess(i.e.their knowledgeis basedon
different underpinning assumptionsand values).Thus, for example,Newell and Swan
r2000) found that the difficulty of knowledge-sharing between different members of the
researchproiect they examined were related to epistemologicaldifferencesin their
knowledge, which stemmed from the different disciplinary backgroundsthey came from.

Table 6.3. Knowledge-related

factorsadverselyaffecting
intercommunityknowledgeprocesses

Limitedamountof commonknowledge
'sticky'
Knowledgepossessedby peopleis anddifficultto share
(highlytacitandcontext-specif
ic)
Fnictomin difforonnoc

(people's
knowledgebasedon differentassumptions,
values)
E S O C I A LA N D C U L T U R A LI S S U E S

The difficulties of sharing sociallyembedded knowledge

Lam (1997)examineda joint technologydevelopmentbetween

a Japaneseand a UK electronics
company While the companieswere competitorsthe Japanese
company had a majorityshare-
holdingin the UK company.However,this collaborative relationship provedproblematic,with staff
frequently referringto, 'probremsof poor communication,
misinterpretationof specifications.
and the clash between their approachesto product development',
with these diffjcultiesbeing
primarilyattributedto, 'differencesin the organization
of knowledgeand work between the part-
ner f irms' (989) Lam foundthe knowledgeof all relevantstaff to
be deeplyembeddedin the social
and organizational context,and that further,the knowledgebaseand organizationat contextof both
divisionswere significantlydifferent.While in the UK companythere
was an emphasison form-
aljzedknowledge,developedthrougheducation,in the Japanese
companytacrtknowledgeaccu-
mulatedthroughexperiencewas more important.Secondly,in the
UK companythere was a clear
demarcationof lob boundaries,limited use of job rotation,and a tendency
for peopleto develop
narrowlyspecialized knowledgebases.ln the Japanesecompanyby contrastoue to the
emphasis
on team-workingthe demarcationbetween jobs was biurred,and
due to the use of job rotation,
people'sknowledgebaseswere typicallybroad.Finally,
there were also signifrcantdifferencesin
the way knowledgewas sharedand developedthroughoutthe product
cycle.ln the UK division,
product design, and the developmentof detailedspecifications
was principalythe domain of
design staff. In the Japanesecompany by contrast production
and design staff both had an
importantrole in the developmentof productspecifications, with this 'interactive'way of working
requiringa significantlevelof 'knowledge-sharing between upstreamand downstreamstaff,(gg0)
These differencestherefore made the process of knowledge-sharing,
and Joint technology
developmentextremelycomplicated.

Thissuggeststhatthe sharingof knowledgebetweenpeoplewith

cultureswhicharequitedifferent
is likelyto be difficult.Froma management
pointof view,what can be doneto addresssuch
problems?

The issueof epistemologicaldifferencesis worth elaboratingon, assuch

differencescan
have a profound effect on attempts to shareor collectively utilize knowledge.
Brown and
Duguid (200I, 2O7)argue that while the advantageof communities
of practice is that
'common
. . practice . createssocial-epistemicbonds,, conversely,,[p]eople with
different practiceshave different assumptions,different outlooks, different
interpretations
of the world around them, and different waysof making senseof their
encounters.,Thus,
people from different communities of practice, or work groups,
may not only have
limited amounts of common, sharedknowledge,but the knowledge
they possessmay be
basedon a fundamentarrydifferent systemof varuesand assumptions.
Suchissuesmay arisein multidisciplinary work (Newell and Swan2000),where
staff from
different organizational subunits require to collaborate (Hansen 1999),
in international
 G W L E D GPER O C E S S E S
B O U N D A R Y - S P A N NKI N O

.ollaborations involving people with significantly different cultures working together

Pauleenand Yoong 2001),where people from different occupational communities lequire
:o shale knowledge (Bechky 2003), or where different organizational functions require to
collaborate(seeFrance-coexample below). The complexity of knowledge-sharingin such
;ircumstances stems from the fact that epistemological differences between people or
basicpre-
aroupscan inhibit the development of even a fundamental understanding of the
nisses,and valuesthat the knowledge of othersis basedon. For example, the feeling of 'cul-
:rre clash,that people can experiencewhen visiting a country with very different cultural
.;aluesstemsfrom difficulties in understanding the basicvaluesunderlying'other' cultures'
\ewell and Swan (2000) suggestthat the gleatel the epistemologicaldifference between
:ollaborating parties,the lesschancethere is that such collaborationswill be successful,and
:1e more likely that they will not be able to effectively integrate their different perspectives
:nd knowledgebases.

France-co:epistemic differencesin cross-functionalcollaboration

:-ance-Co producesspecialistcomponents As partof the company's

for militaryandcivilaircraft.
::.emptsto introduce new management practices following the end of the ColdWar it decided
.: mplementan Enterprise Planning
Resource (ERP) system, which was intendedto improve
This projectrepresented
knowledge-sharing.
,els of interfunctional an enormouschallenge
- France-co, functions
as two of the mostimportant for thisproject,salesandproduction, had
- s:oricallyshared little information.These functionalgroups possessedtheir own specialist
worked
: : Ciesof knowledgeand staff typicallyhada strongsenseof identityf or the f unctionthey
- Further,the knowledgepossessedby staff in these groupswas highlytacit,and was typically

-=velopedthrough practice,over time. Finally,relationsbetween these functionalcommunities

- . J h i s t o r i c a l lbye e na n t a g o n l s t l c .

ls France-Co's ERPprojectdevelopedit becameapparentthat the lackof knowledge-sharing

to
:::ween these communitieswas proving detrimentalto the project Thus, initialattempts
-clement the new Systemproved disastrousand had to be stopped.The main reasonfor this
., Lre was that staff in both the sales and productionfunctionswere not sharingthe type of
,^cwledge and informationthat was necessaryfor the success of the project While this
-: rctance to share knowledgewas partly relatedto the historicalantagonismbetween these
--^ctions, it was also relatedto the specializednatureof the knowledgethey each possessed.
-- had a
s, combinedwith the extensivelackof interactionthat had been typical,meant that they
.:ry poor understandingof how each other worked, or what their constraintsand requirements
,,e:e, Thus,even when staff f rom these communitieswere willingto shareknowledgewith each
of
::^er, effectivelydoing so proved difficult,as each had an extremely limited understanding
',^at knowledgewas relevant,important,or useful to the other'

parties
\\rhen such significant epistemological differences exist it is necessary for the
,rr-olved to develop an improved level of mutual understanding before any knowledge

:an be effectively shared, or collectively utilized (Bechky 2003). From a practice-based

SOCIAL
A N D C U L T U R AI L
SSUES

perspective,developing such an understandinginvolves the sort of perspectivemaking

and taking processesoutlined in Chapter 3. While the practice-basedperspectiveon
knowledge assumesthat processesof perspectivemaking and taking are necessaryfor
the sharing and communication of knowledgein all circumstances,the lack of common
knowledge in intercommunity contexts raisesthe importance of such processes. These
perspectivemaking and taking processes do not result in the integration of the different
knowledgebasesinto a coherentwhole, but should insteadinvolve a processof dialogue,
where 'eachcommunity maintains its own voice while Iisteningto the voice of the other'
(Gherardiand Nicolini 2OO2,42I).Thus, perspectivemaking and taking occursthrough
a processof talking, listening, acknowledging, and being tolerant to any differences
identified.
In conclusion,intercommunity knowledgeprocesses areinhibited by the differencesin
the knowledgepossessed by the people involved in such processes. In generalterms, the
greaterthe degreeof common knowledge that exists,the more straightforwardknow-
ledge processesare likely to be. Further, the characterof knowledge processesin such
circumstancesarealsoaffectedby the degreeof epistemologicaldifferencein the assump-
tions and valuesunderpinning the knowledgebasesinvolved, with a high level of epi-
stemologicaldifferencelikely to significantly increasethe difficulty and complexity of
such knowledgeprocesses.

ldentity, knowledge,trust, and socialrelations

One of the maior conclusionsto emergefrom the previous section was that where the
common knowledgebaseis limited, or where people have a limited senseof sharediden-
tity this means that the socialrelationship betweenpartiesis unlikely to be strong, and
that the foundations for the existenceof trust are relatively weak. Thus in such circum-
stancesnot only is the existenceof strongtrust unlikely, but the developmentof trust will
typically be complicated and difficult. Fundamentally,the level of trust and mutual
understanding between people who do not normally work together and who are not
membersof the samework group or community of practiceis likely to inhibit the sharing
and collectiveutilization of knowledge,aswas discussedin Chapter 4.
The importance of trust in thesesocialcontexts,combined with the complexity of the
concept of trust, means that it is worth elaboratingmore on the topic. Analysesof trust
show it to be a theoreticallycomplex conceptwhich has multiple dimensions(Lane1998;
Newell and Swan 2000; Zucher 1996).Thus, most analysesof trust outline a number of
different types of trust (seeTable 6.4). Further,this work showsthat thesetypes of trust
are distinctive in character, are developed in quite different ways, and have a complex,
mutually interdependentrelationship.
The limited basis for trust which exists in intergroup contexts, and particularly for
newly formed intercommunity project groups can be seenfrom any of the three tlpolo-
gies of trust described.Thus the nature of the social retationship between people in a
newly formed intercommunity work context precludes the existence of what Zucker
(1996) referred to as process-based, and characteristic-based trust, what Lane (1998)
 BOUNDARY-SPANNIN
KGN O W T E D G EP R O C E S S E S

Table6.4. Typologies
of trust

Author/s Typeof trust Descriptionof trust

- - c k e r( 1 9 8 6 ) Processbased Basedon experience,

and builtup overtime
Characteristic
based Basedon socialsimilarities
andculturalcongruence
Institutional
based Trustbasedon institutional
or professional
reputation,
NOTinterpersonalfamiliarity
-,^e (1998) Calculative Trustbasedon someform of caiculation
regarding
costs/benefjts
Norm/value
based Trustbasedon commonsocialvalues
Cognitlve/expectation Trustbasedon commonexpectations aboutfuture
oaseo events,and/orpatternsof behaviour
. : . 1 / e ta n o Companion Trustbasedon judgementsof goodwillof friendship,
: . , a n( 2 0 0 0 ) builtup overtime
Competence Trustbasedon perception of otherscompetence
to carryout relevanttasks
Commitment Truststemmingfrom contractual
obligations

-:erredto asvalue-or expectation-based trust, and what Newell and Swan(2000)refeired

ascompanion- and competence-based trust. Thus in suchcircumstances, the onlybasis
: trust is the most impersonal,and arguablyweakesttypes of trust (institutional-based
. Zucker'sterms, calculative in Lane'sterms, and commitment-basedin Newell and
,"'-an's
terms).Thus, aswill be discussedin the following section,one of the main waysto
,:.litate the development of knowledge processesin intercommunity work contexts is
-:ough the development
of trust basedon better mutual understanding and stronger
:;al relations.

you havewith a rangeof people.To what extentaretheserelationships

.'::t on the relationship
'.-: on differenttypesof trust?Further,
how doesthe levelandtype of trustyou havein different
- -: e affectthe amountandtypeof knowledgeand informatron you sharewith them?

Facilitating/managingknowledge between communities

- ro this point the chapter has emphasizedthe not insignificant difficulties in the
: :.tive, collectiveutilization of knowledgein intercommunity work groups.However,
' =>edifficultiesarenot insurmountable.
Thus, there is much that can be done to address
' ' : rll, drld increasethe chance
of intergroup work processes effectivelymaking collective
. of their knowledge. In generalterms, this involves improving the level of mutual
- :erstanding and developing
the socialrelationship between relevant people.Current
:ing suggests two broad waysin which this can be achieved.First,work can be invested
@ SOCIAL
A N D C U L T U F AI LS S U E S

in managing the social relationship between people, and secondly, developing the
existing areasof overiapbetweenpeople.

Relationshipmanagement

Relationshipmanagementinvolvesattempting to developthe socialrelationshipbetween

the people involved in an intercommunity work group to becomelessbasedon the most
impersonai,and relativelyweakforms of trust outlined above.In Newell and Swan,sterms,
this involvesmoving away from a relationshipbasedon commitment trust, to one where
competenceand companion typesof trust aredeveloped.However,successfully achieving
such a transition is by no means straightforward.Primarity, the development of these
more personaltypes of trust involves group membersdevelopinga greaterlevel of sens-
itivity to and understandingof the knowledge,values,and assumptionsheld by other
membersof the samework group. This requiresthe processes of perspectivemaking and
taking outlined earlierin the chapter,which requiresall partiesto both talk and listen to
each other. However,the more limited the amount of shared,common knowledge,and
the greaterthe levelof epistemologicaldifferencein the valuesand assumptions,the more
time-consumingand complicatedthis processis likely to be. Further,to be effectivesuch
processes may well requirea certainlevel of face-to-face interaction (Bechky2003).This is
because,as Lam (1997, 992) suggestseffectivecollaborationin this context requiresthe
development of, 'direct and intimate social relations . tas] . . . Iearnerswill need to
become"insiders"of the socialcommunity in order to acquireits particularviewpoint.,
Brown and Duguid (1998)identifled two rolesthat key individuals could take in the devel-
opment of intercommunity social relations: brokers and translators.The brokering role is
relevant where there is some pre-existing overlap in the knowledge of the communities/
people involved. A broker is someone who inhabits both communities, and uses their
knowledge and understanding of both to facilitate the development of mutual understand-
ing between other membersof the communities. Gherardi and Nicolini (2002)arguethat a
broker is someonewho has the ability to, 'transfer and translate certain elements of one
practiceto another'. The role of translator is relevantwhere there is no overlapping common
knowledge between communities/people. This requires the translator to have a detailed
knowledge of both communities, and further, the translator requires to be trusted by the
members of both communities as they play such a key role in interacting between them.
Suchroles are acknowledgedto be extremely complex and difficult to successfullymanage.

How lmportantis face-to-face

interaction
for the development of trustandan effectiveworking
relationship
betweenpeoplefrom significantly differentcultures?Cancross-culturalworkinqrelations
be developed withoutanyface-to-face
interaction?

Boundary objects

The third and final method discussedby Brown and Duiguid (1998) to facilitate inter-
community knowledge-sharinginvolves the development and utilization of boundary
 B O UN D A R Y - S P A N N IKNNGO W L E D GPER O C E S S E S

Dbjects.Boundary obiectsare entitiesthat are common to a number of communities and

can be either physical or linguistic/symbolic in character.Boundary objects provide a
iocus for negotiation, discussion, or even shared activity betr,veenpeople from different
communities, and thus can be utilized to help develop and improve the working rela-
rionship between people, and the mutual understanding they have of each other. One of
the most common type of boundary objectsmentioned by Brown and Duguid are con-
tracts, which typically provide a focus for intercommunity negotiation, and which can
help provide an initial stimulus to a processof perspectivemaking and taking at an early
stagein the working relationship of an intercommunity work group.
Gherardi and Nicolini (2OO2)examined a building site, focusing on how safety issues
n'erejointly negotiated by the three communities of practice with some responsibility for
and involvement in safetyissues.Thesethree communities were engineets,site foremen,
and main contractors.Boundary objects in this context included the physical site that
e\.eryoneworked on, the building under construction,aswell asthe assortedrangeof raw
materialsthat were used, and which were dotted around the building site. Howeveq there
\rere someequally important linguistic boundary objects,such asthe term 'safety'itself.
One of the main ways that relationsbetween thesethree communities were developed,
and negotiations of how safetywas managed on site occurred was through discussionand
negotiation over these boundary objects, which provided a common focus which
brought the communities together.

UK-Pension:boundary objects and brokers

-rs outlinedin Chapter5 (seep. 63), UK-Pension had traditionally been structuredinto two
tiscretedivisions that operatedwith suchsignificant levelsof autonomythat they constituted
separate and distinctcommunities of practice. As partof a majorrestructuring processwhich
cegan in the mid-1990s UK-Pension attempted to move towards a more integrated structure,
,vithgreaterlinksbetweentheirtwo mainbusiness areas:lifeassurance andpensions. Onekey
,vaythis was donewas throughsettingup a cross-business callcentre.Thiswas a singlecall
lentrethatwouldhandleworkfrom bothbusiness the callcentrewas staffedby
areas.lnitially,
ceoplefrom bothdivisions, with the leaderof the callcentreimplementation projecthavingthe
'oleof persuading staffto work in the centre.The projectmanager therefore was in the roleof
croker,and the callcentrerepresented a (new)boundary object.Whilethe callcentrewas a
coundary objectcommonto bothcommunities, andwhichwouldprovidea physical sitewhere
stafff rombothdivisions wouldworktogether, staffwereunfamiliar with it. Further,it represented
a radical changein workingpractices for UK-Pension, notonlybecause it required stafffromboth
Civisions to worktogether, but becauseit was the firstlarge-scale useof a callcentrewithinthe
:ompany. Therefore,the brokering roleplayed by the call
centre project manager in communicat-
ng the purposeof the callcentreto staff,andpersuading someof themto work in it was key.ln
:heend,the projectmanager was successf ul in hisbrokering role,as hewas ableto persuade an
adequate numberof staffto changejobsandwork in the callcentre.
E S O C I A LA N D C U L T U R A LI S S U E S

Conclusion

This chapter narrowly focused on cross-community,boundary-spanning knowledge

processes.Arguably, the relevance and importance of cross-community knowledge
processeshas increaseddue the changesin working practicesthat have emergedfrom
the contemporaryrestructuringof work organizations.The differencebetweenintra- and
inter-community knowledgeprocesses relatesto the senseof sharedidentity and typically
high level of common knowledgewhich existswithin communities (seeChapter 5), but
which is relatively absent from intercommunity contexts. Further, it may also be the case
that not only are there limited amounts of common, sharedknowledgebetweenparties,
but that there may be epistemicdifferencesin the knowledgeof the peopleand commun-
ities involved, where their knowledgeis basedon fundamentally different assumptions
and values.
Typically, as illustrated by a number of examples,intercommunity knowledge pro-
cessesare likely to be more complex and difficult to make successfulthan intracom-
munity processes.This is due to both the differencesin identity, which may induce
intercommunity conflict, and the lack of common knowledge.Somewhatsimplistically,
the lesscommon knowledgethat exists,and the greaterthe level of epistemicdifference,
the more complicatedand difficult the knowledge-sharingprocesswill typically be.
Knowledge-sharingacrosscommunities was shown to requiretwo primary, and closely
interrelatedelements,both of which aredevelopedthrough a processof socialinteraction
and communication. First, an adequatelevel of trust requiresto be developedbetween
the individuals from both communities,ideally with the strongestforms of personaltrust
being developed.This gzpeof trust has been variously labelledas process-based (Zucker),
cognitive (Lane),and companion trust (Newell and Swan).Secondly,people from both
communities require to develop a basic understandingof the values,assumptions,and
viewpoints which underpin each other's knowledge base.This processof perspective
making and taking, which was also examined in Chapter 3, requiresnot a merging of
these different knowledgebases,but an appreciationof, sensitivity to, and toleranceof
the differencesin perspectivewhich emerge.
Finally,the chapterexaminedthe waysin which intercommunity knowledgeprocesses
can be facilitated, through brokers/translatorsattempting to bridge communities and
develop relations between them, and through the use of boundary obiects that are
common to all relevantcommunities.

REVIEW
QUESTIONS

t Theprevalence of interorganizational networking canbe gaugedby a simplepieceof

research f r o ma n ys e r i o uds a i l yn e w s p a p earn dy o ua r el i k e l y
E.x a m i nteh eb u s i n e ssse c t i o n
to findrelevant examples. However, is thistypeof workingpractice likelyto be more
commonin somebusiness sectorsmorethanothers?What factors affectthe extentto
whichinterorganizational networksaredeveloped andutilized?
 N O W L E D G EP R O C E S S E S
B O U N D A R Y - S P A N N I NKG

2-:eorySuggeststhatmoreimpersonaIformsoftrust,suchascommitment-basedtrustor
^stitutionallybasedtrust are typicallyweaker and more f ragilethan trust developedthrough
arongoingsocralrelationship,suchasprocess-basedtrust.Doesthisref|ectyourown
experience?
you and your
g leflect on any work experiencethat you have had.To what, if anythingdid
,vorkcolIeaguesmostStronglyfeelasenseofidentityasbeingpartof:yourimmedlatework
group,thefunctionyouworkedin,thedivisionyouworkedfor,ortheoverallcorporate
of an effectiveworklng
group?Are these sensesof identltylikelyto inhibitthe development
.elationship,and the sharingof knowledgewith peoplefrom differentpartsof the
?
crqanization

l
F U R T H ERRE A D I N G
'Knowledgeand Organization:
A SocialPracticePerspective"
o J. Brownand P Duguid(2001).
OrganizationScience,1212:198-213'
A|argetytheoreticat,butwe||writtenandaccessib|epaperwhichref|ectsonthewhatmakes
in tercom m unity knowl edg e-sharing difficult'

rA.Lam(1997).'EmbeddedFirms,EmbeddedKnowledge:ProblemsinCollaborationand
organizationstudies,l 8/6: 973-96
KnowledgeTransferin Globalcooperativeventures"
difficulties of knowledge sharing within
A theoretically groundedcase studywhich examines the
an internationalProiect team
'Trustand lnter-Organizational
Networking"HumanRelattons'
r S. NewellandJ. Swan(2000).
531101281-1328
on the role of trust in shaping the
An empiricaltyrich andtheoreticallyinnovate casestudy
proiect
dynamics of a multi-disciplinary team
of lnterconnected
in a Constellation Practices:Canon
and D. Nicolini(2002).'Learning
r S. Gherardi
' Management Studies'3914 419-36
or Dissonance?Journalof
intercommunitysense-maktng
Examinestherole of boundaryobiectsand brokersin facilitating
and worKng