Perspective Making and Perspective Taking in Communities of Knowing.1995

8/12/2019 Perspective Making and Perspective Taking in Communities of Knowing.1995

1/24

Perspective Making and Perspective Taking in Communities of Knowing

Author(s): Richard J. Boland, Jr. and Ramkrishnan V. TenkasiSource: Organization Science, Vol. 6, No. 4 (Jul. - Aug., 1995), pp. 350-372Published by: INFORMSStable URL: http://www.jstor.org/stable/2634993.

Accessed: 19/04/2014 13:41

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at.http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

INFORMSis collaborating with JSTOR to digitize, preserve and extend access to Organization Science.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=informshttp://www.jstor.org/stable/2634993?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/2634993?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=informs


2/24

erspective a k i n g n d erspectiveT a k i n g in ommunities o nowing

RichardJ. Boland,Jr. * RamkrishnanV. TenkasiWeatherheadSchool of Management, Case WestemReserve University,Cleveland, Ohio 44106-7235School of BusinessAdministration, University f Southem Califomia, Los Angeles, Califomia 90089-1421

AbstractKnowledge-intensiveirmsare composedof multiplecommu-nities with specializedexpertise,and are often characterizedby lateral rather than hierarchical rganizationalorms. Weargue that producingknowledge o create innovativeprod-ucts and processes n such firmsrequires he ability o makestrongperspectiveswithin a community, s well as the abilityto take the perspectiveof another nto account.We presentmodels of language,communication nd cog-nition that can assistin the design of electroniccommunica-tion systems for perspectivemakingand perspective aking.By appreciating ow communications both like a languagegame playedin a local community nd also like a transmis-sion of messages hrougha conduit,andby appreciating owcognition ncludesa capacity o narrativize ur experienceaswell as a capacity o process information,we identifysomeguidelines or designingelectroniccommunication ystems osupport knowledgework. The communication ystemswepropose emphasize hat narratives an help construct trongperspectiveswithina community f knowing, ndthat reflect-ing uponandrepresentinghatperspective an createbound-ary objectswhich allowfor perspective akingbetween com-munities.We conclude by describingour vision of an idealizedknowledge ntensive irmwitha strongcultureof perspectivemakingand perspective aking,and by identifying ome ele-ments of the electroniccommunication ystemswe wouldexpectto see in such a firm.(Knowledge Work; Organization Leaming; DistributedCognition;CommunicationSystems)

IntroductionOrganizations are developing innovative products andservices on faster cycle times (Purser and Pasmore

1992,Lawler 1992),causing an increasein knowledgework (Pava 1983) and a gradualreplacementof capitaland labor intensivefirmsby knowledge ntensivefirms(Starbuck1992).Knowledgework creates new under-standingsof nature,organizations r marketsand ap-plies them in valued technologies,products,or pro-cesses. Knowledge intensive firms are composed ofmultiplecommunitieswithhighlyspecialized echnolo-gies andknowledgedomains Purseret al. 1992).In thepharmaceuticalndustry, or example, developingnewproducts requires integrationof knowledge from abroad array of disciplines such as molecularbiology,physiology,biochemistry, yntheticchemistry,pharma-cology and even esoteric specialties such as molecularkinetics Henderson,1994).A similarpattern s observ-able in other industries.The first generationof cellulartelephones used five sub technologies, but their thirdgeneration incorporated fourteen distinct sub tech-nologies(Granstrand t al. 1992).The increasingpro-liferation of specialized and distinctknowledgecom-munities and the need for their integrationhas alsoresulted n the emergenceof neworganizationalorms,among them the lateral-flexible orm of organization(Galbraithand Lawler1993,Galbraith1994).The lat-eral-flexibleorganizationalorm relies on peer-to-peercollaboration as opposed to a vertical hierarchy) nachievingorganizational bjectives.It is ourcontention hat all organizations re becom-ingmoreknowledgentensiveacross he service, ndus-trial and governmental ectors. It is easiest to see thefundamental mportanceof knowledgework in firmsinvolved with new product development in leadingedge technologies,but the relentlesspace of change nmarket expectationsmeans that all organizationswill

1047-7039/95/0604/0350/$01.25Copyright ? 1995. Institute for Operations Research350 ORGANIZATIONCIENCE/VOl. 6, No. 4, July-August 1995 and the Management Sciences

This content downloaded from 131.123.1.227 on Sat, 19 Apr 2014 13:41:00 PMAll use subject to JSTOR Terms and Conditions
http://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/page/info/about/policies/terms.jsp


3/24

RICHARD J. BOLAND, JR. AND RAMKRISHNAN V. TENKASI Communities of Knowing

increasingly ely on creatingnew knowledgeand adopt-ing lateral organizational orms. The major issue forsuch firms s to find creativeways for representingandintegratingknowledgeacross their lateralunits (Weickand Roberts 1993,Galbraith1994).Knowledge production involves communicationwithin and between a firm's multiple communitiesofknowing.We refer to communication hat strengthensthe unique knowledgeof a communityas perspectivemaking,and communication hat improves ts ability otake the knowledgeof other communities nto accountas perspective aking.In this paperwe employ modelsof language,communication nd cognitionto proposehow electroniccommunication ystemscanbe designedto support perspectivemaking and perspectivetakingin knowledge ntensivefirms.We argue that perspective makingand perspectivetakingare achievedby narrating ur experienceas wellas by rationallyanalyzing t. These processesare likeplaying games with language as well as like transmit-ting messages through a conduit, and' they involveheightened evels of reflexivity.The narrating f expe-rience is a critically mportant but often overlookedelement of knowledgeproduction n knowledge nten-sive firms,even though it is recognizedthat scientificreasoning is often conducted throughnarrativesandthat scientists' nterpretivepracticesare embodied intheir conversation Knorr-Cetina1981, Mulkayet al.1983). Similarly,the importanceof playful situatedaction for strengthening ne'sownperspectiveand theimportanceof reflexivity or appreciatinghe perspec-tiveof another s not sufficiently ecognized n commu-nicationsystemdesign. After analyzing he processesof perspective makingand perspectivetaking,we de-scribe an idealized knowledge-intensiveirm to high-lightsomeof the featuresof an electroniccommunica-tion systemthat wouldsupport hose processes.We first present the concept of a communityofknowingas anopensystemandprovidea briefoverviewof models of language,communicationand cognitionthat can guide the designof electronic communicationsystems.The languagegames model of Wittgenstein(1974) andBruner's 1986, 1990)model of narrationasa cognitivemode are presentedas supplements o thedominantorganizationalmodels of languageas mes-sage transmissionsand cognitionas informationpro-cessing. Science is used as an exampleof knowledgework to draw mplicationsor applying hese models tothe communication equirements n knowledge-inten-sive firms. We then explorethe dynamicsof perspec-tive makingandperspective akingand some potentialbreakdownsn the perspective-taking rocess.This al-

lows us to summarize he strengthsandweaknessesofthe languagegameand the conduitmodels of commu-nication for designingelectronic communicationsys-tems, and to emphasizethe importanceof reflexivityand boundaryobjects n perspective aking.Finally,wepresent our admittedlyutopianvision of some futureapplicationsof electroniccommunicationor support-ing knowledge work in a hypothetical firm with alateral-flexibleorm.

Communitiesof Knowingas OpenSystemsOrganizationsare characterizedby a process of dis-tributed cognition in which multiple communitiesofspecialized knowledgeworkers, each dealing with apart of an overallorganizationalproblem,interacttocreate the patternsof sense makingand behaviordis-played by the organizationas a whole (Boland et al.1994).Organizationsare necessarily characterizedby dis-tributed cognition because their critically importantprocessesand the diversityof environmentsand tech-nologies to be dealt with are too complex for oneperson to understand n its entirety Brehmer1991,p. 4; Nersessian1992).Thisproblem s especiallyacutein knowledge-intensiveirms that rely.onmultiplespe-cialtiesand knowledgedisciplines o achievetheir ob-jectives. Each such communityof specializedknowl-edge workers s whatwe term a community f know-ing.A number of scholars such as Fish (1980), Fleck(1979), Barnes (1983) and Brown and Duguid (1991)havecommentedon the way thatcommunitiesdevelopunique social and cognitive repertoireswhich guidetheir interpretationsof the world. Fleck's(1979) con-cept of thoughtcollective s one such notion thatemphasizes the unique interpretiverepertoiresof adistinctcommunity f knowing.A thoughtworldevolvesin a community f knowingas a readiness or directedperception .Thought worlds with different funds ofknowledgeand systemsof meaningcannoteasilyshareideas, and may view one another's central issues asesoteric, if not meaningless.Other terms which echoour concept of communityof knowing nclude inter-pretive community Fish 1980), contextof learning(Barnes1983),and community f practice LaveandWenger 1990, Brown and Duguid 1991, Orr 1990).However,givenourfocuson knowledge-intensiveirms,and our concern with the interactionof differentex-pert knowledge groups in the process of knowledge

ORGANIZATIONCIENCE/Vol. 6, No. 4, July-August 1995 351



4/24


5/24


6/24

RICHARDJ. BOLAND, JR. AND RAMKRISHNAN V. TENKASI Communitiesof Knowing

Table 1Key assumptions behind conduit model of communication andparadigmatic mode of cognition.

* There is underlying objective knowledge in the world that hasuniversal applicability.

* Language can be a medium for representing objective knowl-edge and words have fixed meaning.* Human beings can achieve universalityof understanding since

fixed meanings of words can be communicated objectively from oneperson to another.

* Realizationof objective knowledge is a rationalprocess. Knowl-edge evolves and progresses through the systematic application oflogic and principles of the scientific method.Key assumptions behind language games model of communica-tion and narrative mode of cognition.

* Knowledge as well as methods for realizing knowledge areobjective only to the extent they are ratifiedas objective by a specificcommunity's interpretiveconventions.

* Words can have consensus of meaning only within a specificcommunity of knowing. However, even within a unique community,the meaning of words change and are never fixed in time or space.

* Language is not a medium for representing our thoughts andobjective underlyingknowledge but language is thought and knowl-edge. The limits of our language are the limits of our knowledgesince we can explain the world only through language and narrativeforms.

* Knowledge evolves by inventing new language and narrativeforms. Re-narrativizing he familiar or coming up with narrativesthatexplain the unfamiliar s the primaryactivity by which new knowledgecomes about.

cognition being drawn upon, they will not be com-bined. Whereas the language games model is philo-sophical and proposed as a more accurate depiction ofhuman language and communication, the conduitmodel is technical and proposed as the necessary re-quirements for a communication system. One focuseson human language, the other focuses on communica-tion technology. The narrative and paradigmatic modesof cognition, on the other hand, are meant by Bruner(1990) as complementary functions of the same whole,each being capacities of the human being. We will notcollapse them, and will treat all four as having aseparate tradition and use. We will employ each toserve different purposes in understanding the pro-cesses of perspective making and perspective takingamong communities of knowing.We will now look to science for insights on howcommunities of knowing develop and change throughcommunication. We will then draw implications fordesigning electronic communication systems for knowl-

edge work in organizations, and also consider howmultiple communities of knowing interact.

Science as Knowledge Workin Communities of KnowingConsidering science as organized knowledge work hasmany insights to offer for understanding perspectivemaking and perspective taking in knowledge intensivefirms and for speculating on how electronic communi-cation can be designed to support knowledge work inlateral organizational forms. A central source for theseinsights is provided by Thomas Kuhn (1970) as hedescribes the historical process of scientific work.Readers are no doubt familiarwith Kuhn's argument ofhow normal science within paradigms leads to crisisand revolution. For Kuhn, a paradigm is a shared senseof what the metaphysical nature of the world is, whatproblems are important, and what serve as good exem-plars for a domain of concern.There are many difficulties with Kuhn's notion ofparadigm. It is often taken to be totalizing, unitary andalmost religiously held. As Masterman (1970) has noted,Kuhn (1962) used the term paradigm in many differentways in the first edition of The Structure of ScientificRevolutions. In the revised edition Kuhn (1970) ac-knowledged the concept's ambiguity and added furtherrefinements, but debates about just how a paradigm isto be defined or isolated for further study in its ownright will not concern us here. We believe his basicinsight is valid, and is in keeping with Polanyi's (1967)idea of tacit knowledge , Boulding's (1956) discussionof the image , Pepper's (1942) notion of world hy-potheses and numerous others who point out thatperception is only accomplished through a perspective(Burrell and Morgan 1979, Bartunek 1984).Kuhn's (1970) insights are particularly relevant forunderstanding how knowledge is produced in a com-munity of knowing by refining and clarifying the per-spective of the community. Development of knowledgein a community is a process of posing and solvingpuzzles, thereby elaborating and refining the vocabu-lary, instruments and theories that embody the per-spective. Agreement that knowledge is progressing isagreement that the perspective is strengthening. Unex-pected events or findings can only be recognized assuch from within a perspective. Without a strong per-spective the community cannot tell an anomaly fromnoise; a challenge to their knowledge from an irrele-vancy.Collins (1983) makes some interesting observationson the dynamics of knowledge development and the

354 ORGANIZATIONCIENCE/Vol. 6, No. 4, July-August 1995



7/24

RICHARD J. BOLAND, JR. AND RAMKRISHNAN V. TENKASI Communitiesof Knowing

different kinds of competence required of the scientist.Working within a perspective has well establishedmethods for externalizing its objects, and the scientistshould be competent in those respects. Collins termsthis native competence . It is the kind of competencethat makes meanings, perceptions, and acts of thenative member follow naturally as a matter of course.However, changing or overturning the taken for grantedrules or replacing them with a completely new setrequires interpretive competence on part of the sci-entists. It lies in perspective taking: being able toreflect upon and renarrativize the familiar to open upnew insights and understandings.The stronger and more well developed a community'sperspective is, the more useful a conduit model ofcommunication and feedback becomes. As theories,puzzles, measures and accepted results are clarifiedand institutionalized within the community, the morelikely it is that messages can be thought of as selectionsfrom a predefined set. The process by which newcommunities of knowing begin to form, however, andthe processes of questioning and changing perspectivesis not as well handled by a conduit model. Work thatquestions a perspective is of a different logic type thanwork within a perspective, and is primarily controlledby the dynamics of change in an open system ratherthan simple feedback (Wiener 1954, von Bertalanffy1968, Bateson 1972). For this second-order knowledgework, the language games communication model ismore helpful than the conduit model. Previously ac-cepted understandings, measurements, and logics arein a sense up for grabs . The perspectives behindways of knowing of the organizational communities arebeing made in real time by the communities' members.The language of their communication is changing astheir practices in forms of life are changing. Messagescannot be separated from the evolving context of mak-ing and using them as in the conduit model.Two final themes from Kuhn that we will considerbefore drawing implications for knowledge work inorganizations are the incommensurabilitybetween per-spectives and the emergence of new perspectives. Ifmembers of a community create a strong perspectiveand do distinctive and important knowledge work, itwill of necessity approach becoming incommensurablewith other perspectives. They may use the same wordsas other communities of knowing, but they will usethem to see things in different ways (Knorr-Cetina1981). They will look at the same phenomena asanother community, but will see different problems,different opportunities, and different challenges(Czarniawska-Joerges 1992). As Kuhn puts it, they will

live in a different world from those in other communi-ties of knowing. Data important to one are irrelevantto another, or are used for entirely different purposes.Arguments that persuade convincingly in one commu-nity of knowing have little or no weight in another.And the more developed and refined the community ofknowing becomes, with an increasingly elaborate anddetailed perspective, the more nearly incommensurableit becomes with others (Fleck 1979, Brown and Duguid1991, Dougherty 1992). If the members' language gameswithin one community of knowing fully understood andappreciated the positions of another, they would not bedifferent communities and would not be doing distinctknowledge work.Knorr-Cetina (1981) presents some grounded exam-ples of how local communities of knowing developtheir unique paradigmatic worlds and are resistant tochanging them. In her sociological study of differentresearch units, she found that research laboratoriesdeveloped local interpretations of methodical rules, ora local know-how with regard to how to make thingswork best in actual research practice. Criteria for whatmattered and what did not matter were neither fullydefined nor standardized throughout the research com-munity. Nor were the rules of official science exemptfrom local interpretations. Many important selectionsof the research process were locally driven, includingquestions of ingredients, instrumentation, and durationof experiments.Implicationsof Kuhn and KnowledgeWorkin Sciencefor UnderstandingKnowledge-intensiveFirmsA first implication of Kuhn for thinking about knowl-edge-intensive firms is that the primaryunit of analysisshould be the community of knowing. The individualdoes not think in isolation and is not an autonomousorigin of knowledge. A community of knowing is alanguage game and neither the language nor theknowledge created within it comes from the actoralone.Secondly, a community of knowing requires perspec-tive making in order to do knowledge work. Without astrong perspective it cannot produce important knowl-edge. A community's perspective develops by refiningits vocabulary, its methods, its theories and values andits accepted logics through language and action withinthe community of knowing. This means that the com-munity must, of necessity, have a space for conversa-tion and action isolated from the larger organization.Thirdly, the ability of one community of knowing towork jointly with another requires an ability to over-come the degree of incommensurability between them.

ORGANIZATION CIENCE/VO1.6, No. 4, July-August 1995 355



8/24

RICHARD J. BOLAND, JR. AND RAMKRISHNANV. TENKASI Communitiesof Knowing

This, of course, must be done without sacrificing theintegrity and distinctiveness of their own perspective.Below we will explore this process of perspective takingin which the perspective of another can be taken intoaccount as part of a community's way of knowing.Fourthly, the conditions for change in the perspec-tive of a way of knowing come from both the inside andfrom the outside. Inside the perspective, conditions ofchange come from the accumulation of anomalies as itis tested and elaborated. From outside the perspective,pressure for change comes from adherents drawn to apromise of the aesthetics, power or excitement of anew perspective. This suggests that memories of errorsand anomalies are important to maintain and reviewopenly, and that the isolation of communities necessaryfor their development should be punctuated by periodsof interaction between communities.Finally, new perspectives need to be nurtured andgiven protection from strong demands for perfor-mance. Of necessity, they will not be able to competewith an established perspective in another community'sway of knowing.For a knowledge-intensive firm, then, we look to itsecology of communities of knowing to understand itspossibilities for doing knowledge work. Electronic com-munication can mediate how the open system of com-munities emerge, develop, elaborate, suffer crisis, andtransformwithin it. Electronic communication can alsomediate how communities of knowing interact andtheir capacity for perspective taking. It is to theseprocesses of perspective making and perspective takingthat we now turn.

Perspective Making and PerspectiveTaking in Communities of KnowingThe Process of PerspectiveMakingPerspective making is the process whereby a commu-nity of knowing develops and strengthens its ownknowledge domain and practices. As a perspectivestrengthens, it complexifies and becomes better able todo knowledge work. Complexification is achieved cog-nitively through the use of paradigmatic analysis withina narrative framing of experience. It is a process ofdeveloping finer language games, and from a paradig-matic standpoint, more precise causal laws. Complexi-fication signifies a movement from a global, undif-ferentiated naming to a more precise explication ofconstructs, where more coherent meaning structuresare developed than preceding ones (Waddington 1957,Werner 1957). Knorr-Cetina (1981) proposed that sci-

entific conceptual systems have to progressively com-plexify themselves over a period of time to successfullysolve scientific problems. This implies the ability torespond to shifts and fluctuations in the novelty of thescientific problem domain by modeling the shifts them-selves (Rubinstein et al. 1984).A good example of complexification in perspectivemaking is presented by Bradshaw (1992) in his analysisof the Wright brothers' invention of the airplane. Healso illustrates the interweaving of narrative framingand paradigmatic analysis in the perspective makingprocess. Bradshaw asks why were the Wright brothersso successful in conquering the challenge of mannedflight, while many of their competitors with bettertraining and resources failed? He answers that first,the Wright brothers narratively framed the phe-nomenon of flying using a different metaphor thantheir competitors, and second, they employed finerproblem solving procedures. Whereas their competi-tors narrated flight with a chauffeurs of the airmetaphor, telling how flying was akin to driving a carinto the air, a group that included the Wright brothersnarrated flight as being like flying a kite . Many ofthe unsuccessful inventors had a propensity to con-struct complete aircrafts and then to test them bymeasuring distance and time in flight. To these design-ers, the airplane as a vehicle to be chauffeured was anassemblage of parts (wings, fuselage, propulsion, etc.)and developing an aircraft meant exploring possibledesigns for configuring these parts.However, for the Wright brothers, the major concernwas to understand how a kite flew, and to achieve itsfunctions (lateral control, sufficient lift, reduced drag,etc.) in the airplane. They first isolated these func-tional problems and then proceeded to solve them oneat a time. The pattern in their work was to exploresolutions to subproblems using directed experiments.For example, a kite was built to explore lateral controland wind tunnel experiments explored lift and thrust.Through extensive testing of models, the Wright broth-ers discovered an important error in aerodynamicsoverlooked by other investigators (Bradshaw 1992,pp. 246-247). Only when each separate problem wasunderstood and solved did the Wright brothers investtime and energy in building a new craft. The Wrightbrothers employed both narrative and paradigmaticmodes of cognition in their perspective making, as theymodeled and developed more complex and finer un-derstandings of the workings of aerodynamic laws. Incontrast, their competitors were exploring the possibil-ity of flight with minimal understandings of aerody-namic laws, and relied on trial and error, hoping one of

356 ORGANIZATIONCIENCE/VO1.6, No. 4, July-August 1995



9/24


their models would fly, without having any conceptionof why. They lacked the strong perspective necessary todo important knowledge work.

The Importance of Narrative in Perspective MakingPerspective making within communities of knowing is asocial practice in a form of life. For insight into howthis process takes place in a community of knowing, wewill return to Jerome Bruner's work on the role ofnarrative in constructing knowledge of self and world(Bruner 1986, 1990). Bruner argues that we must lookto how actors make meaning of their experiencethrough narrative if we are to understand the processof perspective making. Bruner, synthesizing studies ofchild development, language acquisition and conceptformation, proposes an innate narrative capacity as theengine for our cognitive activity. The typical form offraming experience (and our memory of it) is in narra-tive form. What does not get structured narratively islost in memory. (Bruner 1990, p. 56) Paradigmaticthinking is an important part of our cognitive reper-toire, but only a part. Narrativizing our reflexive moni-toring and rationalization of conduct is not ruled by anabstracted logic. Within a community of knowing, anarrative explanation works not only because it is logi-cally acceptable, but also because it is lifelike andplausible; it fits the culturally bound demands of aform of life.In parallel with Giddens' structuration theory(Giddens 1976), Bruner emphasizes that when we nar-rativize experience, we also construct and validate theself. The narrator'sperspective as an essential elementin any story assures this. The self is always at stake inthe individual'snarrativizingof experience, because theself is at least the narrator (recognizing the canonical,indicating and explaining the anti-canonical, determin-ing how the world should be) and often part of thestory (being herself delineated as a causal agent withmotives, intentions and values).The importance of narrative has not gone unnoticedin organizational research. Clark (1972) explored theimportance of sagas and Mitroff and Kilmann (1976)recognized the importance of myth. Myth and saga areimportant, but they can distract our attention from theway that human cognition operates almost continu-ously in a narrative, storytelling mode. We wish toemphasize that narrative is fully equal to paradigmaticanalysis in the construction, maintenance and changeof perspectives in an organization. We see them in atype of figure-ground relation in which paradigmatic,rational-analytic thought takes place in a context pro-

vided by narrative, and narratives are constructedagainsta backdrop f paradigmaticnderstandingsn akind of genuineunion of the two modes (Boland andPondy 1983). The rational analyticelements of a per-spective in a communityof knowingare a productofstorytellingas much as they are a mediumfor it.More recently, he role of stories and storytelling nthe day-to-day unctioning of organizationshas beenaddressedby Boje (1991). The constructive, hangingqualityof stories documentedby Boje in his focus onsituatedpractice s a major tep toward he positionwearguefor here. He moves beyondthe mythic view ofthe storyas an object ,foundn Martinand Meyerson(1988), McConkieand Boss (1986),and Gabriel 1991),and turns our attention to the communitydependentprocess of producing he story.When scientists experienceanomalieswithin a per-spective they often turn to narrative n an attempttomake sense of the noncanonicalobservation.Science,and scientific papers documentingexperimentsandtheories, in retrospect,alwaysseem paradigmatic,in-ear and certain.This is partlydictated by the socialconventions of what good science is (Knorr-Cetina1981). However,an examinationof the informal dis-course of scientistspresentsanotherpicturealtogether.Highlyvariableand inconsistentaccountsof actionandbelief are very much the norm. Actors continuallyconstructand reconstruct he meaningof their scien-tificworldthrough he formulationof divergentnarra-tive accounts. As Mulkay et al. (1983) summarize:Unless we understandhow actors sociallyconstructtheir accounts of actionand how actors constitutethecharacterof their actions primarilyhroughthe use oflanguage,we will continue to fail ... to furnishsatisfac-tory answersto the long-standingquestionsabout thenature of actionandbelief in science pp. 195-196).Others such as Nersessian(1992) and EysenckandKeane(1990)havealso pointedout the important oleof narrativein scientific reasoning. Thought experi-ments are a prevalentform of scientificreasoninginwhich the scientist imagines a sequenceof events andthen narrativizes he sequence in order to communi-cate the experiment o others. Einstein is supposedtohave performed hought experimentsbased on storiesaboutridingon a lightbeam andtravelingn elevators.Rutherfordn his investigations f the structureof theatom is reputed to have imagined the electrons asrevolvingaroundthe nucleusin the sameway as plan-ets revolve around the sun (Gentner 1983). Galileo(Galilei 1638;cited in Nersessian1992) ikewiseused athought experiment n arguingagainstthe Aristoteliantheorythat heavierbodies fall fasterthanlighterones.




10/24


The Process of Perspective TakingIn knowledge-intensive irms, competitive advantageand product success are a result of collaboration nwhich diverse individualsare able to appreciate andsynergistically utilize their distinctive knowledgethrough a process of perspective taking (Dougherty1992, Purser et al. 1992, Nonaka 1994, Henderson1994, Brown 1991). Duncan and Weiss (1979, p. 86)summarize his process as one in which: The overallorganizational nowledgebase emergesout of the pro-cess of exchange,evaluation,andintegrationof knowl-edge. Like any other organizational process, ... [i]t iscomprisedof the interactions of individualsand nottheir isolatedbehavior. t requiresa process of mutualperspective taking where distinctive ndividualknowl-edge is exchanged,evaluated,and integratedwith thatof others in the organization Nonaka and Johansson1985, Shrivastava 983).Much of social behavior s predicatedupon assump-tions an actor makes aboutthe knowledge,beliefs andmotives of others. This is the beginningof the processof perspective aking,andis fundamental o communi-cations.In any communication,he knowingof whatothers know is a necessarycomponent or coordinatedaction to take place (Bakhtin1981, Clark1985,KraussandFussell1991).As Brown(1981) observed,effectivecommunicating equiresthat the point of view of theother be realistically imagined. Others such as-Rommetveit have affirmedthis point: An essentialcomponentof communicativeompetence n a pluralis-tic social world .. is our capacity o adopt the perspec-tives of different others (Rommetveit 1980, p. 126).The fundamentalmportance f taking he other'spointof view into account is seen in Mead (1934) whoreferred to it as takingthe attitude of the other andequatedour abilityto be fully humanwith our abilityto maintain an inner conversationwith a generalizedother.In order for perspective taking to proceed, the di-verseknowledgeheldbyindividualsn the organizationmustbe represented n its uniqueness,and madeavail-able for others to incorporate n a perspective-takingprocess. Valuing diversityof knowledge by enablingeach type of expertiseto makeunique representationsof their understandings, nd assistingactors with dif-ferent expertise to better recognize and accept thedifferentwaysof knowingof others, is the foundationfor perspectivetaking.It can be encouragedby com-municationsystemsthat include an emphasison sup-portingthe distinctiveneeds of separatecommunitiesof knowing.

The task of takingeach other'sknowledgeandback-ground into account is a complex process, and canfrequently break down. For example, Purser et al.(1992) did a comparative tudyof twoknowledge nten-sive product developmentprojectsof equal technicalcomplexity n a high-technologyirm. One projectsuc-ceeded while the other failed. Two essential factorsaccountedfor the differencesin results between thetwo projects.The first was a higher ncidence of barri-ers to knowledgesharing among the members on thefailed projectteam. But behind this first factor was asecond,causalfactorof failedperspective aking.Teammemberswere unableto surface and reconciledissimi-larities in their knowledge and cognitive frames ofreference.Failure o achieveperspective aking hroughdepicting and exchanging representations of theiruniqueunderstandings ramaticallyeducedtheirpos-sibilitiesfor successful eam knowledgework.Perspective taking involves a variety of inferentialand judgmentalprocesses. Individualsmay utilize anassortmentof techniques ncluding tereotypesand in-ference heuristics o estimatewhat others know. Suchheuristics can induce systematic errors and biases(Kahnemanet al. 1982,Nisbett and Ross 1980).Theready availabilityof the actor's own perspective maylead the actor to overestimate he likelihood that theperspective will be shared by others (Steedman andJohnson-Laird1980). This false consensus effect, inwhichsubjectsassumethat others are more similartothemselves han is actually he case (Ross et al. 1977) sa form of bias particularlyelevantto the perspective-takingprocess.This heuristic eads to overestimatesofthe extent to which a person'sknowledge s sharedbyothers,and studiessupport he existenceof such a bias(Dougherty1992,KraussandFussell 1991).Dougherty (1992) provides an insightful analysisofbreakdowns n the perspective-taking rocess due toactors' nability o surface and examinetheir differinginterpretiveschemes. She found that in unsuccessfulcases of new product development,the key playersinterpretedand understood ssues aroundtechnology-market inkingandnewproducts n qualitatively iffer-ent waysfromeach other andwere not able to recon-cile these differences.The differences n interpretationcentered around three themes. The first theme waswhatpeople see whentheylookinto the future, nclud-ingwhichissuesare seen as mostuncertain.Whattheysaw seemeduncertain,while whattheydidnot see, didnot seem particularlyuncertainor even noteworthy.The businessplannerworriedaboutpositioningagainstcompetitionwhile the field personworriedaboutiden-




11/24


tifyingthe right potentialcustomers.A second themecharacteristic f failed teams involvedpeople's under-standingof the developmentprocess itself. People notonly ignored he activitiesof othersandfailed to argueover relativepriorities, hey glossed over the concernsof others,and tended not to appreciate heircomplexi-ties. A third theme characteristic of failed teamsinvolved he different thoughtworlds f team mem-bers. For new productdevelopment,differentdepart-mental thought worlds were coherent and consistentwithin themselves.Thisreduced the possibility or cre-ative perspectivetaking,since members of a depart-ment thought that they already knew everything(Dougherty 1992). As lucidly worded by Dougherty(1992), Nor is the problem ike the proverbial et ofblindmen touchinga differentpart of an elephant.It ismore like the tales of eye witnessesat an accident, orof individualsn a troubledrelationship-each tells usa complete story,but tells a differentone (p. 191).

In summary then, the problem of integration ofknowledge n knowledge-intensiveirms is not a prob-lem of simplycombining, haringor makingdata com-monlyavailable. t is a problemof perspectiveaking nwhichthe unique thoughtworldsof differentcommuni-ties of knowing are made visible and accessible toothers.Makingexplicitrepresentations f one's knowl-edge and understandingso exchangewith others en-ables one to better appreciatethe distinct ways ofknowing that those others will attempt to communi-cate. In order to integrateknowledge hroughperspec-tive taking, communicationystems must first supportdiversityof knowledge hrough he differentiationpro-vided by perspective makingwithin communities ofknowing.Onlyafter a perspective s differentiatedandstrengthened an it be reflectedupon and representedso the actors in other communitiesof knowinghavesomethingto integrate through a perspective takingcommunication.Implications for ElectronicCommunication Systems and PoliciesThe design of electronic communication ystems af-fects how organizationmembersare able to engageinperspective making and perspective taking and thusbuild communitiesof knowing.In knowledge-intensivefirms, the problemof designing systems and policiesfor electroniccommunications a problemof providingan environmentn which an ecologyof communitiesofknowing can develop through complexificationovertime. In perspectivemaking,a communityof knowing

complexifiesby enrichingand refining ts distinct per-spective and way of knowing.Its categories for parti-tioning the world become more numerousand subtle;the distinctions t makes as to the appropriateness fproblem statements, measures, tests and logics for agiven situation become more esoteric and precise. Inperspective taking, complexification nvolves an in-creased capacityfor communitiesof knowing to takeeach other into account within their own languagegames,and to constructnew languag'e amesfor theirinteraction.The developmentof complexifiedperspec-tive taking represents the integrativecapacity of theecology of communities.These two dynamics,perspective making and per-spective taking,are instantiatedonly throughspeakingand actingin a community.Electroniccommunicationmedia provide an importantpart of the physicalandsymbolicenvironment available for engaging in theforms of life of the organization'scommunitiesofknowing,but only a part. Other concernssuch as task,technology, structure, culture, reward systems andleadership tyle, all play a role in mediating he type oflanguage games that will emerge. Although the entireset of these issues is beyondthe scopeof this paperwewill discuss some of the issues further in the nextsectionwhen we describesome examplesof communi-cation systems that would support these processes.Here, we will concern ourselves with presenting acertain sensibility as a way of thinking about howelectroniccommunicationmediaprovideconditions orthe two dynamicsof perspectivemaking and perspec-tive taking.

A first element in the sensibilitywe propose is torecognizethe strengthsand weaknessesof the modelsof communication nd cognitionwe are drawinguponin designing hese systems.The conduitmodel,with itsassumptionof messagesthatcarryunambiguousmean-ing if they are coded and decoded errorfree, is a goodmodel for thinkingabout the communicationof wellestablishedelements in a communityof knowing's o-cabulary and methods of practice. Communicationwithin establishedcommunity outinescan and shouldbe addressedwith a conduitmodel.The knowledge ssemi-fixedand reliably interpretablewithin the com-munity,so the assumptionsof a conduit model matchthe communication eeds well. The organization-widecommunitywhere culture and identity are acted outand a sense of institution is developed is also wellsuited to a conduit model. It is appropriate or ques-tions of broadcastbandwidthand for developmentof afirm-wide vocabulary.Recent research in corporate




12/24

RICHARDJ. BOLAND, JR. AND RAMKRISHNANV. TENKASI Communitiesof Knowing

strategyemphasizing he importanceof shared inter-pretiveschemes(Bartunek1984, Ranson et al. 1980),commonvisions(Collinsand Porras1991, Bennis andNanus 1985) or shared strategic image (Hamel andPrahalad1991, Bertado1990) are examplesof this typeof communication t the level of the organizationas awhole.The symbolicqualityof this culture-buildingommu-nication,with its reliance on evocativeimages ratherthan precise languageis somewhatat odds with theconduit model, but can generallybe adequatelyhan-dled by redundancyor repetition.The conduitmodelcan supportactivities hat broadcastand reinforce m-portantsymbols,stories,and exemplarswhich becomecommonlyavailable o membersof the community s awhole and incorporatedn their languagegames.Verylittle in the way of distinctive,organizationalknowl-edge workis accomplishedat the cultural evel of thecommunityas a whole. It is better thought of as abackdropagainst which the more esoteric languagegamesof morelocallysituatedforms of life are playedout.The conduit model, however,does have some dis-tinct weaknesses. The perspective-making rocess re-quiresa nurturingof emergentcommunitiesof know-ing,andrequiresa respectforthe uniquenessof a localcommunity's istinctiveorm of life. The conduitmodelstands in oppositionto this requirementwith its em-phasison developingdata models,decision modelsandcommunicationormatsthat are commonand sharedacrossthe organization.Currentresearchin informa-tion technologyoften reflects this inappropriate se ofthe conduit model with its emphasis on enterprisemodelingand data architecturewith a single, unifieddatastructure Scheer1992,Deng andChaudhry 992,Targowski1988, Richardsonet al. 1990, Chen 1976).Similarly,model managementsystemsconcern them-selves withunifying he diversityof knowledge n man-agement decision models througha varietyof meta-level integrativeechniques Geoffrion1987,Dolk 1988,Elam and Konsynski1987). Finally, it seems that aprincipalconcernwith end-user computing s the re-duction of diversityand the establishment f standardsand common structures or data and models (Brownand Bostrom1989,Munroet al. 1987,Rivardand Huff1988).We disagreewith these calls for commonalitynvocabularyand knowledge practices,and call insteadfor recognizing he importanceof strong perspectivemakingand differentiation f knowledgeamonga firm'scommunitiesof knowing.Electronicmedia basedon the wrongmodel of com-municationcan hinderperspectivemakingand taking

in interactions among communitiesof knowing.Anexamplefrom research on new product developmentprocesseswill illustrate he point.The task forthisnewproductdevelopment eam was to choose a nonhumananaloguesuch as a rat, rabbit or primatemodel withwhich to conduct tests of a new drug compoundtheywere developing or certain afflictionsassociatedwiththe human intestine. The team had membersrepre-senting differentdisciplines uchas life sciences,chem-istry,toxicologyand biopharmaceutics. here were dif-ferences of opinionas to the nonhumananaloguemostappropriate or the task. As a result, the team mem-bers resortedto a populargroupwareproductand itsvoting system to reach a consensus. Based on thevoting procedure,a rat analoguewas chosen.Unfortu-nately, the rat was not suitable for the task of repre-sentingthe human intestine,but the team only foundthat out at the human clinicaltrials. The poor choicehad by then cost the companyconsiderable expenseand three years of development ime. The groupwarevoting system,with its emphasison findingconsensus,hampered he team membersfrom first strengtheningand representing heir own perspectivesand then en-gagingin a dialogueof perspective takingwith eachother.The groupwarehelpedreducenoise in the com-municationandprovidedan illusion of certainty.Whatwas required,however,was a languagegames modelofcommunicationo complexify he unique understand-ing of each throughdialoguewithin their community fknowing.Then, they should have employeda technol-ogy that would supportreflexivity,and creation of avisible representationof their unique knowledgethatwould have enabledperspective aking amongthem.The languagegamesmodel also has its strengthsandweaknesses. One strength s helpingus think throughissues of perspectivemakingwith its insistence on theprimacyof speakingand actingin a local community.Electronic communicationmedia may reduce boundsof space and time for such communities(Giddens1991),but the languagegames model can help us torecover the importanceof enablingand protecting o-cal logics, practicesand vocabularies Jonsson 1992),even within dispersed communities. The languagegames model is also useful for emphasizing he needfor isolation o createidentity n a community f know-ing. Time for participatingn communities s limited,and identifiedspacesfor members o engage the com-munity's anguage games and develop its perspectiveare an importantcondition for its persistence anddevelopment.Schon (1979) providesa vividexampleof the need torespectthe importance f communicationn localcom-




13/24

RICHARD J. BOLAND,JR. AND RAMKRISHNAN V. TENKASI Communitiesof Knowing

munities from the history of town planning. Whentown planners saw their task as a need to cure ablighted area, they intervened with all manner ofplanned renewals to tear down and remake wholesections of a city, often disturbing the patterns ofcommunicationwithin neighborhoods.But theireffortswent terriblywrong,again and again, until they cameto see such areas of town not as blighted, but as folkcommunitieswith a strong networkof communicationand support hat sustained hem quite well in the faceof substantial difficulty. The problem for the townplanners hen became how to design systemsand poli-cies that would enable that emergent capacity of thelocal communitiesof knowingto strengthenand self-organize.We hope to build such an awareness nto ourapproach o thinking about electroniccommunicationfrom the start.The language games model is also a good basis forthinkingabout narrativen a community f knowing. temphasizesthat narrative is experientiallygroundedand that it is a search for ways to make issues andevents of interestto the community ensiblewithinitsway of knowing. The causal implicationsand actionsequences in narrativeare the source of perspectivemakingfor the community,as members reflect uponthe underlying ogics,values and identitiesof the com-munityof knowing.A majorlimitationof the language games model isthe epistemic inhibitions of its own paradigm(Rubinsteinet al. 1984). The strongera communityofknowingis supportedby communication ystems re-flecting a languagegame model of communication,hestrongeris perspective making complexities,and theless able it may become to allow for other ways ofseeing.A vividexampleof this dynamics presented nDougherty's 1992) field study.The various functionsinvolved n the productdevelopmentprocess agreedonthe need for the productto be marketoriented. How-ever, in the languagegamesof the researchand devel-opment group,marketorientationmeantproductspec-ificationsandtechnical eatures: he market s whattheproductcan do. For the manufacturing eople, on theother hand, a market-oriented roductwas a durableand reliableone. Lowering he numberof features andspecificationswould improve its market orientation.Further, the marketing group considered customerneeds on a customerby customerapproach.For theplanninggroup,to be marketorientedmeant to posi-tion the productin the rightmarket niche. They didnot worryaboutproduct features,customerneeds, orreliableproductperformance.Thisis wherethe rewardsystemsand cultureof the organization ecomeimpor-

Table 2 Two Models of Communication and Their RelativeMerits for Supporting Electronic Media in Systemsof Knowledge WorkCONDUIT MODELStrengths

* Reliable and precise channel for communicating well estab-lished elements in the vocabulary of a community of knowing andtechniques of practice.

* Can facilitate culture building, organization-wide integration ac-tivities through shared and common images.Weaknesses

* Does not value diversity; emphasis on uniform data and deci-sion models and communication format across the organization canhamper the emergence of unique communities of knowing.

* Inappropriate or supporting the narrative orms of cognition thatare central to the perspective making process.

* Common vocabulary and set of decision models denies theimportance of perspective taking.LANGUAGEGAME MODELStrengths* Facilitates perspective making by virtue of its insistence onprimacy of speaking and action in a community of knowing.

* Underscores the importance of enabling and protecting locallogics, local practices and local vocabularies.

* Implicates the importance of narrative n a community of know-ing.

* Emphasis on narratives enables reflection on underlying logics,values and identities of the community of knowing.Weaknesses

* Increasingly specialized language games results in epistemicinhibitions (imposed by each community's unique paradigm) andcomes in the way of perspective taking.* May heighten conflict among communities.

tant in maintaininga balance between perspectivemakingand perspectivetaking.One importantdesignelement in this regard s the establishment f an issue-specific space for perspective taking between strongcommunitiesof knowingto take place. Isaacs (1993)refers to this spaceas a container or dialogue,andwewill think of it as a forumwithin an electroniccommu-nicationsystem.As we have seen, both models have strengthsandweaknessesthat primarilyrelate to their role in per-spective making,but both models alone have distinctweaknesses with respect to perspective taking (seeTable 2). The conduitmodel, with its emphasison acommonlyavailableandexhaustive et of messagesandcodingtechniquesdenies the importance f perspectivetaking.A commonvocabulary nd set of decisionmod-els presumes that each member of the organizationparticipates n the samewayof knowingand needs no




14/24


special support or opening a spacewithinthe dialogueof theirown local communityortaking he perspectiveof another.The languagegame model,as we have justseen, also does not help in thinkingaboutperspectivetakingbecause of its emphasison speakingand actingwithin a form of life and its increasingly pecializedlanguage games. Another aspect of communicationmust be considered or thinkingabout perspective ak-ing, one that is absentor overlooked n the conduitandlanguagegamesmodels.This aspect concernshow therichness of representationsand the reflexive capacityof a communication ystem enables the creation andexchangeof boundaryobjects(Star 1989, 1993),whichwe will discuss n the next section.Reflexivity, Boundary Objectsand Perspective TakingIn our discussionof perspectivemaking n communitiesof knowing,we sawthe individual peakingand actingwithin the community'sorm of life. For perspectivetakingwe need a shift in emphasis,to focus on theindividual's bility o make his or her own understand-ing visible for self-reflection.Oncea visiblerepresenta-tion of an individual's nowledge s made available oranalysisand communication, t becomes a boundaryobjectandprovidesa basis for perspective aking.Representationsof ways of knowingfrom membersin one communitycan then be exchangedwith mem-bers of another,who, having hemselvesengagedin aneffort to make rich representationsof their under-standings, an nowengagein communication bout theperspectivesof another.This takingof the other intoaccount, n lightof a reflexiveknowledgeof one's ownperspective, s the perspective-taking rocess.Perspective aking s never a one-to-one mappingofmeanings.Membersof the same communityof know-ingwillnot have full consensus,and membersof differ-ent communitiescannotsimplyadoptthe meaningsofanother.But as Star(1989, 1993)has observed,scien-tistswithin and betweencommunitiesdo find a wayofbringing their distinctive perspectives into dialoguethroughthe constructionand discussionof boundaryobjects. An indexed collection of items, a map, anidealized image, or a label can all serve as boundaryobjects around which sense making can take place.Such boundary objects do not convey unambiguousmeaning,but have instead a kind of symbolicadequacythat enables conversationwithoutenforcingcommonlysharedmeanings.Boundary bjectscan,of course,be acenter of intense conflict as easilyas one of coopera-tive effort. Creatingand reshapingboundaryobjectsis

an exercise of power that can be collaborativeorunilateral. Nonetheless, in the absence of boundaryobjects,the possibilityof perspectivetakingis limitedand the opportunityor knowledgeworkin the firm isreduced.Reflection on our own perspectives s difficult andoften not attempted.As Rubinsteinet al. comment, Ifpracticingscientists were more conscious of the pro-cesses of science, it would go a long way toward cir-cumventinghe epistemological nhibitions mposed byparadigms 1984,p. 138).Collins(1983)also notes thehidden nature of such processes.He argues that manytimes it is only when the rules go wrong that thescientistquestionsthe nature of his or her interpreta-tion. Otherwise, urgivingof meaning o objects-ourinterpretivepracticesare so automatic hat we do notnoticethatany interpretations involved (Collins983,p. 90). In Schutz's 1964)terms, reflexivitys the abilityto periodically uspendour naturalattitude and noticethe matter-of-course,aken-for-grantedwaysin whichour communities f knowingare constructedand inter-preted, which can open possibilities to change them(Collins 1983).Rubinstein et al. (1984) posit that be-coming awareof, evaluating,and modifyingperspec-tives is requiredfor maintainingadaptiveknowledge.There are many possible forms for boundaryobjectsthat can representknowledge romone communityorperspective akingby another, ncludingphysicalmod-els, spreadsheets,or diagrams.We will present twoexamples hat could be incorporatedn communicationsystems: cognitive maps (Axelrod 1976, Huff 1990,Boland et al. 1994, Weick and Bougon 1986, Weick1990, Eden 1992), and narrative structures(Tenkasiand Boland 1993, Mulkayet al. 1983, Knorr-Cetina1981).A cognitive map is a directed graphwhose nodesrepresent concepts or factors in the actor'sdecisiondomain,and whose arcsrepresentcause-and-effect e-lationships between source and destination nodes(Boland et al. 1992, Burgess et al. 1992). Figure 1presentsan exampleof a causemapdepictinga physi-cian'sunderstanding f quality n medical care. Creat-ing this map is an exercisein perspectivemaking,andexchanging t with actors from other communitiesofknowingwithin the hospitalmakes t a boundary bjectand opens the possibility or perspective akingin thesearchfor quality n medicalcare.Buildingsucha mapcan be evocative or the map creator,as well as infor-mative to its recipient. Creating cognitive maps canreveal personal cause-and-effect ogic, which in turnforces the individual to confront the reasonablenessandvalidityof previouslyacit cause-effectassumptions

362 ORGANIZATIONCIENCE/VO1. 6, No. 4, July-August 1995



15/24


Figure 1 A Physician's Map of Quality in Medical Care

Medical ~ ~ ~ ~ ~ civ

+_

Treatmentj m }d, uestl(-.)ues nmg

ancl ~ ~ Bac c~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~Lc

KOMedice ile

EffectsPereNte Arcs wo+Ptieive

- r \)fCareferal ompassi)

Xpectatio Cureaiorehwniniats nerait ofNI

c Sientific 5 \> _/


16/24

RICHARD J. BOLAND, JR. AND RAMKRISHNANV. TENKASI Communitiesof Knowing

(Fiol and Huff 1992,Weick and Bougon 1986). Creat-ing maps of one's understanding f a problemdomainand reflecting on them can also facilitate new andmorecomplexunderstandings f the situationat hand,improving the chances for scientific success (Weick1990).Cognitivemaps are a good beginning ormakingrichrepresentations f an understandingwithin a perspec-tive. But a key ingredient or communicative uccessisa way to link elements and relations n a map,as wellas the map itself to unstatedelementsand assumptionsof the perspective.That is, the knowledgerepresenta-tion growsricheras contextis added, layer by layer,toindividual lements in the cognitivemap.This suggestsa hypertext or hyper media communicationenviron-ment in which actors find a self-reflectivespace tobuild rich knowledgerepresentationswherebyfactorsin a cause map are linked to underlyingbeliefs, valuesor assumptions n the form of spreadsheets,notes, or

graphsor other cause maps (Bolandet al. 1992, 1994).Another kind of boundaryobjectthat can serve as afocal point for perspectivetaking is a narrative truc-ture. Narratives,f bracketedand approached or anal-ysis with an interpretivestance can also provideele-ments of the reflexivequalitywe see as necessaryforperspective taking. Narrativeanalysiscan reflexivelygive access to the implicitand unstated assumptionsthat are guiding perspectivemaking,and in so doinghelp enable a perspective taking process. We willdemonstrate hisbyfirstpresentinga narrative rom anactual incident collected duringfield work in a phar-maceuticalcompany,andthen analyzingand interpret-ing its narrative tructure.The Story of Norman, a Chemist

Norman stood up from his work bench in mid-morning andwent to the men's toilet where he used the urinal. Shortlyafter returning to his work bench, Norman felt a numbness inhis penis. He was startled, but he immediately thought thattrace amounts of the XV75 hypertension compounds he hadbeen working with that morning had been on his hands andmay just be a powerful topical anesthetic.

He told two colleagues about this potential discovery andcreated an informal team to explore its possibilities. Afterabout six weeks, he obtained formal approval from the Assis-tant Director for this project and his team. At this stage, theidea was to go for a topical application of the compound.After several weeks, Norman went to see the Assistant Direc-tor to inform him of a metabolic study of the compound in acell culture that showed some indications of toxicity. Helearned from the Assistant Director that a market study hadjust been completed showing that an oral form of the drugwould be very successful and highly profitable, whereas the

Figure 2 Narrative Structure of the Story of NormanIgnore Focus on topicaliumibiiess versioii

T Seoise ioew Forixiprofret Develop oralTito liscovei-y teamnut versioi- \ ~~~~~~~~~report

Coomplaixif lax Discoolinueprocedures peojectLepeod(: d = Keroel Everrt

( =SatelliteEveolt

topical version would actually have a very limited marketpotential.

.The Assistant Director told Norman that the toxicity reportwas uncertain and that he should reorient his team toward anoral form of the compound. Think positive, he told Norman.

We have to move on and we have to take risks if we expectto reap rewards. Market projections of the proposed oralform of the drug were presented to the Executive Committeeof the corporation, and were enthusiastically received.Chatman 1978)presentsan elaborate ramework ordiagramming arrative tructure,and we can use someof his techniques in a simplifiedform to show hownarrativeanalysiscan surfaceassumptionsand aid re-flexivityin perspective making. In diagramming hestructure of events in a story plot, Chatman(1978)distinguishes etweenmajorand minorevents. He callsmajorevents kernelsand shows them as a square n hisdiagrams.Chatmanrefersto minoreventsas satellites(1978,p. 54) and showsthem as circles n his diagrams.Satellites are events which enrich the story aestheti-cally,but are not crucial to the plot. Satellite eventsnecessarily mply he existenceof kernels,butnotviceversa. (Chatman 1978, p. 54). Figure 2 is a partialdiagramof the plot of Norman's tory.In diagramminghe first part of this story,we haveidentified two kernels, treatingthe other elements assatellites.Other readersmight interpretthe structuredifferently,but that is whatkeeps an interpretive on-versation ively.The two kernelswe isolate are: the tripto the toilet and the initialtoxicityreport.Forpurposesof an example,we will providea brief analysisof thetwo kernels.First Kernel: A Tripto the Toilet. From this kernel,the story could have taken several different paths.First,Norman could have simplyreturnedto his workstation, and waited for the numbness to go away.Eventually t would have and this episode would be

364 ORGANIZATIONCIENCE/VO1., No. 4, July-August 1995



17/24

RICHARDJ. BOLAND,JR. AND RAMKRISHNAN V. TENKASI Communities of Knowing

over. Or, Norman might have become enraged that noone had warned him that the XV75 compounds couldhave this effect. After much finger-pointing and theestablishment of stricter chemical handling policies,this story would also eventually end without a newproject being instituted. Instead, Norman used theevent of the numbness to engage in perspective taking,looking at XV75 and his own experience from a per-spective other than that of hypertension or personaldiscomfort. In so doing, much about the canonicallityof the world of the lab is revealed.First, the way the kernel is resolved shows that it iscanonical to be open to the meaning of an unexpectedevent, that science will take strange twists and theseemingly irrelevant could be the basis for an impor-tant new discovery. It is canonical in this lab for ascientist to take any event, no matter how bizarre orpersonal, and view it as a potential for creating newknowledge. Second, we learn that it is canonical to seethe event of numbness as an experiment on oneself.This lesson of the narrative is supported by field workwhich confirmed that self-experimentation is a fre-quent practice among the lab workers. The first kernel,then, can tell us much about the values and lab prac-tices in this community.

Second Kernel: The PreliminaryToxicity Report. Inthis kernel, we can imagine several alternatives that didnot happen; the project could have been focused on atopical version only, because of the risk of toxicity, orthe project could have been dropped altogether. In-stead, canonicallity is restored by a call for positivethinking and the lure of a large profitable market foran oral version. The tension between the market/profit-seeking perspective in product innovation andthe toxicology perspective is lopsidedly made canonicalin favor of the market. In this kernel there is a distinctfailure of perspective taking on the part of the Assis-tant Director. As a result, the possibilities for knowl-edge creation in this network are diminished, the fram-ing of the problem is constrained, and opportunitiesfor a complex exploration of how risk, rewards, toxicityand efficacy can become a topic of open dialogue arediminished.There is obviously more that could be done in read-ing the canonicallity of the lab in these kernels, butthese examples should suffice. The important point isthat the kernel is a hinge in the structure of the storyand interpretation of the kernel gives access to what iscanonical in a community that may be difficult tosurface otherwise.

ImplicationsPerspective taking through boundary objects is a rela-tively unexplored frontier in electronic communication.One can expect that tools and media to support reflex-ivity, representation of knowledge structures and theirexchange with others in a perspective-taking processwill increase over time. Paradoxically, it is a kind ofcommunication with others that grows out of an im-proved communication with self. Communication withone's self is the basic stance of reflexivity; an innerconversation that builds and reflects upon a represen-tation of one's understanding of a situation. Being ableto do so implies that the perspective making in acommunity of knowing has progressed far enough toprovide a sufficiently strong perspective to reflect upon.Having had this type of communication with one's self,the actor is equipped to enter into a new kind ofcommunication with others, that of perspective taking.We now present some examples of the types of elec-tronic communication systems suggested by our an-alysis thus far, by describing an idealized firm thatdisplays strong capacities for perspective making andperspective taking.

Some Examples of DesigningCommunication Systems to SupportPerspective Making and PerspectiveTakingThe implications of using information technologies toprovide support for perspective making and perspec-tive taking are best understood as the interrelationshipof organizational, cultural and technological elements.This insight was evident in the first experiences withindustrial research laboratories (Marcson 1960, Carlson1992), in the Manhattan project (Davis 1969) and alsoin recent studies of new product innovation (Law andCallon 1992, Carlson 1992, Dougherty 1992). In keep-ing with an emphasis on how a narrative and languagegame orientation can be interweaved with paradig-matic reasoning, this section will present a plausible,but admittedly utopian form of a knowledge intensivefirm. In this idealized firm, a reflexive hermeneuticattitude (Gadamer 1975, Boland 1993, Boland et al.1994) and an open recognition of language games andthe process of perspective taking is assumed to be wellestablished. We will first describe the technological,organizational and cultural backdrop for such a hypo-thetical knowledge-intensive firm of the near future.We will then describe some applications of informationtechnologies that could be employed for perspective




18/24


makingand perspective takingby its communitiesofknowing.Technologically,we expect to see that computing,imagingand communication eviceshave become ubiq-uitous. The information nvironmentn this hypotheti-cal firm s a seamless ntegrationof multimediadevicesfor collection, storage, processing and display. Theorganization s repletewith systemsbased on the con-duit model and language games model. Once certainkinds of knowledgeare establishedand the perspectiveof a community f knowingbecomes mature,the deci-sion routines are embedded in project managementand other kinds of software,althoughsuch decisionpremises are always subject to question and revision.Graphics,texts, models, audio and video applicationsare all radically ailorable o a user'sneeds. Hyperlinksfrom an element in any one application o elementsinanyother applicationare fullysupported,makingcon-textuallyrich, complexly layered representationsthenorm. Groupware s highlydeveloped,with multimediameetings,and discussiongroups in a wide variety ofissue forums.A sophisticatedvocabularyof electronicforms for initiating,replyingor commentingon deci-sions models and discussion opicshasemerged hroughan open processof structurationGiddens 1979).Organizationally,he firm is characterized y a criti-cal densityof interdependentknowledgecommunities.There is a post-modem (Harvey1989) qualityto theorganization,and groupware ommunication rocessesare markedby multiplevoiceswith shiftingpatternsofinterest, givinga sense of a fragmented,almostchaoticcommunication nvironment ompared o the predom-inantly hierarchical ne of the late 1980s. The organi-zationuses lateral teams extensivelyn whichthe verti-cal authoritystructureplays a muted role while theprinciplevalue addingactivitiesof knowledgecreationand knowledgeapplicationare carriedout in a chang-ing mosaic of lateral project teams. Because of thefirm'sstronglateral form and collaboration-based e-wardstructure,parochial nterestgroupsand fiefdom-like powerbases which used to subvertefforts at freeand informed communication have largely disap-peared. Individualswho play important liaison rolesbetweenstrongcommunities f knowinguse their newlydeveloped skills as semiotic brokers Lyotard 1984)to help facilitate the perspective-takingrocess.Culturally,he idea that doingwork in a knowledge-intensive firm means perspectivemakingand perspec-tive takingin communitiesof knowinghas taken holdand has shaped both individualand group identities.Individualshavea reflexiveawarenessof theirparadig-matic as well as their narrativemodesof cognition.The

culture reinforcesan awarenessof the individual's a-pacityto step outsideof a messagestream and engagein meta communicative nalysis Bateson 1972). Mem-bers of the firm are used to taking an interpretivestance, playingwith possible meanings,searchingforunderlying tructures,questioning he social construc-tion of new nouns and verbsin their languagegames.They enter into and makereadingsof communicationepisodes with an open awarenessof the hermeneuticcircle in which they tack back and forth from aninterpretation f the larger context of a perspective oan interpretation f the detailed elements of the mes-sage at hand (Palmer 1969). Their hermeneuticatti-tude means they avoid debate in favor of dialogueunless compellingreasons call for a dialectic commu-nicativeprocess. They realize that debate is a win-losepolarizingstrategy hat rarely results in true synthesisor creative nsights.Dialogue,in contrast, s a mutuallyreinforcing,working ogether through anguage.It is arealization that we can assume a perspective-takingorientationand benefit fromopeningourselves to thehorizon of another.Withinthe organizational,ulturalandtechnologicalenvironment ketchedabove,communitiesof knowingare using advanced groupwarefacilities to conductmeetings,constructmulti-author ocuments,and coor-dinatetheirpromisesand deadlines,all with the capa-bility to access data and knowledge through a world-wide network of knowledge repositories.As is truetoday,the groupware ystemsare composedof a seriesof forumswhichserveas containers or dialogueoncertaintopics, issues, concerns, projectsor tasks.Fo-rums reflect the way the knowledge work is beingfocused,andthe kindsof knowledge tructures hat areemerging n the firm,and are thus one avenueinto itscommunities of knowing. Individualsparticipate inmanyforums n an evolvingpattern.The lateralgroupsto whichthey belongand theirunique expertisedefinesthe types and kinds of forumsin which they partici-pate.What takes place in these forums are languagegames.The modeof cognition s a mixed one in whichparadigmatic easoning s interwovenwith stories andnarration. The applicationsof advanced informationtechnologies for perspective making and perspectivetaking hat we describebelowdepend upontherebeinga higher level of reflexivityin knowledge intensivefirmsthan is presently he case. These communicationsystems depend not only on talkingabout issues andproblemswithina groupware nvironment, ut on talk-ing about how they are talking (Bateson 1972). Itdepends upon a criticalhermeneuticattitudein which




19/24

RICHARDJ. BOLAND, JR. AND RAMKRISHNAN V. TENKASI Communities of Knowing

the strangeness and multiple possibilities for makingmeaning in our conversations are constantly in ourawareness (Gadamer 1975, Ricoeur 1981, Boland 1993).As groups form and reform in a knowledge-intensivefirm employing a lateral organization structure, weanticipate five new classes of electronic communicationforums as examples of ones that would enhance theprocesses of perspective making and perspective tak-ing. Within each class there would be several differenttypes of forums as we will discuss below. The five newclasses of forums we propose as examples are:1. Task Narrative Forums;2. Knowledge Representation Forums;3. Interpretive Reading Forums;4. Theory Building Forums;5. Intelligent Agent Forums.Task Narrative ForumsThis type of forum has been envisaged by Brown andDuguid (1991) and Galbraith (1994) among others, andis an implicit recognition of the importance of narra-tivizing our experience and sharing the narratives withothers in our community. Through narrative, the com-munity constructs its practices and its social world bybuilding and restoring its sense of the canonical. Nar-rative, by making the implicit and the tacit inferable tothe reader or the listener, is a critically important firststep in achieving perspective taking within and amongcommunities of knowing. Because these task narrativeswould be multimedia, and include video and audio,they enable the benefits of learning by experience toextend beyond normal constraints of space and time.Task-narrative forums serve as perspective makingfor those creating the narratives and also serve as aperspective-taking experience for those reading thenarratives. The narrative is always incomplete and thereader must read into the story in making it sensible.Bruner (1990) refers to this reading into as a sub-junctive process and is a primary vehicle for openingoneself up to the perspective of another and makingreal its possibilities for seeing the world differently.KnowledgeRepresentation ForumsCurrent groupware enables linking from a text docu-ment to a spreadsheet, decision model, graphic depic-tion, or picture. Once a document is hyper-linked inthis way, the context it carries with itself is enrichedand its possibilities for interpretation are increased. Sowe are already used to seeing a message with otherdocuments linked to or embedded within it. A knowl-edge representational forum, in contrast, is one whichfocuses on the understanding that lies behind such

Figure 3 Perspective Making and Perspective Taking

Community ofCmuiyoKnowing A' Knowing'B'

Perspective Makiing Perspective Making- Narratives of experience - Narratives of experience- Paradigmatic analysis - Paradigmatic analysis- Reflexivity and - Reflexivity andRepresentation RepresentationPerspective Taking Perspective Taking

- Reflexivity and - Reflexivity andnterpretive Reading Interpretive Readi

BoundaryOb'jects- Cause Maps- Narrative miaps- Models- ClassificationSchemes

complex documents. It is a forum that captures acommunity's cooperative efforts to reflect upon, inter-pret and depict an understanding of their situation tothemselves.It is not a problem-solving or task-practice forum somuch as a sense-making forum in which the objects ofdiscussion are visual representations of their under-standing of a situation, a problem or an objective. It isan openly reflexive forum in which communities ofknowing explicitly talk about their understandings. Suchforums could use storyboards in which still or animatedpictures are assembled in a sequence, in a kind ofvisual depiction of an understanding, or these forumscould use cause maps as in Figure 1, or other diagramsand models for representing an understanding.Like narrative forums, representational forums serveas a perspective-making experience for those construct-ing, revising, or commenting on an emerging represen-tation within a community of knowing. They also serveas a perspective-taking experience for those who readthem with a hermeneutic attitude of engaging thehorizons of another thought world (see Figure 3).InterpretiveReadingForumsWhereas representation forums are overtly reflexive inthat participants are tryingto reflect upon their currentstate of understanding of some issue, interpretive read-ing forums are a space for reflecting upon the assump-tions and meanings revealed by the communications inother forums. In this forum, participants are subjectingother texts to re-readings in hopes of portraying thetacit and implicit meanings characterizing a community




20/24


of knowing, their own or others' (Czarniawska-Joergesand Guillet de Monthoux 1994).Discussions in this forum could resemble dialoguesof literary criticisms in which critical-reflexive readingsare made of the streams of entries in other forums.Eventually, re-readings could even be made of thedialogues in the interpretive reading forum itself, aslayers of reflexivitybegin to compound (Ashmore 1989).In addition to interpretive essays, such a forum wouldalso be used for discussing the narrative structures inthe task narratives forum and in the narrativizationthat is evident in all other forums as well. Here, thediagramming of narrative structures, the isolation ofkernels, and the unpacking of how the canonical andnoncanonical are revealed for different communities ofknowing would take place.Another type of forum within this class could befocused on words and might be known as a word-talkforum. This would be another type of reflexive analysisin which words (new words, especially nouns and verbs,as well as familiar ones) were systematicallyconsideredas to their changing meanings and uses, their shiftingcontexts and connotations, and the implicit and tacitassumptions they reveal.The interpretive reading forums discussed above arethe most explicitly hermeneutic-interpretive ones weenvision and have the greatest dependence on thesupportive organizational and cultural qualities dis-cussed above. Without a widely shared sense of theimportance of perspective making and perspective tak-ing in knowledge creation, and a well established senseof the value of a hermeneutic attitude, these forumswould not be possible.Theory-buildingForumsThese forums would most closely reflect the dialogueof theory that is woven throughout scientific practice.We envision this as a series of forums in which differ-ent communities of knowing articulate, critique, extendand explore the theories that do or should guide theirwork. Theory-building forums are not just for sciencework, however, and we would anticipate that in aknowledge-intensive firm the ethos of perspective mak-ing and perspective taking also would be held by thefinancial, marketing and other nonscientific fields ofdiscourse within the organization, and between theorganization and its many environments.In addition to Theory Corners , or forums dedi-cated to dialogue on theories within and across specificcommunities of knowing, we would also expect theory-building forums to include Thought Experiment fo-rums, where individuals played with theories and their

implications by narrativizing thought experiments.Thought experiments entail the construction of mentalmodels by a scientist who imagines a sequence ofevents, and then uses a narrative form to describethe sequence in order to communicate the experimentto others (Nersessian 1992). Thought-experiment fo-rums would also include the construction and playfulexploration of simulation models, especially multime-dia simulations and virtual reality systems.Once again, we see this class of forums playing animportant role in perspective making and also perspec-tive taking. Constructing theories and conductingthought experiments in dialogue within a community ofknowing is essential for strong forms of perspectivemaking. Participating in these forums from the fringesof the community, or reading and interpreting thetheory building from outside the community is a pow-erful means of perspective taking.Intelligent Agent and Expert SystemForumsThe final class of conversational forums we will presenthave to do with intelligent agents and expert systems.By intelligent agents we mean personal assistants inthe form of software systems that can roam the net-work of forums within a firm as well as libraries,repositories and information sources outside the firm.These agents help an individual to direct her attentionwithin the burgeoning field of forums that could be ofimportance and interest to her, and also help assemblecontextual materials for building cross-document linksin complexly layered representations.We see these intelligent agents as well as expertsystems as another important form of reflexivitywithinthe firm. Both are classes of software systems whereexpertise and interests have been reflected upon, madevisible and embodied within these artificial agents. Inthe forums we envisage, individuals would develop andshare insights into how these systems can best beconstructed and deployed, and how their results canbest be interpreted. The forums would thus be directlyconcerned with thinking about thinking, especially asthought processes are embodied in the active agentsand expert systems.

SummaryThe applications described above are not intended asan exhaustive listing, nor as a taxonomy of ways inwhich perspective making and perspective taking canbe supported by electronic communication. Rather, weintend merely to open a discussion of some possibilitiesfor appreciating language games, narrative cognitionand reflexivity in the design of electronic communica-




21/24


tion systems. No doubt, many of these kinds of activi-ties are already being explored in nascent form bythose organizations that are installing extensive group-ware capabilities.Concluding ThoughtsAny design of an electronic communication systemimplies a model of human communication and of hu-man cognition. We have explored how principles andpolicies for the design of electronic communicationsystems are affected by incorporating a language gamesmodel of communication and an awareness*of thenarrative mode of cognition. In so doing, we haveargued that perspective making and perspective takingin the science work of knowledge-intensive firms and infirms generally using lateral organizational forms wouldbenefit from systems designed with this sensibility inmind. We have also pro

Perspective Making and Perspective Taking in Communities of Knowing.1995

Documents

Transcript of Perspective Making and Perspective Taking in Communities of Knowing.1995