knowledge management: a threefold framework

The Information Society, 18:47–64, 2002Copyright c° 2002 Taylor & Francis0197-2243 /02 $12.00 + .00

Knowledge Management: A Threefold Framework

C. W. HolsappleSchool of Management, Gatton College of Business and Economics, University of Kentucky,Lexington, Kentucky, USA

K. D. JoshiSchool of Accounting, Information Systems, and Business Law, College of Businessand Economics, Washington State University, Pullman, Washington, USA

It is widely claimed by a number of business and academicgurus that in order for organizations to have a lasting competi-tive advantage they will have to be knowledge driven. If knowledgeis viewed as a resource that is critical to an organization’s survivaland success in the global market, then like any other resource itdemands good management. However, the bulk of organizationsstill have not approached knowledge management (KM) activityformally or deliberately. The cause for this inattention could bethat most organizations are still struggling to comprehend the KMconcept. To ease the struggle, the fundamental issue of identifyingsalient characteristics of KM phenomena needs to be addressed.This article helps address this need by introducing a threefolddescriptive framework that identi� es and characterizes the mainelements of KM phenomena and their relationships. The � rstcomponent provides a generic description of an organization’sknowledge resources. A second component introduces elementalknowledge manipulation activities an organization performs indealing with those resources. The third component identi� es majorin� uences that impact an organization’s conduct of KM. Results ofa survey to assess the framework are reported. They indicate gen-eral satisfaction with the framework.

Keywords chief knowledge of� cer, framework, knowledge-basedorganization, knowledge management, knowledgeresources

Received 12 August 1997; accepted 28 February 2000.The authors are grateful for the suggestions made by the editor-in-

chief, associate editor, and referees. Authors are listed alphabetically.Address correspondence to K. D. Joshi, Todd 240B, School of Ac-

counting, Information Systems, and Business Law, College of Businessand Economics, Washington State University, Pullman, WA 99164,USA. E-mail: [email protected]

The world in general, and the business world in par-ticular, is experiencing a paradigm shift: a shift towardknowledge-based organizations (Holsapple & Whinston,1987) in a knowledge-based society (Tof� er, 1990;Drucker, 1993). In such organizations , knowledge is re-garded as a crucial resource, and harnessing knowledge-processing skills to maximize the value of this resourceis recognized as a paramount concern. An organization’sknowledge resources are complex and multifaceted, rang-ing from tacit components to knowledge that is explic-itly represented (Nonaka, 1991), and including descriptiveknowledge, procedural knowledge, and reasoning knowl-edge (Holsapple & Whinston, 1987). Both computer andhuman participants in an organization possess variousskills for manipulating its knowledge resources. This port-folio of skills and how they are deployed in manipulat-ing available knowledge resources go a long way towarddetermining the nature of an organization’s innovationsand outputs, and hence its competitiveness in a dynamicenvironment.

It is clear that knowledge resources need to be carefullymanaged rather than being left to serendipity (Amidon,1996). Documented cases of organizations that haveachieved some success in this include Chaparral Steel,Honda, Canon, Buckman Laboratories, and Skandia(Rifkin, 1997; Stewart, 1997; Leonard-Barton, 1995;Nonaka, 1991). However, as Nonaka (1991) points out,“despite all the talk about brainpower and intellectual cap-ital, few managers grasp the true nature of the knowledge-creating company—let alone how to manage it.” Zack andSerino (1996) maintain that “While the business case forknowledge management is becoming widely accepted, feworganizations today are fully capable of developing andleveraging critical organizationa l knowledge to improvetheir performance.”

47

48 C. W. HOLSAPPLE AND K. D. JOSHI

One reason why the full potential of deliberate, system-atic knowledge management (KM) efforts has yet to bewidely ful� lled is that organizations are still struggling tocomprehend the KM concept. Anne Stuart, senior editor ofCIO; writes, “Many managers would be hard pressed to ex-plain precisely and concisely, what this evolving businesstrend (KM) means. What they probably do know is knowl-edge management has been billed as a critical tool for the21st century corporation . : : : And they know it is some-thing they just have to get too—even if they don’t knowexactly what it is” (Stuart, 1996). To foster a commonunderstanding of what it is, we need to address the fun-damental issue of identifying the salient characteristics ofKM phenomena in organizations. This is a prerequisite forsystematic research into the nature and possibilitie s of KM,as well as for easing the emergence of KM into practice.

This article advances a framework that identi� es keycharacteristics to consider in the study and implementa-tion of KM. It is the result of synthesizing elements fromexisting descriptive frameworks and related KM litera-ture to yield a relatively complete, uni� ed perspective.The framework is comprised of three main components.The � rst of these provides a taxonomy that identi� es thekinds of knowledge resources that an organization canhave and manage. The second component identi� es basictypes of activities that can be used to manipulate an orga-nization’s knowledge resources. An organization’s partic-ipants (human and computer-based) use their knowledge-handling skills to execute these manipulation activities,producing and consuming knowledge � ows in the pro-cess. The framework’s third component characterizes threeclasses of in� uences that shape the conduct of knowledgemanagement in an organization (e.g., they in� uence thecon� guration of manipulation activities used to satisfyknowledge needs).

We begin with a brief background discussion of the na-ture of knowledge and the need for descriptive KM frame-works. Next, the threefold framework is described. As itselements are introduced, they are tied to the KM literatureand illustrated with examples. In addition to this concep-tual grounding, some external validation is then presentedin results of a survey to assess the threefold framework’scompleteness, accuracy, clarity, and conciseness. We closewith a description of future research avenues emanatingfrom the framework.

BACKGROUND

Throughout the ages, philosophers , scientists , and othershave debated the nature of knowledge. Although it is notour intent to review or add to this debate, a brief con-sideration of knowledge provides useful background forappreciating the framework.

According to Newell (1982), knowledge is embeddedin usable representations. Simply put, representations arepatterns. This includes patterns of materials, energy (e.g.,sound, light electro-magnetic), actions, behaviors, andother symbolic systems. Such representations conveyknowledge to the extent they are usable by some pro-cessor (i.e., a knower). That is, knowledge is relative tothe knower; one processor may attain great utility fromthe knowledge embedded in a representation that anotherprocessor is unable to use (Holsapple & Whinston,1996).

The KM framework advanced here is not oriented to-ward any particular kind of representation or processor.It admits the possibility of both explicit representations(e.g., the words displayed here) and tacit representations(e.g., mental states). It also admits the possibility of bothhuman-based and computer-based processors. However,its focus is not on enumerating types of representations,transformations of representations , or the mechanisms forprocessing a particular type of representation. Its focus ison identifying classes of knowledge resources, basic ac-tivities for manipulating them, and factors in� uencing theconduct of KM.

Commentators on the knowledge management sceneoften strive to draw distinctions between the notions ofdata, information, and knowledge. Some of these samecommentators, as well as others, proceed to use the termsknowledge and information interchangeably. Still othersargue that knowledge management is simply a renamingof information management. The framework is not depen-dent or predicated on any of these views; nor is its purposeone of grappling with such distinctions . However, the man-ner in which the framework is understood and applied isdependent on one’s view of knowledge.

Like much of the KM literature, the framework proba-bly furnishes the least value to those who equate knowl-edge with information. For those who treat knowledge andinformation as mutually exclusive, the framework can beapplied to whatever is regarded as being in the knowledgecategory. Between these two extremes is the view thatregards information as a kind of knowledge. A progres-sion of several knowledge states including data, informa-tion, and knowledge has been identi� ed by van Lohuizen(1986); the progression proceeds from the least re� ned,least focused, least directly usable state of knowledge (i.e.,data) through states that are increasingly re� ned, focused,and directly usable. A researcher applying the frameworkcan consider knowledge to encompass all such states,or can choose to apply it to some subset of thesestates.

Machlup (1982) contends that all information is knowl-edge, but not the reverse; knowledge is more than justinformation. For example, one knowledge taxonomy

KNOWLEDGE MANAGEMENT: A THREEFOLD FRAMEWORK 49

distinguishe s among descriptive, procedural, and reason-ing knowledge (Holsapple, 1995). Descriptive knowledge,commonly called information, describes the state of someworld (e.g., past, current, predicted, hypothetical). In itsvarious states from raw data to structured information,descriptive knowledge has long been the focus of infor-mation management and information systems for repre-senting and processing it (Holsapple & Whinston, 1988).In contrast, procedural knowledge is concerned with howto do something (e.g., processes, techniques) and reason-ing knowledge indicates what conclusion is valid ina particular situation (e.g., cause-and-effect principles,correlations, heuristics). Like descriptive knowledge,these exist in various gradations of re� nement andusability.

An organization’s descriptive , procedural, and reason-ing knowledge is explicitly and/or tacitly represented andprocessed by human and/or computer-based participantsin the organization. Although the framework does not in-corporate or commit to this view, we suggest that such aview is suf� ciently inclusive to be a helpful basis for under-standing and applying the framework’s characterizationsof what an organization’s knowledge resources are, whatactivities can be exercised in manipulating those resources,and what factors in� uence the organization’s conduct ofKM.

In the absence of a comprehensive framework, a � eld’s“progress is but a fortunate combination of circumstances,research is fumbling in the dark, and dissemination ofknowledge is a cumbersome process” (Vatter, 1947). Al-though a few descriptive frameworks have been positedfor KM (e.g., Nonaka, 1994; Wiig, 1993), each seems toaddress only certain KM elements. A comparative anal-ysis of such frameworks shows that individually, nonedescribes the full scope of KM phenomena (nor is thisthe stated intent of any); collectively, it is not obvioushow they might � t together and it appears that they over-look some KM issues reported elsewhere in the literature(Holsapple & Joshi, 1999a). A more comprehensive de-scriptive framework for characterizing KM phenomenacan bene� t both researchers and practitioners by furnish-ing an organized foundation for future progress in under-standing the conduct of KM.

Here, we help address this need by introducing a de-scriptive framework that identi� es and characterizes themain elements of KM phenomena and their relationships .A descriptive framework provides a perspective for fullyunderstanding (as well as organizing) concepts in a uni� edfashion. It describes a phenomenon in the form of key fac-tors, constructs, or variables and their relationships (Miles& Huberman, 1994). One � nding from a survey evaluatingthe threefold framework introduced later in this paper isthat respondents are unanimous in their view that a KM

framework is needed; representative respondent commentsabout this need are shown in Table 1.

A THREEFOLD KNOWLEDGE MANAGEMENTFRAMEWORK

For many years organizations have formally captured andmanaged explicit knowledge by way of computer systemssuch as management information systems, decision sup-port systems, and expert systems. One premise of this ar-ticle is that such technologies are just a portion of whatis possible and what is necessary for effective KM. Asecond premise of this article is that understanding KMphenomena depends on (1) characterizing knowledge re-sources that need to be managed, (2) identifying and ex-plaining activities involved in manipulating these knowl-edge resources, and (3) recognizing factors that in� uencethe conduct of knowledge management. These three as-pects correspond to components of the descriptive frame-work introduced here: a knowledge resources component,an activities component, and a KM in� uences component.The knowledge resources are an organization’s reservoirsof knowledge; this knowledge (e.g., descriptive , proce-dural, reasoning) is embedded in usable representations(e.g., tacit, explicit). The knowledge manipulation activi-ties are elementary functions an organization performs inprocessing its knowledge resources. The KM in� uencesare factors that affect the conduct or manifestation of KMin an organization.

A major objective of this framework is generality. It isnot geared toward the description of KM within any par-ticular organization, but rather can be applied to the studyof KM phenomena in diverse organizations . As such, itselements involve a certain level of conceptual abstractionthat becomes more concrete only when applied to the studyof a particular organization. As the elements of each com-ponent are introduced, we ground the descriptions in ex-amples. These examples are included to illustrate conceptsand are by no means exhaustive.

In creating the framework, pains were taken to ensurethat its concepts accommodate features found in prior KMframeworks, plus ideas gleaned from the KM literature atlarge (Holsapple & Joshi, 1999a). This involved an itera-tive process of synthesis in the direction of a unifying andmore complete view of KM. In keeping with the objectiveof generality, the framework recognizes that technologycan play an important role in KM, but does not emphasizeits role relative to nontechnologica l aspects. That is, theframework is applicable regardless of whether one wantsto use it from a technological , human, or hybrid vantagepoint.

A survey was conducted to assess the threefold frame-work in terms of its completeness, accuracy, clarity, and


TABLE 1The need for a KM framework: Survey respondents’ comments about the need for a KM framework

The framework is important for the practical & theoretical structure it can provide. It can provide a contextfor all work in the � eld.

It is important to create (or begin the process of converging on) a consensual de� nitionof the phenomenon so that we can begin to perform “normal science” and startto coordinate accumulated research.

A framework helps people understand what KM is, what knowledge activities are involvedand how the knowledge activities affect organizational effectiveness. Most of the confusionabout knowledge management results from the lack of a comprehensive framework.

This framework gives academicians and practitioners a common set of well-de� ned constructsfor research and practice in KM.

A framework can help place people’s efforts in a bigger perspective. It can also help both practitionersand researchers have a way to identify if they have covered all the appropriate issues pertainingto their situation.

For (a) awareness and understanding, (b) common communication, (c) scoping of initiatives and projects,(d) further development of the � eld. It is especially important at this early stage in the developmentof the practice and theory of KM to be able to discuss what it is and “is not”; what entities and activitiesit is concerned with (as you have done). The importance of models and frameworks in the communicationof subtle issues, such as one involved in KM, can hardly be overstated.

A framework facilitates communication.It is also extremely useful to have a common and understood vocabulary.A beginning for purposeful research in an emerging area.(1) Need to be able to de� ne for students the range of activities that they will be prepared for,

if they concentrate in KM. (2) Clari� es what people mean when they talk about KMand intellectual property. (3) Reveals the “culture” of KM-ers to those ofus who study this “information renaissance” as a social phenomena.

You must be able to visualize it in order to manage it and continuously improve it. Managing knowledgeis not new—what is new and exciting is the development of a frameworkand language that allows us to talk/study it.

Much confusion exists surrounding the notion of knowledge management. Most of this is based on a lackof clarity with respect to the de� nition and domain of KM. A framework is needed that de� nes the boundaryof KM as well as its components.

We tend to have (create) tacit frameworks, so an explicit one helps in re� ection and communicationin a wider circle.

(1) Determines scope of action/management. (2) Acts as visual support to aid communications.(3) Can be a diagnosis/resource allocation tool.

To understand how big a problem is and that it is necessary to consider it in global dimension. To identify all resources,actors and in� uences involved in the process. Framework helps to understands all (quite all) facets of KMto make a difference between true KM and just marketing keywords.

The KM label is being applied to everything and anything by consultants and “scholars” looking to getin on the hottest issues of the day. We need to reclaim the concept and provide some in� uential foundation,rigor, and consistency. On the other hand, we must realize that something as basic and fundamental as knowledgecannot be captured by a single view.

It matters not what framework or architecture you use, but having one enables systematic knowledge identi� cationand leverage.

The experience I have with my clients is that until they have a coherent vision (the perspective based on an overallframework model), they cannot focus on priorities, identify how to coordinate cross-organizational efforts,[focus] on identifying overall long-term bene� ts.


conciseness. Survey results indicate an appreciable degreeof success for these criteria. They also suggest the possi-bility of future enhancements. Future investigations willdetermine the extent to which the threefold frameworkserves to unify viewpoints and ideas in the � eld, the de-gree of its descriptive power, its utility in the generationof KM theory and research hypothesis , its value as a guideto considerations that practitioners need to address, andits contribution as a stepping stone in the stimulation offuture KM frameworks.

The Knowledge Resource Component

A conventional view of an organization sees it as hav-ing three main types of resources: material, human, andmonetary (Miner, 1978). A knowledge-based view recog-nizes knowledge as another major organizationa l resource(Holsapple & Whinston, 1987). Indeed, Peter Drucker(1993) and others have proclaimed that knowledge is themost important of an organization’s resources. Figure 1a

FIG. 1a. Organizational resources.

portrays this knowledge-centric view of an organization.Arrows in the � gure indicate that knowledge resourcesimpact the other three resources and the external environ-ment, and vice versa. More broadly each type of resourcecan affect each of the other resources as well as the externalenvironment, and vice versa (e.g., there can be tradeoffs,exchanges, substitutions , synergies).

Figure 1a also highlights one of these impacts by rec-ognizing the existence of knowledge management skills inthe organization’s participants. The dashed lines are meantto suggest that some of these skills are human resources(e.g., a person’s cognitive skills), while others may be ma-terial resources (e.g., a computer system’s processing abil-ities). That is, an organization can have both human andmachine-based knowledge workers each having certaincapabilities for representing and processing knowledge(Holsapple et al., 1996). An organization’s knowledge re-sources can be exploited only as far as its participants’knowledge manipulation skills permit. Conversely, the ex-ercise of knowledge manipulation skills depends on what


FIG. 1b. Types of knowledge resources.

knowledge resources are available in the organization andits environment.

In Figure 1b, the knowledge-centric view of an orga-nization is further developed to identify primary typesof knowledge resources. This constitutes the KM frame-work’s resource component. The issue of how participants’knowledge manipulation skills are exercised (i.e., chan-neled into a set of interrelated knowledge manipulationactivities ) is addressed later in discussion of the frame-work’s activities component. Knowledge can be stored,embedded, or represented in an organization as anyof six distinct kinds of resources: (1) participants’knowledge, (2) culture, (3) infrastructure, (4) knowledgeartifacts, (5) purpose, and (6) strategy.

A researcher investigating knowledge management in aparticular organization can study any of these six resourcesor dependencies among them. A practitioner likely wouldbe concerned with all six types of knowledge resources andtheir interdependencies . Aside from the six types of knowl-edge resources within its boundaries, an organization has

access to knowledge existing in its environment. Althoughit does not belong to an organization, it is a crucial sourcefor replenishing and augmenting an organization’s knowl-edge resources. Both researchers and practitioners mustbe cognizant of the relationships between environmentalknowledge resources and the six types of organizationalknowledge resources.

Participants ’ Knowledge. An organization’s partici-pants can be human resources (e.g., employees) and/ormaterial resources (e.g., computer systems). Each partici-pant has certain knowledge management skills and a store-house of knowledge. The knowledge that a participantbrings to bear in the execution of its role within an orga-nization is a knowledge resource of that organization. Be-ing the combination of such knowledge, an organization’sparticipant knowledge resource is thus affected by the ar-rival and departure of participants, participant learning, theportion of each participant’s knowledge that is brought tobear on organizationa l work, and the interrelationship s that


participants’ knowledge are allowed to have (e.g., by or-ganization’s infrastructure and culture).

The portion of an employee’s knowledge that is usedin accomplishing work for an organization is a part of theorganization’s participant knowledge. An organization’sknowledge can also be stored on computer systems, forinstance in a database or an expert system. For example,General Electric’s answer center, USA, has collected cus-tomers’ complaints in a database. This knowledge is partof a system that aids operators in handling customer com-plaints and concerns for 1.5 million potential problems(Sveiby, 1997).

The knowledge stored by a computer system preserves,formalizes, and consolidates knowledge from varioussources. Joe Daniele, corporate manager for intellectualproperty at Xerox Corp., in Rochester, NY, in 1990 af-ter identifying a retirement trend that would have createda huge knowledge de� cit throughout Xerox, recognizedthe need to capture the valuable knowledge of key play-ers in the organization before they left (Crowley, 1997).Similarly, the consulting � rm Integral, Inc., recognizedthe need to create a system that captured and stored theparticipants’ knowledge when 25% of its 40-person staffplanned to go to graduate school (Crowley, 1997). Oncethe knowledge of the human participants is captured andformalized, it can be transferred and shared throughout theorganization.

Culture. Culture is de� ned by Schein (1985) as the“basic assumptions and beliefs that are shared by membersof an organization, that operate unconsciously, and that de-� ne in a basic taken-for-granted fashion an organization’sview of itself and its environment.” An organization’svalues, principles, norms, and unwritten rules and pro-cedures comprise its cultural knowledge resource. Thisresource exists independently of the presence of any par-ticular participants’ knowledge, yet it in� uences each par-ticipant’s use of knowledge as well as the interactionsamong participants’ knowledge. The cultural resource iscomprised of basic assumptions and beliefs that governparticipants’ activities. It is important for KM researchersand practitioners to appreciate this knowledge resourceand the mechanisms whereby it persists and canbe altered.

Historically, culture has been a storehouse of knowl-edge. For instance, persons involved in karate and Japaneseculture point out that karate is not only a mechanismto achieve physical � tness, but also serves as a vehiclethrough which Japanese philosophy is transferred fromgeneration to generation. It is an approach to preservingan organization’s (i.e., a community’s) cultural knowledgeThis transfer is highly nonverbal in nature, communicatedmainly through watching and doing. Another example ofcultural knowledge persistence involves fasting. In India,

the practice of fasting has been observed for centuries.Practiced as frequently as once or twice a week, this ritualis a body-cleansing mechanism and a means to achieveself-control over eating. The cultural knowledge of bodycleansing and self-control is embedded in the form of abelief in the practice of fasting. Interestingly, an activitysimilar to fasting is advocated by several diet programsin the Western world. Also, cultural knowledge has longbeen stored in the form of stories and proverbs, whichare transferred from generation to generation by means ofstorytelling.

The perspective of culture as a knowledge resource canbe recognized by observing organization’s participants’behaviors. For instance, a mill superintendent “champi-oned the ultimately disastrous installation of a $1.5 millionarc saw for cutting � nished beams. He was not penalized,but promoted” (Leonard-Barton, 1995). This encouragedthe values of high tolerance for risk taking and failureat Chaparral Steel. Slowly, the knowledge that a posi-tive attitude toward risk taking is crucial to the organiza-tion’s success became ingrained in its culture. This knowl-edge is manifested in the form of frequent experimentationperformed by employees to solve problems, which allowsChaparral to be innovative and creative.

An organization’s cultural knowledge resource governsparticipants’ behaviors (e.g., knowledge sharing vs.knowledge hoarding). It affects what knowledge is ac-quired and internalized. Leonard-Barton (1995) points outthat “Values serve as a knowledge-screening and -controlmechanism.” That is, the cultural knowledge resource canfunction as a kind of meta-knowledge. It also in� uencesand is in� uenced by the other kinds of organizationa lknowledge resources such as infrastructure, strategy, andpurpose.

Infrastructure. Infrastructure is the knowledge that isused to structure an organization’s participants. It is a for-mal counterpart to an organization’s cultural knowledgeresource. Infrastructure refers to, “the roles that have beende� ned for participants to � ll, the relationships amongthose roles, and regulations that govern the use of rolesand relationships” (Holsapple & Luo, 1996). The roles,relationships , and regulations in force for an organizationare knowledge about the formal structuring of work per-formed by both human and computer participants. Thisinfrastructure knowledge resource governs not only or-dinary organizationa l operations, but also the designing,enabling, monitoring, evaluating, enforcing, and modi-fying of organizational infrastructure (Holsapple & Luo,1996).

Representing knowledge as infrastructure is a means offormalizing existing organizational knowledge that can beused to generate new knowledge (Marshall et al., 1996).Role de� nitions are knowledge about what needs to be


done by participants, about expectations for the partici-pants assigned to the roles (e.g., what knowledge eachis expected to handle or generate). Relationship de� ni-tions are knowledge about what interactions are availableamong participant-� lled roles. The interactions that occurfor one relationship pattern may generate different knowl-edge from those that occur for a different pattern (Chinget al., 1992). Regulation de� nitions are knowledge of for-mal rules and procedures that participants are expectedto observe in � lling their roles and in engaging in rela-tionships . Examples include manufacturing and serviceprocesses, hiring processes, performance appraisal, andreward processes.

Knowledge Artifacts. A knowledge artifact is an ob-ject that represents knowledge. An artifact is not a partici-pant, as it does not have any innate knowledge-processingcapability. Common examples of knowledge artifacts arevideo training tapes, books, memos, business plans in print,manuals, patent documents, � ling cabinet contents, andproducts (e.g., knowledge embedded in a manufacturedcar). An artifact belongs to an organization, but it maybe under the control of or accessible to only certainparticipants.

Knowledge embedded in artifacts can also be repre-sented in other knowledge resources. Representing knowl-edge as an artifact involves explicit embodiment of thatknowledge in an object, thus positively affecting its abil-ity to be transferred, shared, and preserved. For exam-ple, Chaparral Steel’s near-net-shape casting process, inwhich both mold and process are patented (Leonard-Barton, 1995), is represented as two knowledge artifacts:the physical system and a document describing a patentedprocess (thereby preserving and protecting it). It is con-ceivable that this process knowledge also resides with par-ticipants; however, it is the representation as a patent doc-ument that furnishes legal protection and preservation.

Organizational knowledge manifests itself in form ofproducts (Wiig, 1993). Products are not simply a resultof material, capital, and labor resources, but also ofknowledge resources. Knowledge resources guide thetransformation of material, labor, and capital resourcesinto a product. In other words, products in an organiza-tion’s inventory are artifacts representing the knowledgeused to build them. Once a product is released into theenvironment it is no longer an organizationa l resource.But these products can be exchanged for other kinds ofresources (e.g., � nancial). A product’s exchange value isin� uenced by what a customer is willing to pay for thatknowledge. For example, the value of a can of Coke to aconsumer derives largely from the marketing, packaging,and recipe knowledge embodied in it, rather than from thecosts of assembling certain ingredients and distributingthem in containers. A subtle acknowledgment that knowl-

edge is represented in products can be seen in competitors’attempts at reverse engineering.

Purpose. Purpose is knowledge of the reason forwhich an organization exists. It is a knowledge resourcethat indicates an organization’s mission, vision, objectives,and goals. It strongly in� uences the other knowledge re-sources that an organization does or needs to have. Thepurpose resource guides strategy formulation, the resultof which then drives knowledge manipulation activities.If this knowledge is unclear, inadequate, and not carefullyevaluated, then an organization may formulate and im-plement strategies that are detrimental to organizationalperformance. For example, Sears, whose purpose is to sellconsumer goods, bought the investment � rm Dean Witter.This turned out to be a failure because consumers did notconsider their � nancial needs to be satis� able by a “con-sumer product” (Drucker, 1994). In this case, the knowl-edge about purpose (i.e., what comprises consumer goods)was inadequate or unused.

Strategy. Strategy is knowledge about what to do inorder to achieve organizational purpose in an effectivemanner. This knowledge resource indicates plans for us-ing an organization’s infrastructure, culture, knowledgeartifacts, and participants’ knowledge (as well as other or-ganizational resources). For instance, it can be plans forpromoting a product or achieving effective resource allo-cation. A purpose of Pepsi and Chaparral Steel is to sus-tain their leadership positions in their respective markets.However, strategies needed to achieve the same purposeare very different for each � rm. Pepsi’s strategies focusmainly on gaining competencies in the areas of marketingand sales, whereas Chaparral Steel strategies focus primar-ily on competencies that allow it to continually improveand innovate its production processes (Leonard-Barton,1995). In this example, Pepsi and Chaparral Steel haveverydifferent sets of strategic knowledge on how to sustain mar-ket leadership. And thus activities for acquiring and cul-tivating such knowledge would be different for Pepsi andChaparral Steel.

External Environment. An organization’s environ-ment is populated with many entities that are potentialsources of knowledge. Through contact with these entities,an organization can augment and replenish its knowledgeresources. Environmental knowledge may be viewed asa virtual knowledge resource for an organization. It doesnot actually belong to an organization, nor is it controlledby the organization, but it can be accessed or acquiredfrom the environment. This may or may not be dif� cultor expensive. The World Wide Web’s content is a virtualknowledge resource that is relatively easy and inexpen-sive to tap, albeit of variable quality. The Johns Hopkins


networked database for genetic research is an example ofa virtual knowledge resource. This database is the “oneand only one of� cial record of every gene and piece ofDNA that’s mapped in the world. : : : [It also] captures andre� ects the ongoing wisdom” of experts from all over theworld (Anthes, 1991, p. 28). Thousands of medical re-searchers and practitioners access this knowledge resourcethat is virtually their own.

Discussion. The six types of knowledge resourcesidenti� ed here are distinct. A given purpose does not dic-tate a particular strategy, although it may limit candidatestrategies. Conversely, a given strategy does not imply aparticular purpose. Similarly, a given culture does not dic-tate a particular strategy or purpose; nor is a given strategyor purpose necessarily incompatible with all but a singlecultural variation. Infrastructure does not dictate culture,and vice versa. Participant knowledge is different fromany of these (e.g., it is a knowledge resource that can walkout the door). Although artifacts may hold renditions ofany of the other knowledge resources, they are not thoseother resources. Although distinct, the six types of organi-zational knowledge resources are also highly interrelated.They constrain, facilitate, and reinforce each other. It isoutside the scope of this article, but we do note the impor-tance of appreciating relationships among knowledge re-sources as a basis for bringing them into proper alignment.

An organization’s knowledge resources can be char-acterized in terms of various attributes. Although the re-source types are nouns, the attributes are adjectives thatqualify them in various ways. The literature contains avariety of attribute dimensions, including knowledgeusage—practical versus intellectual versus pastime ver-sus spiritual versus unwanted (Machlup, 1982), knowledgetype—descriptive versus procedural versus reasoning(Holsapple & Whinston, 1988), knowledge mode—explicit versus tacit (Nonaka, 1994), knowledge subject—domain versus relational versus self (Holsapple et al.,1996), and knowledge quality, involving knowledge util-ity and validity measures (Holsapple & Whinston, 1996).For instance, usage of the latter attribute dimension can beseen in Integral Incorporated’s Knowledge Bank System,which requires that each internalized unit of knowledgepass through a value check or quality assurance procedureand be assessed for assignment into one of three qualitygrades (Crowley, 1997). Studying the application of at-tribute dimensions to each of the framework’s resourcetypes is an interesting direction for future research.

The Knowledge Manipulation Activities Component

Having identi� ed thegeneric types of knowledge resourcesan organization possesses, we now formulate a set ofgeneric types of manipulation activities an organization

undertakes in working with those resources. These knowl-edge manipulation activities are an expression of partici-pants’ knowledge manipulation skills. Skill is the abilityto apply one’s knowledge effectively and readily to execu-tion and performance (Merriam Webster’s Collegiate Dic-tionary, 10th ed., 1995). Knowledge manipulation skillsare exhibited by human and/or material resources as theyoperate on knowledge resources.

From one organization to the next; we contend thereis a set of interrelated knowledge manipulation activitiesthat is common. These form a starting point for KM re-searchers investigating how knowledge is processed in par-ticular organizations and for comparisons across time ororganizations . They highlight major activities with whicha chief knowledge of� cer needs to be concerned. Partici-pants’ knowledge manipulation skills should be cultivated,harnessed, and organized in the performance of these ac-tivities. Thus, an organization’s knowledge resources willbe used in creating value.

The second component of the framework is illustrated inFigure 2. It retains the knowledge resources from Figure 1band further develops the participants’ knowledge manipu-lation skills from the � gure in terms of activities performedwith those skills. The framework’s four knowledge ma-nipulation activities are acquiring knowledge, selectingknowledge, internalizing knowledge, and using knowl-edge. The latter refers to the activities of externalizingknowledge and generating knowledge. The arrows inFigure 2 indicate knowledge � ows from one activity orresource to the another. Aside from the knowledge � ows,activities interact by sending and receiving messages (e.g.,requests, commands). For simplicity, these message � owsare not represented in the � gure. A message in the form ofa request can range from procedural (specifying how theactivity should be carried out) to nonprocedural (merelyindicating what is needed). It can range from explicit (e.g.,a command) to implicit (e.g., involving recognition of aneed). It may require a fast response or tolerate perfor-mance of an activity in the background. It can range froma one-time request to a standing request that requires con-tinual monitoring.

Each activity can be performed individually by partici-pants in an organization or may be carried out by con� gura-tions of multiple participants. For instance, an individual ’sknowledge manipulation skills can be applied to acquir-ing a unit of knowledge for the organization or the skillsof multiple participants may jointly be brought to bearon acquiring that knowledge. An individual participantmay exercise knowledge manipulation skills to engagein multiple activities. For instance, a person or a com-puter system may participate in both acquiring and gen-erating knowledge. How participants’ skills are deployedwith respect to accomplishing the activities identi� ed inFigure 2 depends on an assortment of factors: resource


FIG. 2. Major knowledge management activities.

in� uences (e.g., culture, infrastructure, � nancial resources,participants’ knowledge manipulation skills), environmen-tal in� uences (e.g., dynamics of the environment), andmanagerial in� uences (e.g., leadership style, coordinationapproach that governs assignment of participants to roles).These in� uences on the conduct of KM are addressed inthe framework’s third component. Here the activities alongwith their relationships are described.

Acquiring Knowledge. Acquiring knowledge is theactivity of accepting a unit of knowledge from the externalenvironment and transforming it into a representation thatcan be internalized, and/or used within an organization.Subactivities involved in acquiring knowledge include:

� Extracting knowledge from external sources. Thisincludes locating, accessing, capturing, andcollecting knowledge from external sources intothe organization (e.g., customers, competitors,suppliers, universities , consultants , governmentagencies).

� Interpreting the extracted knowledge. This in-volves transforming extracted knowledge into

representations that can be understood and pro-cessed by another knowledge manipulation activity.

� Transferring the interpreted knowledge. Thistransferal can be to an activity that immediatelyuses the knowledge or to one that internalizes itwithin an organization for subsequent use.

Acquiring knowledge from customers is illustrated bythe example of Nike’s innovative foot-measuring technol-ogy, called NGAGE (Levine, 1997). A customer placesa foot on a screen and a personalized card detailing cus-tomer’s foot length, forefoot length and width, arch length,height and heel width, as well as recommended productsize is printed. This not only helps customers get right-sized shoes, but also allows Nike to update its sizing tablesand later to use this information for research and develop-ment of new footwear. Acquisition of knowledge in thiscase involves extraction from a customer, interpretation ofa foot image, and transferal of resultant foot characteristicsfor internalization in sizing tables.

Selecting Knowledge. Selecting knowledge is an ac-tivity of extracting a requested unit of knowledge from


internal knowledge resources and providing it in an ap-propriate representation to a requesting activity (i.e., to theacquiring, using, or internalizing activities). Subactivitie sinvolved in selecting knowledge include:

� Locating requested knowledge within the internalknowledge resources.

� Retrieving the located knowledge. This involvescapturing and/or collecting knowledge from lo-cated organizational knowledge resources, and as-sembling/organizing/packaging it in a representa-tion appropriate for the requesting activity.

� Transferring retrieved knowledge to an appropri-ate activity. This transferal can be used to supportthe acquiring, internalizing, or using activities.

An example of the selecting activity is illustrated byBuckman Labs’ knowledge sharing system, K’Netix(Rifkin, 1997). For instance, a managing director of Asianfacilities sent a message to all the employees in thecompany requesting knowledge about pitch-control strate-gies in their parts of the world. Within a few hours he re-ceived 11 suggestions addressing his request and enablinghim to secure a $6 million order. Here, the K’Netix partici-pant broadcasted the message to all the human participantsfor the purpose of locating an appropriate unit of knowl-edge. Once this knowledge was located, K’Netix collected,packaged, and transferred it to the person requesting theknowledge.

Internalizing Knowledge. Internalizing is an activitythat alters an organization’s knowledge resources based onacquired or generated knowledge. Subactivities involvedin internalizing knowledge include:

� Assessing the knowledge to be internalized. Thisis concerned with determining the suitability ofthe knowledge for internalization.

� Targeting the assessed knowledge. This identi� esknowledge resources that are to be impacted bythe new knowledge.

� Depositing the knowledge as targeted. This in-volves modifying existing knowledge resourcesby adding to them, deleting from them, or per-haps fundamentally restructuring them. It involvesdisseminating , distributing , sharing, and diffusingknowledge to targeted knowledge resources.

McKinsey and Bain & Co. has established a computersystem that holds experiences from various team assign-ments (Sveiby, 1997). Knowledge generated from eachassignment is internalized in the system’s database. Later,it can be selected by employees for future assignments.In this example, suitability of knowledge to be internal-ized is assessed by team members, who then target a

computer system where it is to reside and take care ofdepositing it.

Using Knowledge. Using knowledge is the activity ofapplying existing knowledge to generate new knowledgeand/or to produce an externalization of knowledge.

Generating Knowledge. Generating knowledge is anactivity of producing a unit of knowledgeby processing ex-isting units of knowledge, where the latter are the results ofselection, acquisition, and/or prior generation. The knowl-edge generated may be new to the organization. This is verycrucial because “new knowledge provides the basis fororganization renewal and sustainable competitive advan-tage” (Quinn, 1992). Alternatively, it may currently existor have previously existed in the organization’s knowledgeresources. Generation of knowledge that is not “new” canoccur for economic reasons (e.g., it is cheaper to producethan to select or acquire), for training reasons, for validitychecking, due to a lack of awareness about its existence,or due to not having been internalized when previously ac-quired or generated. Subactivities involved in generatingknowledge include:

� Monitoring the organization’s knowledge re-sources and the external environment.

� Evaluating selected or acquired knowledge.

� Producing knowledge. This can involve creating,developing, analyzing, constructing, synthesizing ,re� ning, and/or assembling knowledge from ex-isting knowledge.

� Transferring the generated knowledge for exter-nalization and/or internalization.

There are two types of generation: deriving knowledgeand discovering knowledge. Deriving generates knowl-edge by structured use of procedures or rules that areknown to the organization. Examples include calculatinga forecast for product demand, deriving an optimal orderquantity using the EOQ formula, or inferring a recommen-dation from a set of rules. Discovery generates knowledgein less structured ways. The exact path toward the gen-erated knowledge cannot be fully preconceived or eventraced. Herman Helmholtz, a German physiologis t andphysicist, described his scienti� c discoveries by sayingthat he went through three stages: saturation , � nding outeverything he could learn on the subject; incubation , re-� ecting on what has been absorbed, by thinking about andmulling over what he has learned through the research; andillumination , arriving at a sudden solution. French math-ematician Henry Poincare added a fourth stage to this,which he called veri� cation (Carson, 1992).

Externalizing knowledge. Externalizing knowledge isthe activity of using existing knowledge to produceorganizationa l outputs for release into the environment.


Externalization is only partially a knowledge manipula-tion activity because it can involve physical activities suchas the act of producing a product through transformationof raw materials. However, the � ow of material can beseen as secondary to the � ow of knowledge that enables,facilitates, and guides it (Cook et al., 1995). The external-ization box in Figure 2 is not completely subsumed withinthe participants’ knowledge manipulation skills area as itcan use other skills (e.g., manual skills) participants have.Subactivities involved in externalizing knowledge include:

� Targeting the output. This is a determination ofwhat needs to be produced for targeted elementsof the environment (e.g., certain customers).

� Producing the output. This involves applying, em-bodying, controlling, and leveraging existingknowledge to produce output for the target. Theoutput is a representation of the knowledge usedto produce it.

� Transferring the output. This is concerned withpackaging and distributing the representations ofknowledge that have been produced for targets inthe environment.

Examples of externalization include manufacturing acar, providing technical support, or developing a � nan-cial portfolio. For instance, manufacturing a car involvestargeting the car to a speci� c market (low-income groupvs. high-income group). Producing a car by applying theproduct and process design knowledge to manufacture itand then transferring the car into the external environment.The car that is released into the market is a representationof knowledge (disassembling the car during reverse engi-neering would reveal at least some of the knowledge thatwent into manufacturing it).

Knowledge Management In�uences

The conduct of KM in an organization involves the useof knowledge manipulation skills to perform knowledgemanipulation activities with respect to the organization’sknowledge resources. Although both knowledge andknowledge manipulation skills are crucial and necessaryfor the operation of a knowledge-based organization, theyare not suf� cient. Which of the activities in Figure 2 areexecuted, when, by which participants, and with respectto which knowledge resources? The third component ofthe framework aims to help practitioners and researchersanswer such questions for speci� c organizations by rec-ognizing factors that in� uence this conduct of KM.

Stewart (1997) describes the experience of FrankOstroff of Perot Systems as a college student working ata tire-making factory. He and others were given the jobof applying glue to rubber for tire after tire. These samepeople upon leaving work for the day would spend their

evenings rebuilding cars or running volunteer organiza-tions. Although this tire-making organization had knowl-edgeable and skilled employees, who could rebuild a caror run a volunteer organization, it did not use its partici-pants’ expertise to create value for the organization. Littleof the participants’ individua l knowledge was made into anorganizational knowledge resource. Their knowledge ma-nipulation skills were largely unused in the organization’sexecution of knowledge manipulation activities. Althoughsome level of KM undoubtably occurred in this organiza-tion, its conduct of KM was probably far less effective thanit could have been.

The conduct of KM in an organization is in� uencedby a variety of factors. The third component of the KMframework identi� es three classes of factors: resource in-� uences, managerial in� uences, and environmental in� u-ences. In Figure 3, the solid circle represents KM conduct,its inner core represents the essential results of KM con-duct (i.e., projection and learning), and each angle of thetriangle represents a class of in� uences. Appreciating thesein� uences is important for KM researchers in describingand prescribing how KM is or should be accomplished.Chief knowledge of� cers need to be cognizant of such in-� uences as constraints on their efforts to create knowledge-based organizations as well as levers that can help them todo so.

Resource In� uences. Both knowledge resources(Andersen & APQC, 1996) and other resources (Inkpen,1996) affect the way in which KM is conducted in an orga-nization. Financial resources could put a ceiling on the cap-ital expended on knowledge manipulation activities. Sim-ilarly, participants’ knowledge manipulation skills (e.g.,human resources) both constrain and facilitate knowledgemanipulation activities.

Each of the six types of knowledge resources in� uencesthe conduct of KM in an organization. That is, the knowl-edge an organization has in� uences the nature and out-come of its knowledge work. For instance, Kodak’s culturehas valued chemical engineering knowledge related to � lmdesign more than the mechanical engineering associatedwith equipment design (Leonard-Barton, 1995). Chemicalengineers were assured good compensation and challeng-ing assignments; consequently, Kodak attracted the bestchemical engineers, but not top-notch mechanical engi-neers. This proclivity likely enhanced the organization’sinnovations and projections related to chemical processesrather than mechanical processes. Similarly, KM conductis in� uenced by infrastructure, strategy, purpose, knowl-edge artifacts, and available participant knowledge.

Managerial In� uences. Theconduct of KM is affectednot only by the existence of various resources, butalso by the deployment of these resources. Here is where


FIG. 3. Knowledge management in� uences.

managerial in� uences on KM conduct come into play.Such in� uences govern the state of an organization’sknowledge resources and the use of knowledge manip-ulation skills in performing the activities in Figure 2. Asindicated in Figure 3, managerial in� uences on the con-duct of KM include the factors of leadership, coordination,and measurement.

Leadership. Leadership has been recognized as a fac-tor in� uencing the conduct of KM (Andersen & APQC,1996; Inkpen, 1996). The characteristics of leaders rangefrom being manipulators of culture (Schein, 1985) to ar-chitects and catalysts (Hedlund, 1994). In writing aboutleadership, Mort Meyerson (1997), CEO for Perot Sys-tems, says “the way to be a leader today is different” asleaders can no longer function as sole decision makers theway they could 15 years ago at EDS when competitionwas stable. Meyerson states that the essence of leadershiptoday is to make sure that an organization knows itself.

In today’s knowledge-based economy a successfulleader will be one who can effectively manage both or-ganizational knowledge resources and associated knowl-

edge manipulation skills. He or she creates conditions thatallow participants to readily exercise and cultivate theirknowledge manipulation skills, to contribute their own in-dividual knowledge resources to the organization’s pool ofparticipant knowledge, and to have easy access to relevantknowledge resources. A study conducted by Andersen andAPQC revealed that one crucial reason why organizationsare unable to effectively leverage knowledge is because ofa “lack of commitment of top leadership to sharing organi-zational knowledge or there are too few role models whoexhibit the desired behavior” (Hiebeler, 1996).

Coordination. Coordination refers to managing de-pendencies among activities (Malone & Crowston, 1994).It aims to harmonize activities in an organization by ensur-ing that proper resources are brought into play at appro-priate times and that they adequately relate to each otherduring the conduct of activities (Holsapple & Whinston,1996). In the conduct of KM, dependencies that need tobe managed include those among knowledge resources(e.g., alignment of participants’ knowledge with strategy,diffusion of knowledge among participants), those among


knowledge management activities (e.g., which activitiesare undertaken under varying circumstances), those be-tween knowledge resources and other resources (e.g., what� nancial resources are to be allocated for knowledge ma-nipulation activities, which participants are assigned towhich infrastructure roles), and those between resourcesand knowledge management activities (e.g., use of knowl-edge manipulation activities to improve knowledge re-sources, allocating knowledge resources among compet-ing knowledge manipulation activities). The conduct ofKM in an organization is strongly in� uenced by how suchdependencies are managed.

How the knowledge manipulation activities of Figure 2are coordinated de� nes an organization’s approach toproblem solving, decision making, experimentation, andorganizational learning—all of which are knowledge-intensive endeavors. Coordination approaches suggestedand used to manage dependencies in a knowledge-basedorganization include linking reward structures to sharing,establishing communications for knowledge sharing, andconstructing programs to encourage learning. For example,in order to create a knowledge sharing culture at Integral,Inc., management made knowledge sharing part of the per-formance review (Crowley, 1995). To get good reviews(and large bonuses), participants had to share their knowl-edge. At one consulting � rm, professionals are expectedto document what they have learned about what works andwhat does not work, and they are partially compensatedbased on how often their documentation is accessed froma central knowledge repository (Marshall et al., 1996).

At Buckman Labs, incentive, evaluation, and promotionsystems are designed to reward employees who share andtransfer knowledge and punish those who do not (Rifkin,1997). This knowledge sharing and transfer is facilitatedby establishing open communications via trust buildingand the installation of the K’Netix communication sup-port system. What makes the K’Netix system work is theculture of trust instilled by company leadership (Rifkin,1997). This illustrates an important point about KM in-� uences identi� ed in the framework: Although the in� u-ences are distinct concepts, they are related to and canimpact each other. The ef� cacy of coordination can beimpacted by leadership (another managerial in� uence onKM). Culture (a knowledge resource in� uence) can affectthe fashion in which coordination can be conducted andcoordination, in turn, can lead to cultural change.

Measurement. Another managerial in� uence on theconduct of KM is the installation of mechanisms for mea-suring knowledge resources, knowledge manipulationskills and activities, and the results of KM in terms of or-ganizational learning and projection. Such measurementbecomes a basis for evaluation of leadership, coordina-tion, and resources. It can indicate where adjustments inthese may be needed.

Ultimately, measurement may help evaluate the impactof an organization’s KM on bottom-line performance. In-terestingly, this is an under-implemented area (Hiebeler,1996). A few organizations have developed a set of indi-cators for KM measurement. For example, the Swedish� rm Celemi published the world’s � rst audit of its in-tangible assets in its 1995 Annual Report; Skandia usesnon� nancial indicators to measure its processes and pub-lished the � rst annual report supplement on intellectualcapital (Sveiby, 1997). Chris Turner, “learning person”for Xerox, believes that anecdotal evidence is more valu-able than hard data in measuring if KM is occurring orworking for a company (Webber, 1997). Turner empha-sizes the point that measurement indicators need not behard and � nancial, but can be soft and non� nancial. Whatis important is linking these indicators to the � nancialresults.

The feasibility of measuring knowledge resources orprocesses and linking them to � nancial results is not onlydif� cult but also controversial. There exist two schools ofthought: One believes knowledge assets and processes canbe measured (Lev, 1997; Malone, 1997; Stewart, 1997),and the other contends “you’re a fool if you buy intothis” (Rutledge, 1997). Whatever the case, the frameworkin Figure 3 posits that the conduct of KM is in� uencedby (1) whether an organization attempts to measure itsknowledge resources and/or performance of its knowledgemanipulation activities, (2) how it goes about measuringthese, and (3) how effective the measures are.

Environmental In� uences. Aside from internal factors(resource and managerial in� uences), entities outside anorganization also affect its conduct of KM. The environ-ment determines or constrains what knowledge resourcesshould or can be acquired, as well as what the knowl-edge manipulation skills are available (via a labor pool oravailable technology). As Figure 3 illustrates, external in-� uences include such factors as competition, customers,markets, suppliers, and the GEPSE (governmental, eco-nomic, political, social, and educational) climate. Exam-ples of these are many, varied, and largely self-evident.Detailed investigation s of these are warranted.

SURVEY AND RESULTS

Aside from conceptual justi� cation for the framework, sur-vey results provide external validation. A list was compiledof contributors to the KM literature, presenters at KM con-ferences, and faculty who designate KM as a major re-search area. Each of the 122 persons on this list receivedthe framework description as presented in this paper, plusa questionnaire . Of the 31 respondents, 13 identi� ed them-selves as researchers, 13 as practitioners, and the rest asboth researcher and practitioner.


Respondents were asked to assess each of the threeframework components on four criteria: completeness, ac-curacy, clarity, and conciseness. Their perceptions regard-ing the overall framework were also captured by askinghow satis� ed they were with the overall framework andthe extent of the framework’s success in providing a uni-� ed and comprehensive view of KM phenomena. Theseperceptions were gathered using Likert-scale items andclustered into three categories: low, medium, high. For in-stance, responses indicating no success to slight successform the low category; those in the somewhat successful tomoderately successful range form the medium category;responses in the successful to extremely successful rangebelong to the high success category. In addition to scaleditems, respondents were also asked to provide observa-tions on the need for a KM framework (responses wereshown earlier in Table 1).

Distributions of responses are presented in Figures 4and 5. Figure 4 shows frequency distribution s of responsesconcerning the overall framework. The respondents arenearly unanimous in their perception that it is highlyimportant to have a KM framework. Overall satisfactionwith the threefold framework was indicated to be mediumor high by over 80% of the respondents. Similar resultswere obtained for the framework’s success in providing a

FIG. 4. Responses concerning the overall framework.

uni� ed view and a comprehensive view of the main factorsinvolved in KM.

Figure 5 shows frequency distribution s of responses onthe four criteria for each of the framework’s three compo-nents. For each of the 12 assessments, at least 70% regardthe framework as in the medium to high success categories.The mode is high success for 9 of the 12 and medium suc-cess for the other three assessments. On the whole, theseresults are supportive of the framework, but do suggestthat future modi� cations to the framework should focus onenhancing the accuracy and clarity of the resource com-ponent and the completeness of the activity component.

IMPLICATIONS AND CONCLUSION

In its current stage of development, the framework hasseveral uses. It provides a language (i.e., a system of termsand concepts) for discourse about and study of KM phe-nomena. It can be used to generate and frame KM re-search issues. It identi� es factors with which KM practi-tioners should deal. Survey results are positive, but alsosuggest there is room for improvement in the framework.The framework itself serves as a basis for thinking aboutextensions, re� nements, or corrections that could yieldan improved KM framework. We conclude with a brief


FIG. 5. Responses for each of the framework’s component.

discussion that highlights a few avenues of subsequent in-vestigation implied by the framework.

Each of the three main components and any of theirsubcomponents can be analyzed in greater detail. That is,the framework is a starting point for gaining a deeper un-derstanding of any of its elements and of relationshipsthat an element may have with other elements. For exam-ple, the knowledge selection activity has been examined ingreater detail, leading to an identi� cation of issues relatedto knowledge selection that deserve consideration by KMresearchers and practitioners, plus a characterization ofcurrent technological offerings for performing and sup-porting knowledge selection (Holsapple & Joshi, 1999b).The threefold framework can be applied similarly to guidethe study of other knowledge manipulation activities or ofelements in the in� uence and resource components.

The framework provides building blocks for devisingand investigating prescriptive frameworks. KM methodscan be designed and developed to guide the temporal pat-tern of knowledge manipulation activities that should be

used to satisfy a knowledge need. Prescriptions about whatknowledge resources deserve greater cultivation and howto conduct that cultivation are another possibility.

Each of the manipulation activities identi� ed in Figure 2can be explored from such standpoints as how to effec-tively perform it relative to available knowledge resources,how to employ technology in implementing the activity,and how to utilize it in the scope of more composite ac-tivities. Fulcrum, for example, has implemented softwarethat performs the selection activity with respect to a diverseset of knowledge repositories (Fulcrum, 1997). Compositeknowledge manipulation activities include problem solv-ing, experimenting, decision making, and learning. Eachcan be characterized in terms of patterns of the elementalmanipulation activities and each is susceptible to techno-logical support (e.g., a knowledge-based view of decisionmaking; Holsapple, 1995).

The threefold framework can be employed to orga-nize an exploration of available technologies . What typesof technologies can support and/or perform each of the


knowledge manipulation activities? What types of tech-nologies facilitate KM initiatives (e.g., measuring, control-ling, and coordinating knowledge manipulation activitiesand resources)? What types of technologies can be usedfor storing, representing, and embedding knowledge? Howcan technology affect projection and learning? Investiga-tions to answer these questions can help clarify the role oftechnology in KM.

The framework suggests a need for investigating pos-sible linkages between resource, managerial, and envi-ronmental in� uences on the one hand and the outcomeof KM conduct (i.e., organizational learning and projec-tions) on the other hand. For a given set of environmentaland resource in� uences, how might we select from coor-dination, leadership, and measurement alternatives? Suchquestions appear to be amenable to empirical investiga-tions. At a more fundamental level, there is the issue ofidentifying these alternatives as a prelude to assessing theiref� cacy.

A study of in� uences on the conduct of KM is alsoimportant from the standpoint of detecting de� cienciesin knowledge resources and in the management of thoseresources. Such de� ciencies, if untreated, can lead to mal-function in organizational processes and damage of overallperformance. The in� uences identi� ed in Figure 3, as wellas the activities noted in Figure 2, need to be investigatedwith respect to how they can be shaped to remedy problemsin the conduct of KM.

One formal way to further develop the threefold frame-work is by means of a Delphi-like methodology (Bacon &Fitzgerald, 1996). This involves identifying a panel of ex-perts on KM. The experts are provided with the threefoldframework and asked to comment on its accuracy, com-prehensiveness , clarity, and conciseness. These commentsgive a basis for revising the framework. The revised frame-work is then submitted to the panel for a second round ofcomment and critique. This process continues until con-sensus is reached or panelists express no major reserva-tions about the framework. In this way, the framework isbene� cial in leading to its own obsolesence.

REFERENCES

Amidon, D. 1996. The momentum of knowledge management.hhttp://www.entovation.com/momentum/momentum.htmi

Anthes, G. 1991. A step beyond a database. COMPUTERWORLD4 March:28.

Andersen, Arthur, and the American Productivity and Quality Center.1996. The knowledge management assessment tool: External bench-marking version.

Bacon, J., and Fitzgerald, B. 1996. The � eld of IST: A name, a frame-work, and a central focus. hhttp://www.ucc.ie/esrc/ist.htmli

Carson, B. 1992. Think big—Unleashing your potential for excellence.New York: Harper.

Ching, C., Holsapple, C., and Whinston, A. 1992. Reputation, learning,and organizational coordination. Organization Science 3(2):85–98.

Cook, D., Chung, C., and Holsapple, C. 1995. Information � ow � rst,material � ow next! APICS—The Performance Advantage 5(1):38–

39.Crowley, A. 1997. Memory bank. PC Week 6 January:101.Drucker, P. 1993. Post-capitalist society. New York: Harper & Collins.Drucker, P. 1994. The theory of business. Harvard Business Review

72(5):95–104.Fulcrum. 1997. The search for knowledge begins here! hhttp://www.

fulcrum.com/homepage.htmiHedlund, G. 1994. A model of knowledge management and the

N-form corporation. Strategic Management Journal 15(summer):73–90.

Hiebeler, R. 1996. Benchmarking knowledge management. Strategyand Leadership 24(2):22–29.

Holsapple, C. 1995. Knowledge management in decision making anddecision support. Knowledge & Policy 8(1):5–22.

Holsapple, C., and Joshi, K. D. 1999a. Description and analysis ofexisting knowledge management frameworks. Proceedings of theHawaiian International Conference on System Sciences, Maui.

Holsapple, C., and Joshi, K. D. 1999b. Knowledge selection: Concepts,issues, and technology. In The handbook of knowledge management,ed. J. Liebowitz, pp. 7-1–717, Boca Raton, FL: CRC Press.

Holsapple, C., and Luo, W. 1996. A framework for studying computersupport of organizational infrastructure. Information and Manage-ment 31(1):13–24.

Holsapple, C., and Whinston, A. 1987. Knowledge-based organiza-tions. The Information Society 5(2):77–90.

Holsapple, C., and Whinston, A. 1988. The information jungle: A quasi-novel approach to managing corporate knowledge. Homewood, IL:Dow Jones-Irwin.

Holsapple, C., and Whinston, A. 1996. Decision support systems: Aknowledge-based approach. St. Paul, MN: West.

Holsapple, C., Johnson, L., and Waldron, V. 1996. A formal model forthe study of communication support systems. Human Communica-tions Research 22(3):422–447.

Inkpen, A. 1996. Creating knowledge through collaboration. CaliforniaManagement Review 39(1):123–139.

Leonard-Barton, D. 1995. Wellsprings of knowledge. Boston: HarvardBusiness School Press.

Lev, B. 1997. The old rules no longer apply. Forbes ASAP 7 April:35–

36.Machlup, F. 1982. Knowledge: Its creation, distribution, and eco-

nomic signi� cance—The branches of learning, vol. II. Princeton,NJ: Princeton University Press.

Malone, M. 1997. New metrics for new age. Forbes ASAP 7 April:40–

41.Malone, T., and Crowston, K. 1994. The interdisciplinary study of

coordination. ACM Computing Surveys 26(1):87–119.Marshall, C., Prusak, L., and Shpilberg, D. 1996. Financial risk and the

need for superior knowledge management. California ManagementReview 38(3):77–101.

Meyerson, M. 1997. Everything I thought I knew about leadership iswrong. Fast Company [Special Collector’s Edition: New Rules ofBusiness] May:4–11.

Miles, B., and Huberman, M. 1994. Qualitative data analysis. London:Sage.


Miner, J. B. 1978. The management process: Theory, research, andpractice, 2nd ed. New York: Macmillan.

Newell, A. 1982. The knowledge level. Arti� cial Intelligence 18(1):87–

127.Nonaka, I. 1991. The knowledge creating company. Harvard Business

Review 69(6):96–104.Nonaka, I. 1994. A dynamic theory of organizational knowledge

creation. Organization Science 5(1):14–37.Quinn, J. B. 1992. The intelligent enterprise. New York: Free Press.Rifkin, G. 1997. Buckman Labs nothing but Net. Fast Company.

hhttp://www.fastcompany.com/03/buckman.htmliRutledge, J. 1997. You’re a fool if you buy into this. Forbes ASAP

7 April:43–45.Schein, E. 1985. Organizational culture and leadership. San Francisco:

Jossey-Bass.Stewart, T. 1997. Intellectual capital. New York: Doubleday/Currency.

Stuart, A. 1996. Uneasy pieces—Part 5. hhttp://www.cio.com/CIO/060196 uneasy 1.htmli

Sveiby, K. 1997. What is knowledge management? hhttp://www.svieby.com.au/KnowledgeManagement.htmli

Tof� er, A. 1990. Powershift: Knowledge, wealth and violence at theedge of the 21st century. New York: Bantam Books.

van Lohuizen, C. 1986. Knowledge management and policy making.Knowledge: Creation, Diffusion, Utilization 8(1):12–38.

Vatter, W. J. 1947. The fund theory of accounting and its implicationsfor � nancial reports. Chicago: University of Chicago Press.

Webber, A. 1997. XBS learns to grow. Fast Company. [Special Collec-tor’s Edition: New Rules of Business] May:44–51.

Wiig, K. 1993. Knowledge management foundations. Arlington, VA:Schema Press.

Zack, M., and Serino, M. 1997. Knowledge management and collabo-ration technologies. hhttp://www.lotus.com/institute/2102.htmi

knowledge management: a threefold framework

Documents