New Knowledge Management

June 200312

special focus>the new knowledge management

The future value of knowledge management in a corporate context is dependent on the discipline’s ability to overcome many of the limitations of its current guise. Drawing on research conducted for their most recent book,Joseph M. Firestone and Mark W. McElroy discuss eight issues that they feel will define what they call ‘new knowledge management’.

What is the future of KM? If KM is tohave a future, it must give better

answers to such fundamental questions aswhat is knowledge, what is knowledgemanagement, where does knowledge comefrom, and what roles do learning andknowledge play in business performance?We believe, also, that the ‘new knowledgemanagement’, a perspective we and othershave developed over the past few years,answers these questions well, and that thefuture of KM lies within its vision. While we’dlike to present a detailed description of thisperspective in this article, the broad range ofits various frameworks would greatly exceedthe space available. We’ve decided, instead,to present brief discussions of eight issuesthat, based on analysis in our new book, KeyIssues in the New Knowledge Management1,will be very important over the next fiveyears. We intend that these discussions willprovide some of the flavour of the new KMand our very great excitement in developingit. The issues are:

! Freeing KM from the bonds of strategy;! Transcending the Nonaka and Takeuchi

SECI model;! Developing the enterprise knowledge portal;! Developing a comprehensive system of

KM metrics;! Developing the open enterprise;! Creating communities of inquiry;! Developing value theory in KM;! Transcending KM standards development.

This article will briefly characterise eachissue, evaluate its importance and relate itto the future of KM. But before we get to theissues, we need some background on someimportant distinctions in the new KM.

First, the new KM makes a distinctionbetween knowledge management,knowledge processing (KP) and businessprocessing (BP) (see figure 1). This three-

tier perspective is key to the future of KMbecause it formally specifies the role thatKM should play relative to a range ofbehaviours in organisations that shapeknowledge production and integration.Such KP behaviours are always present in organisations, but with KM they can be enhanced. In fact, the new KM says that the purpose of KM is to enhanceknowledge processing, which, in turn,enhances knowledge outcomes, and BPperformance and related outcomes.

Second, the conventional practice of KM begins with the assumption thatvaluable knowledge already exists. KM is all about getting the right information to theright people at the right time. Knowledgedoes not simply exist, however – peoplecreate it. And KM can help them do thisbetter through its impact on knowledgemaking or production.

The new KM focuses on the whole ofknowledge processing, both knowledge inte-gration (including sharing) and knowledgeproduction. We refer to approaches to KMthat deal only with knowledge sharing andintegration as first-generation or supply-sideKM. We refer to newer approaches to KMthat deal with both knowledge integrationand knowledge making as second-generationKM, or demand and supply-side KM.

Second-generation KM, or at least its new variant, provides something essential to KM. By focusing on the sub-process ofknowledge production called knowledge-claim evaluation (KCE), one can distinguishknowledge from information. Put simply,knowledge is comprised of those knowledgeclaims that survive the KCE process. Otherknowledge claims are either false or ‘just’information. So a focus on knowledgeproduction and on KCE and its outcomes arekey to clearly distinguishing both knowledgefrom information and KM from informationmanagement (IM). Therefore, until and

The new knowledge management

Joseph M. Firestone is chief knowledgeofficer for Executive Information Systems andexecutive vice president, Education, Research

and Membership at the KnowledgeManagement Consortium International. He

can be contacted at [email protected]

Mark McElroy is president and CEO ofMacroinnovation Associates. He can be contacted at [email protected]

June 2003 13


unless we expand the scope ofKM to address knowledgeproduction and KCE, KM willbe forever seen as little morethan IM in disguise. Its valuepropositions will be discountedaccordingly.

Freeing KM from the bonds of strategyIn contemporary theory andpractice, KM is subordinate tostrategy. This idea is frequentlyexpressed in the form of dicta, which suggest that KMinitiatives must always bealigned with and supportstrategy and, in turn, besupported by management.According to the new KM,however, this perspective hasthings exactly backwards.

Strategy is a type ofknowledge and is itself anoutcome of knowledgeprocessing (see figure 1). If the purpose of KM is toenhance knowledge processing, then KMprecedes strategy and every other knowledgeoutcome. To argue the reverse is to grantstrategy an exception that it does notdeserve. Why should strategy be any lesssubject to knowledge-production processesthan other knowledge outcomes? We call the idea that strategy comes first the‘strategy exception error’. If KM is to have a future, we must eliminate this error and recognise strategy as just another set of knowledge claims that flow out ofknowledge processing.

The most important form of strategyaddresses an organisation’s capacity tolearn and adapt. Strategies come and go,but in order to survive over the long haul, thequality of an organisation’s systemic capacityto learn and adapt must be high and sustain-able. This is ‘sustainable innovation’, thefundamental strategy of every organisationwishing to survive and prosper.

Most of what currently passes for KM initiatives aligned with strategy are really IM (information-management) projects. Their focus is on capture and delivery ofinformation required to support strategy.While valid and useful, these are not KM initiatives per se. The purpose of KM is to enhance knowledge processing, which, in turn, enhances an organisation’s capacity to produce strategies. It is IM thatafterwards must be aligned with and supportstrategy, not KM.

In our vision of the future of KM, therewill be two kinds of strategy: knowledge-processing strategy and business-processingstrategy. Today, only the latter receives

attention and KM is mistakenly seen as itsservant. Once firms rise to the challenge ofenhancing their capacity to learn and adapt,the relevance and importance of KP willcome into full view. Only then will KM berecognised as the precursor to strategy thatit is, and not its dependent slave.

Transcending the Nonaka and Takeuchi SECI modelAmong the many important implications ofthe new KM is that the still popular SECI(socialisation/externalisation/combination/internalisation) knowledge-conversion modelput forth by Nonaka and Takeuchi in TheKnowledge Creating Company2 suffers fromat least two important limitations. First, SECIhas flaws in its psychological and cognitivetheory. It neglects to include consideration ofimplicit knowledge and in the processprovides us with an ambiguous account oftacit knowledge. As Polanyi points out, tacitknowledge consists of that which one canknow but never tell. It is inexpressible.Implicit knowledge, as Polanyi also says,however, can be converted to explicit form.But Nonaka and Takeuchi do not defineimplicit knowledge, even though theirdiscussion refers to similar beliefs as onetype of tacit knowledge. Had they distin-guished implicit knowledge, SECI would havedefined more modes of conversion. Thissuggests that SECI model is incomplete.

SECI also fails to distinguish betweenknowledge predispositions and situationalorientations. The distinction between tacitand explicit knowledge may either be inter-preted as applying to predispositions or to

orientations, or both. If it’s applied to predispositions, it has no meaning becausethere are no explicit predispositions. On the other hand, if tacit knowledge isapplied to orientations, then it is clear thatmuch of the tacit knowledge referenced in examples, such as the ability to ride abicycle, doesn’t fit a situational interpreta-tion of tacit knowledge, because abilitiesare predispositions.

Second, the new KM distinguishesbetween subjective knowledge in minds andobjective knowledge in artefacts. This, too,materially expands the range of possibilitiesfrom which knowledge can be converted.Instead of just tacit, implicit and explicitbeliefs, we now have all three and bothsubjective and objective forms ofknowledge to consider. Therefore we haveup to five conversion possibilities to dealwith in considering a new model, not justtwo. This takes us from a two-by-two matrixwith four cells, to a five-by-five matrix with25 possible conversions, though closeranalysis shows only 17 viable ones. Thismeans that all those KM practices andpractitioners who have based theirknowledge conversion efforts on SECI havebeen overlooking 13 modes of conversion.Also, there can be no conversion of trulytacit knowledge or predispositions toexplicit knowledge. At best, those applyingSECI in this regard have been ‘capturing’implicit knowledge, not tacit knowledge.

For these and other reasons we couldnot cover here, SECI must be reformulatedwith a broader, more complete conversionmodel. One of the concerns of the new KM in

Business outcomes

Business processing

Knlg. processing outcomes

Knowledge processing (KP)

Knowledge management (KM)

For example:• Business Strategies• Organizational Models• Business Processes• Product Strategies

For example:• Business strategies• Organisational models• Business processes• Product strategies

Knowledgemanagement

(Meta-epistemicbehaviours) KM outcomes

Knowledgeprocessing

(Epistemicbehaviours)

Businessprocessing

(Operationalbehaviours)

For example:• Profitability• Market Share• Growth• Sustainability

For example:• Profitability• Market share• Growth• Sustainability

For example:• Knowledge Processing

(KP) Strategies• KP Policies and Rules• KP Infrastructures• Learning Programs

For example:• Knowledge processing

(KP) strategies• KP policies and rules• KP infrastructures• Learning programmes

Figure 1 – the three-tier new KM model

June 200314


the short term will be to produce this modeland bring it into common practice.

Developing the enterprise knowledge portalThough vendors and writers on the subjectclaim that enterprise knowledge portals(EKP) already exist, if by EKP we mean asoftware application that comprehensivelysupports knowledge processing and KM,then today’s applications fall far short ofthis goal.3 The EKP is an application on theverge of development. The technology itrequires is in existence now. The cost of itsdevelopment is low as software applicationsgo, since its implementation is largely amatter of systems integration, with theexception of its intelligent-agent component,which must be developed.

On the other hand, the benefitsassociated with the EKP are great. Theyamount to nothing less than the realisationof the promise of the enterprise informationportal (EIP) to achieve increased ROI, com-petitive advantage, increased effectivenessand accelerated innovation. EIPs are riskybecause, generally, they fail to evaluate the information they produce and deliver for quality and validity. Nothing, includingEKPs, can ensure certainty about informa-tion, models or knowledge claims, but EKP applications incorporate a systematicapproach to knowledge-claim testing andevaluation that eliminates errors andproduces quality-assured information. In the category of portal technology, they, not

EIPs, are the best we can do. They, notEIPs, are necessary for supporting the openenterprise (see below). They, not EIPs, arethe future of portal technology.

Developing a comprehensive system of KM metricsKM needs improvement in metrics development. KM metrics have focused onintangible assets and sometimes on impactanalysis, but a comprehensive approach islacking, because a framework underlyingsuch an approach has not so far becomeavailable. A systematic metrics frameworkis needed to continue to make progress inKM. The present ad hoc approach is too

slow and uncertain, since the absence of ametrics framework also hampers evaluationof the validity of proposed metrics. Ameasure labelled a ‘KM metric’ is not nec-essarily what it claims to be. To evaluatewhether it measures what it purports tomeasure, we need a broader theoreticalcontext in which the new metric, as well as

other metrics to which itrelates, are incorporated.

The three-tier model infigure 1 provides aframework for classifyingmetrics as well as forviewing the scope of KMmore generally. Figure 1 iseasily specified to provide amore granular map of KMmetrics. At the levels ofbusiness processing andbusiness outcomes, avariety of metrics developedfor use with balanced-scorecard, quality-basedapproaches, and sales andmarketing measurementsystems may be used toproduce detailed sub-cate-gorisations for the map. Atthe level of knowledgeprocessing and knowledgeoutcomes, sub-categoriesare provided by theknowledge-lifecycle (KLC)framework (see figure 2).The KLC sub-processcategories include informa-

tion acquisition, individual and grouplearning, knowledge-claim formulation,knowledge-claim evaluation (KCE),knowledge and information broadcasting,search and retrieval, teaching, and sharing.The KLC outcome categories include organi-sational knowledge and the distributedorganisational knowledge base (DOKB).

Sub-categories at the KM process levelinclude symbolic representing, buildingexternal relationships with others practisingKM, leading, KM-level knowledge production,KM-level knowledge integration, crisishandling, changing KP rules, negotiating forresources with representatives of otherorganisational processes, and allocating

resources for knowledge processes and forother KM processes. Outcomes at this levelinclude knowledge processing outcomes andsocio-technical outcomes.

Developing the open enterpriseBecause most contemporary approaches toKM fail to make the all-important distinctions

Experiential feedback

= Knowledge processes

= Knowledge sets

CKC = Codified knowledge claimCOK = Codified organisational knowledge DOKB = Distributed organisational knowledge baseFKC = Falsified knowledge claimOK = Organisational knowledgeSKC = Surviving knowledge claimUKC = Undecided knowledge claim

Individualand grouplearning

Knowledgeclaimformulation

Informationacquisition

CKCKnowledgeclaimevaluation

Broad-casting

Searching

Teaching

Sharing

InfoaboutSKC

SKC

InfoaboutFKC

FKC

InfoaboutUKC

UKC

OK

DOKBObjective knowledgeSubjective knowledge

DOKB Containers! Agents (Indiv. & groups)! Artefacts (Docs, IT, etc)

Business processbehaviours

of interacting agents

External inputs

Knowledge integration

Business processing environmentFeedback

(including thedetection of

epistemic problems)

Knowledge production

“

“

A systematic metrics framework isneeded to continue to make progress inknowledge management. The present adhoc approach is too slow and uncertain.

Figure 2 – the knowledge lifecycle

June 2003 15


among KM, KP and BP, they usually do notprovide us with any visions of how knowledgeprocessing might be improved as a conse-quence of KM strategies and interventions.Instead, they tend to focus on streamliningindividual processes of information retrievaland use, but not so much on learning orknowledge production. We believe that thegoal of KM should be achieving and main-taining sustainable innovation in knowledgeprocessing, and that to accomplish thisorganisations need openness in knowledgeprocessing, including openness in all sub-processes of the KLC. We call the resultingnormative model, or target knowledge-processing environment, the ‘openenterprise’ (OE).

From a new-KM perspective, the mosteffective knowledge-processing environ-ment for learning and innovation is one inwhich problems are openly recognised,knowledge claims are openly formulated,tested and evaluated on a continuing basisby all stakeholders, and transparency, trust,inclusiveness and other correlates andoutcomes of openness prevail. In the OEimage of the future, ideas, strategies,processes and plans in business are validonly if they survive our tests and evalua-tions, and not simply because of theirsource. We disagree with the Nonaka andTakeuchi position, for example that the justification criteria for knowledge in organi-sations should be set by top management,as though truth is simply a function of whatmanagement happens to think or say.Rather, we believe that truth is independentof rank or title in organisations, and thatthe advice offered by Nonaka and Takeuchiand others, that justification criteria shouldflow from the top, is a recipe for moreEnrons and Global Crossings.

This is not to say that managementshould be democratic in the OE. Not at all.We make a sharp distinction between operational decision making (the province ofmanagement) and knowledge making (theprovince of us all). We envision organisationswhere managers continue to wieldcommand-and-control authority incommitting resources of the firm to action,even as their ideas and those of others aresubjected to open testing, evaluation andcriticism. In the OE, however, knowledgeprocessing will be a transparent andinclusive affair. The protection of opennesswill rise to the level of fiduciary duty, whichwill rest with the board. Indeed, in the OE,KM will very likely report to the board ofdirectors, a position that fits the criticality ofknowledge processing in most organisations.

Creating communities of inquiryCommunities of practice (Cop) frequentlyfacilitate knowledge sharing. Practitioners in

the Cop area often believe that they areuseful for knowledge production, too. Butnew-KM perspectives raise the concern thatknowledge production in Cops is likely to becharacterised by the use of consensus as acriterion for validating knowledge claims.That is, knowledge outcomes, as opposed to information outcomes, may in certainCops be determined by community opinionabout which knowledge claims are moststrongly supported by evidence or otherevaluation criteria.

This communitarian form of knowledgeproduction is inconsistent with new KM’scommitments to anti-justificationism andfalliblism, the ideas that no knowledgeclaim can be justified, only criticised, andthat no knowledge claim is certain. Inaddition, communitarianism is alsoopposed to the new KM’s fundamental ideathat knowledge grows by eliminating errorsin knowledge claims through testing andevaluation, and that testing and evaluationinvolves applying multiple criteria. In thefuture, KM will need to develop an alterna-tive to the communitarian Cop construct,specifying the attributes and characteristicsof communities dedicated to knowledgeproduction and the discovery and elimina-tion of errors in knowledge claims. Suchcommunities are better called communitiesof inquiry rather than Cops. They are thecounterparts of the OE at the group orcommunity level.

Developing value theory in KMProblem recognition, as well as every sub-process of the KLC, involves making valuejudgements. However, the sub-processwhere value judgements seem most con-troversial is the KCE sub-process, the keyto distinguishing knowledge from informa-tion. In developing our model for KCE, weincluded a category of criteria calledpragmatic priority.1 All our other criteria fallinto the category of traditional epistemiccriteria for comparatively evaluating factualknowledge claims. But pragmatic prioritytakes account of the valuational conse-quences of rejecting knowledge claims asfalse, and relying on surviving knowledgeclaims as a basis for action.

The risks we take are a combination ofthe likelihood that our evaluations rejectingparticular knowledge-claim networks are inerror, and the benefits and costs associatedwith such errors. If we are in error we mustsuffer the cost and benefit consequencespredicted by the true knowledge-claimnetwork we have rejected. To take account of these risks in estimating pragmaticpriority, we must formulate knowledge claimsthat provide a value interpretation of ourdescriptive knowledge-claim networks. So to estimate pragmatic priority we have nochoice but to formulate a value theory, andto use it in making our estimates and incomparatively evaluating factual knowledgeclaims. The implication of the role of valuetheory in KCE just outlined is that objectiveinquiry, and our views about truth and falsity,require formulating and also testing andevaluating value theory along with ourtheories about fact.

Transcending KM standards developmentOne of the more profound ironies in con-temporary KM can be found in the area ofstandards. There are many KM standardsinitiatives under way around the world, allof which seem to suffer from the sameinternal problem. Even as they purport tobe aiming at the development of standardsfor enhancing knowledge processing inorganisations, they show no compunctionwhatsoever, much less acknowledgement,about adopting the existing standards of the very organisations under whoseauspices their efforts are unfolding. Thus,if the knowledge-processing systems andpractices of, say, the American NationalStandards Institute (ANSI) are acceptablefor purposes of developing standards forKM – a form of knowledge – why not justadopt the ANSI’s approach to knowledgemaking and call it a day?

Here, again, we can see a basic failureto make the fundamental distinctionbetween KM and knowledge processing.For if this distinction were widely made, itwould be clear to all that standards-makingorganisations have, first, already specifieda standard approach for knowledge making

“ “If the processes defined and managedby standards-making organisations arenot viewed as acceptable bases for KMstandards, how can standards developedunder their auspices be viewed asdefensible by the KM community?

June 200316


and, second, are themselves engaging inKM. Therefore, their approaches to bothrealms of activity, if considered acceptableto those who would develop standards foreither KM or knowledge processing or both,ought to be considered themselves as KMstandards as such. After all, what is astandards-making organisation and itsprocedures if not a KM-enabled knowledge-processing system?

And if the processes defined andmanaged by such standards-making organisations are not viewed as acceptablebases for KM standards, then how can KMstandards developed under their auspices be viewed as defensible by the KMcommunity? Wouldn’t the KM standardsdeveloped via an unacceptable KP systemthat is managed by an organisation whoseKM practices are seen as being at odds withhow KM should be done be suspect, eveninvalid, themselves?

The irony and contradictions here areobvious. Our response to the questionsraised above is that most of the current

standards-making efforts in KM are, for thereasons we suggest, confused, and thatthe processes enforced by the standardsorganisations involved are epistemologi-cally biased in the wrong direction. At base,they are communitarian or consensus-based systems, according to which a claimpasses for knowledge through a kind ofpopularity contest. Never mind what thebest outcome might be; if an idea managesto capture the majority vote, it wins.

KM standards adopted under such aprocedure don’t deserve our respect. They unwittingly condone an approach todetermining truth based on an appeal toauthority: in this case, an appeal to theauthority of the majority. If KM is to have afuture worthy of our respect, it must riseabove these deficiencies in conventionalKP. Before we can have standards in KM,we must have standards in knowledgeprocessing. Standards-making efforts ofthat kind, however, are nowhere to befound. In our vision of the new KM, theywill be everywhere.

The new knowledge managementOur purpose in this article has been toprovide some of the flavour of the new KMby discussing eight issues that reflect itsview of the future of KM and some of itsconceptual perspectives. But the ideasdiscussed here are not all there is to the newKM perspective. Nor do they exhaust theissues that may be important in the future.

Our discussion of the SECI model beginsto hint at the unified theory of knowledgeand its basic distinction between subjectiveand objective knowledge. This theory extendsto a new approach to error elimination invalue theory. In addition, the new KMprovides a broader organisational learningperspective than other frameworks, synthe-sising it with evolutionary epistemology,complex-adaptive-social-systems theory,motivational psychology and culturalanalysis. In this way it foreshadows our beliefthat the future of KM will see the develop-ment of general theory synthesising theframeworks of the many disciplines thatcontribute to knowledge management. !

RReeffeerreenncceess1. Firestone, J.M. & McElroy, M., Key Issues in the NewKnowledge Management (KMCI Press/Butterworth-Heinemann, 2003)2. Nonaka, I. & Takeuchi, H., The Knowledge CreatingCompany (OUP, 1995)3. Firestone, J.M., Enterprise Information Portals andKnowledge Management (KMCI Press/Butterworth-Heinemann, 2003)

Taxonomies: Frameworks for Corporate KnowledgeThe Shape of Things to Come

Written by expert content analyst Jan Wyllie, in association with David Skyrme Associates and Knowledge Management magazine, this report outlines:

! What taxonomies are;! Why they are important;! How to implement a taxonomy;! Individual company case studies;! What the experts say;! Scenarios for the future.

Order your copy now for ONLY £195.Call Kevin Hallett on +44 (0)20 8785 5900 to reserve your copy.

“

“

Before we can have standards in KM, we must have standards in knowledge processing. Standards-making efforts of thatkind, however, are nowhere to be found.

Documents

New Knowledge Management