complex knowledge

Complex Knowledge

Dave Snowden

Cynefin Centre for Organisational Complexity

Global Services© IBM UK Ltd 2003 COMMERCIAL IN CONFIDENCE

The three rules of Knowledge Management

Knowledge can only be volunteered, it cannot be conscripted• Camouflage – motivated by fear of abuse

• Conformity – motivated by time I only know what I know when I need to know it

• Human knowledge requires contextual stimulation

• The way that we know things is not the way that we say we know them, and that goes for decision making too

We always know more than we can say, and we will always say more than we can write down• Context Management

• Narrative Management

• Content Management


Knowledge is the means by which we inform, not a higher order of information

Data

Information

Knowledge

Wisdom


Knowledge is the means by which we inform, not a higher order of information

Dat

a

Information

Knowledge

Wisdom

Context


The model that launched a thousand failed KM initiatives: Nonaka’s SECI

Originally a paper on knowledge transfer in the consumer goods manufacturing industry in Japan

Rapidly generalised as a model for knowledge transfer in all sectors

Closely linked with Business Process Re-engineering

Used way beyond Nonaka’s original intent

Tacit Explicit

Exp

lici

tT

acit

Socialisation Externalisation

CombinationInternalisation

S E

CI


ASHEN Modela meaningful question

ArtefactsAny thing made by people, processes, documents, tools in which knowledge is imbedded

SkillsAbilities that can be trained and measured without ambiguity, but remember the time issue

HeuristicsRules of thumb, the outcome of experience, the main repository of knowledge mostly unarticulated

ExperienceAccumulated experience of failure and success which allows the right pattern to be triggered in the right context

Natural TalentSome people are just better at doing things than other people – and they are often not the people you expect


Knowledge Disclosure Pointsa meaningful context

Any event or activity which reveals knowledge through use• Decisions

• Problems resolution

• Solution creation

• Judgement

• Learning points Used to create the context in which the ASHEN question can be

asked Can be achieved through virtual and physical means

• Virtual allows anonymity and multi-persona techniques to be used

• Physical provides a different dynamic of interaction to vitual (not better or worse but virtual)

DO NOT judge or value or move to solution

First Exercise

Knowledge Disclosure Points


Discovering KDP & ASHEN

Anecdote Circles• A group with cohesive experience• May use a structured task (e.g.

future history)• Stimulate Ditting• Use of Actors, Cartoonists etc • Use of Proxies

Mass Capture Techniques• Naïve Interviews• The Set up

Participative techniques• Field Interviews• Participative Observation• Deep Emersion


From KDP to Knowledge Objects

HAA S

KDP CLUSTERS

ASK ASHEN EAA S

N

HAA S

ASHEN CLUSTERS

KN

OW

LE

DG

E

OB

JEC

TS

Communities and JIT Knowledge management

Using the natural flow patterns of an organisation


Acceptable levels of Abstraction

Ab

str

acti

on

Cost of Codification

ULAA

LLAA


The Cynefin sense making framework applied to knowledge management

High Abstraction

Low Abstraction

Teach

ing

Cu

lture

sLearn

ing

Cu

ltu

res

Communities of Practice

The Domain 150

Committees & Review BodiesBureaucracy

The Shadow OrganisationDomain of

15

Chaos the threat of the

newCrisis

management & opportunity


The Cynefin sense making framework applied to knowledge management

Good Practice

Best Practice

Emergent Practice

Decisive Action

High Abstraction

Low Abstraction

Teach

ing

Cu

lture

sLearn

ing

Cu

ltu

res


Communitiesan emergent model


Social Network Analysis


Knowledge Project MatrixK

NO

WL

ED

GE

OB

JEC

TS

!!!!

!!!!!!

!!

!!!!!!

5 2 4 1 3 7 6

THE THINGS DECISION MAKERS VALUE


Knowledge Project Matrixeasy to sell, hard to implement

KN

OW

LE

DG

E O

BJE

CT

S

!!!!

!!!!!!

!!

!!!!!!

5 2 4 1 3 7 6



Knowledge Project Matrixhard to sell, easy to implement

KN

OW

LE

DG

E O

BJE

CT

S

!!!!

!!!!!!

!!

!!!!!!

5 2 4 1 3 7 6



Knowledge Project Matrixa portfolio of knowledge projects

KN

OW

LE

DG

E O

BJE

CT

S

!!!!

!!!!!!

!!

!!!!!!

5 2 4 1 3 7 6



Creating a Knowledge Strategy“what” related to value, where and flow

KN

OW

LE

DG

E O

BJE

CT

S

!!!!

!!!!!!

!!

!!!!!!

5 2 4 1 3 7 6



The history and context of knowledge management

The first generation before the SECI model• Information Management

• Technology supports human decision making The second generation the SECI period (corruption of Nonaka)

• Knowledge is tacit or explicit

• KM practice focuses on the conversion of tacit to explicit knowledge

• Closely linked to Business Process Re-engineering (BPR)

• Focus on maximising efficiency The third generation post SECI

the impact of complexity science• Knowledge is never wholly tacit or explicit

• Just in Time Knowledge Management

• A focus on effectiveness rather than efficiency

To improve the effectiveness (not the efficiency) of decision

making and to create the conditions for innovation

The purpose of Knowledge Management


3rd Generation or Post SECI KM

Content Management Narrative Management

• Narrative representation, neither qualitative, quantitative but emergent• Archetypes & themes & & & &

• Narrative databases• Indexed on emergent, NPML and Demographic data• Enable Serendipitous encounter

• Myth Management• Fable Construction• Story Virus• Monitoring

Context Management• Social network stimulation• Expertise location• Narrative (metaphor)• JIT KM

Technology is a tool: if we pick it up and it fits our hand it is useful, if we have to bio-reengineer our hand to fit the tool something is

going badly wrong!


Knowledge Managementthe final metaphor

How do you learn to ride a bicycle?


The Cynefin Centre

Nasrudin found a weary falconsitting one day on his window-sill. He had never seen a bird like this before.‘You poor thing’, he said, ‘how ever were you to allowed to get into this state?’He clipped the falcon’s talons and cut its beak straight, and trimmed its feathers.‘Now you look more like a bird,’ said Nasrudin.

From Idries Shah, The exploits of the Incomparable Mulla Nasrudin & The subtleties of the Inimitable Mulla Nasrudin. Octagon Press, London 1985

The Cynefin Centre

A new paradigm in research and method development


The presumptions of management science and consultant practice

The presumption of rational choice• human actors make a “rational” decision based on self interest• individual and collective behavior can be managed by manipulation of

pain/pleasure• education to make those consequences evident will change behavior

The presumption of intentional capability• the acquisition of capability indicates an intention to use that capability• actions from competitors are the result of intentional behaviour: every “blink” we

see is in effect a “wink” and act accordingly. • hypocrisy of intent, we have accidents, they don’t

The presumption of unitary and stable identity• organisations are aggregations of discrete individuals• categorisation and stereotyping• scalability – the rule of 15 & 150

The presumption of order• relationships between cause and effect capable of discovery and empirical

verification and in consequence, prescriptive and predictive models are possible


Contextual Complexitythe landscape of management

Empirically KnownThe domain of

the actualCause and effect relationsCause and effect relations

repeating & predictablerepeating & predictable

Empirically KnowableThe domain ofthe probable

Cause and effect separatedCause and effect separatedover time & space but repeatover time & space but repeat

ComplexThe domain of many

possibilitiesCause and effect coherent in Cause and effect coherent in

retrospect, repeat accidentallyretrospect, repeat accidentally

ChaosThe domain of the

inconceivableNo cause and effect relationships No cause and effect relationships

generally perceivablegenerally perceivable


Contextual Complexitydecision models

The domain ofthe actual

SenseCategoriseRespond

The domain ofthe probable

SenseSenseAnalyseAnalyseRespondRespond

The domain of many possibilities

ProbeProbeSenseSense

RespondRespond

The domain of the inconceivable

ActSense

Respond


Contextual ComplexityIntervention types

Empirically KnownThe domain of

the actualLegitimate best practice

STANDARD PROCEEDURESPROCESS RE-ENGINEERING

Empirically KnowableThe domain ofthe probable

Analytical/ReductionistSCENARIO PLANNINGSYSTEMS THINKING

ComplexThe domain of many

possibilitiesPattern Management

PERSPECTIVE FILTERSCOMPLEX ADAPTIVE SYSTEMS

ChaosThe domain of the

inconceivableStability focused intervention

ENACTMENT TOOLSCRISIS MANAGEMENT


Interpretation of state determined by preference for action

Known

Knowable

Complex

Chaos

DISORDER


Crisis managementmulti-point attractors

Asymmetric Collapseexcessive order leads to the collapse of order – chaos

Single point attractorsthe charismatic dictator stabilise to order, with the danger of asymmetric collapse

Multi point attractorscreate several points of attraction (chaos to complexity) to seed possibilities

Select favourable patterns to exploit (Complex to Knowable)

Kill unfavourable patterns (but you don’t know which they will be

X

X


Discovery and InterventionNetwork linkages

Strong centralStrong central

strong strong distributeddistributed

Strong centralStrong central

weak weak distributeddistributed

WeakWeak

centralcentral

strong strong distributeddistributed

WeakWeak

CentralCentral

weak weak distributeddistributed

Domain of the Actual

Domain of the ProbableDomain of Possibilities

Inconceivable Domain


Diagnosis - principles

Domain of the ActualWe can teach people how to do We can teach people how to do things very quickly and ensure things very quickly and ensure

compliancecomplianceWe control the space and can

determine behaviourAny diagnostic is a simple check

against predetermined criteria

Domain of the ProbableWe have learnt how to solve this We have learnt how to solve this problem and are now in a stable problem and are now in a stable

situationsituationWe have been here before and any deviations are readily understood

Diagnostics can be separatedfrom intervention

Domain of PossibilitiesEvery time we think we have Every time we think we have

solved the problem it returns in a solved the problem it returns in a different formdifferent form

Minor, troublesome events occur but are easily rationalised Diagnostics are interventions –

they influence the patterns

Inconceivable DomainThis is a new situation in which This is a new situation in which we no previous experience and we no previous experience and

have no reflective timehave no reflective timeSomething completely unexpected at this time

There is no time or need for any diagnostic


Intervention

Domain of the ActualStandard process with reviewStandard process with review

cycle & clear measurescycle & clear measures

Establish indicators to prevent Establish indicators to prevent catastrophic failurecatastrophic failure

SENSE – CATEGORISE - RESPONDSENSE – CATEGORISE - RESPOND

Domain of the ProbableAnalytical techniques to determine Analytical techniques to determine

facts and option rangefacts and option range

Create ritual/cyclical process to Create ritual/cyclical process to prevent entrainment of expertiseprevent entrainment of expertise

SENSE – ANALYSE - RESPONDSENSE – ANALYSE - RESPOND

Domain of PossibilitiesMultiple small and diverse Multiple small and diverse

interventions to create optionsinterventions to create options

Always have an exit strategy to the Always have an exit strategy to the ordered domains for exploitationordered domains for exploitation

PROBE – SENSE - PROBE – SENSE - RESPONDRESPOND

Inconceivable DomainSingle or multi point attractor(s) Single or multi point attractor(s)

to stabilise situationto stabilise situation

Avoid creating long term Avoid creating long term dependency on single attractordependency on single attractor

ACT – SENSE - RESPONDACT – SENSE - RESPOND


The Cynefin Life Cycle

Licensed Methods

Comprise Components Models and ToolsWhich have Attributes and UsesAnd combine into Assemblies and RecipesInterface with Other bounded

techniques

Emergent Methods

Are Experimental Innovative and ChallengingThey are Sold at cost Using key staff and Aim to become

LicensedThey can Use the elements of

licensed method Create new elements

Discovery Programs

Address Intractable Problems Re-vitalising other

approachesComprise Catalytic and

Patterning Events Focused

Participatory Learning Streams

Emergent ResearchTypically Run over 3-6 months Are Multi-client

complex knowledge

Documents

expectglobal services

knowledge transfer

ithuman knowledge

main repository of knowledge

contextual stimulationthe

nonakas secioriginally

outcome of experience

ashen question