Slidedeck for Integrated-EA conference, February 2014.
(It's a conference on enterprise-architecture in the Defence context, hence a somewhat military flavour and various military in-jokes.)
The dung-beetle's tale: systems-thinking, complexity and the real-world
1. the futures of business
The dung-beetleās tale
systems-thinking, complexity and the real world
Tom Graves, Tetradian Consulting
Integrated EA Conference, London, March 2014
21. It seems that every attempt
to control the complexity
makes it more complex.
22. It may seem like every attempt
to control the complexity
makes it more complex.
23. Beyond a certain point,
it may seem like every attempt
to control the complexity
makes it more complex.
24. Beyond a not-so-certain point,
it may seem like every attempt
to control the complexity
makes it more complex.
25. Subject to certain provisos,
beyond a not-so-certain point
it may seem like every attempt
to control the complexity
makes it more complex.
26. Subject to certain provisos
and special-cases,
beyond a not-so-certain point
it may seem like every attempt
to control the complexity
makes it more complex.
27. Sometimes,
subject to certain provisos
and special-cases,
beyond a not-so-certain point
it may seem like every attempt
to control the complexity
makes it more complex.
28. Sometimes,
subject to certain provisos
and special-cases,
and in unpredictable ways,
beyond a not-so-certain point
it may seem like every attempt
to control the complexity
makes it more complex.
65. Overall, itās another pattern:
ā¢ sense
ā¢ make-sense
ā¢ decide
ā¢ act
(rinse-and-repeat, indefinitely,
at every required level)
(yep ā another kind of fractal-recursionā¦)
66. When its Simple doesnāt work
for some level or contextā¦
our beetle has to stop
and think for a bit.
Itās kinda Complicatedā¦
(for a beetle, anywayā¦)
67. On the mapā¦
before
(this axis represents ātime available
for sensemaking and decision-makingā
before we must take action)
āSimpleā
NOW!
certain
uncertain
68. On the mapā¦
before
weād place
āComplicatedā
in this region of
our map
(because itās distant
in time from the
āNOW!ā of action)
(this axis represents ātime available
for sensemaking and decision-makingā
before we must take action)
NOW!
certain
uncertain
69. On the mapā¦
before
āComplicatedā
(this axis represents ātime available
for sensemaking and decision-makingā
before we must take action)
(for āSimpleā, time-tostop-and-think is a
luxury we donāt haveā¦)
NOW!
certain
uncertain
70. Our beetle formulates a planā¦
(See beetle. See beetle stop and thinkā¦)
CC-BY-NC-ND scotproof via Flickr
75. Theory and practice
before
THEORY
What we plan to do, in the expected conditions
What we actually do, in the actual conditions
PRACTICE
NOW!
certain
uncertain
76. Theory and practice
before
āIn THEORY,
thereās no difference
between theory
and practiceā¦
ā¦in PRACTICE,
there is!ā
NOW!
certain
uncertain
114. Weāve noted its vertical-axisā¦
before
(represents ātime available for
sensemaking and decision-makingā
before we must take action)
NOW!
certain
uncertain
115. ā¦letās look at its horizontal-axis
before
(represents an arbitrary spectrum
from āabsolute-samenessā [left-side]
to āabsolute-differenceā [right-side])
NOW!
certain
uncertain
116. A bunch of similar scalesā¦
sameness
uniqueness
high-probability
standardised
low-probability
customised
high-dependability
reusability
low rate of change
unique
low-dependability
bespoke
high rate of change
117. What, no numbers?
before
by intent, there are no numbers
on either axisā¦
itās to map criteria for tactics
ā relationships, not āabsolutesā
NOW!
certain
uncertain
118. We donāt need no steenkinā
numbers!
(but we do need options for varying tacticsā¦)
CC-BY-NC-SA isaachsieh via Flickr
121. Take control! Impose order!
āInsanity
is doing
the same thing
and expecting
different resultsā
(Albert Einstein)
ORDER
(IT-type rules do work here)
certain
uncertain
122. Order and unorder
āInsanity
is doing
the same thing
and expecting
different resultsā
edge of
uncertainty
āInsanity
is doing
the same thing
and expecting
the same resultsā
(Albert Einstein)
(not Albert Einstein)
ORDER
UNORDER
(IT-type rules do work here)
(IT-type rules donāt work here)
certain
uncertain
123. Same and different
A quest for certainty:
analysis, algorithms,
identicality, efficiency,
business-rule engines,
executable models,
Six Sigma...
An acceptance of
uncertainty:
experiment, patterns,
probabilities, ādesignthinkingā, unstructured
process...
SAMENESS
UNIQUENESS
(IT-systems do work
well here)
(IT-systems donāt work
well here)
certain
uncertain
124. Why skills are neededā¦
What is always going to be
uncertain or unique?
What will always be āmessyā?
(āMessyā ā politics, management, wickedproblems, āshouldā vs āisā, etc.)
Wherever these occur,
weāre going to need human skillā¦
127. Tame- and wicked-problems
ā¢ definable formulation
ā¢ static āsolutionā
ā¢ clear end-point
ā¢ solution is true/false
ā¢ each essentially same
ā¢ finite dependency
ā¢ no defined formulation
ā¢ dynamic āre-solutionā
ā¢ no clear end-point
ā¢ solution is good/bad
ā¢ essentially unique
ā¢ infinite dependency
āTAMEā
āWICKEDā
(ācontrolā can work here)
(ācontrolā canāt work here)
certain
uncertain
128. Terms such as
ācomplexity scienceā
may unintentionally mislead:
physical-sciences apply
mostly in the ātameā-spaceā¦
most ācomplexity-scienceā
applies in the āwickedā-space.
129. Use the right models
for each domain...
donāt mix them up!
134. Where to map complexity?
āComplicatedā
āCOMPLEXā
some people might place
Complexity in just one domain
āSimpleā
āChaoticā
135. Where to map complexity?
before
āComplicatedā
āAmbiguousā
yet in reality,
Complexity is everywhere!
āSimpleā
āNot-knownā
NOW!
certain
uncertain
136. Many meanings of
āComplexityāā¦
(and of āChaosā or āChaoticā, too)
ā we need to embrace them all
(not make the Simplistic assertion
that only one kind is āthe real complexityāā¦)
137. Complexity: theyāre both rightā¦
Roger Sessions:
John Seddon:
āeliminate
complexity!ā
āembrace
complexity!ā
(Simple Iterative
Partitions; Snowman)
(Vanguard Method;
āfailure-demandā)
SAMENESS
UNIQUENESS
(most IT-type models
do work well for this)
(most IT-type models
donāt work well for this)
certain
uncertain
138. Theyāre both right, because
of fractal recursionā¦
- a point we can illustrate via Mandelbrotā¦
139. not this kind of Mandel Brotā¦
CC-BY-NC squeakychu via Flickr
140. this kind of Mandelbrotā¦
CC-BY-NC-SA Michael-Maclean via Flickr
146. every point expresses the patternā¦
every point describes every other pointā¦
CC-BY-NC-SA sharman via Flickr
147. Fractal recursion means that
every point includes its own:
ā¢ Simple
ā¢ Complicated
ā¢ Ambiguous (āComplexā)
ā¢ Not-known (āChaoticā)
148. NOTE:
āself-similarā is not the same as
āthe sameāā¦
āhigh-probabilityā does not mean
āwill always happenāā¦
ālow-probabilityā does not mean
āwill never happenāā¦
149. Use the right tactics
for each domain...
donāt mix them up!
150. But the beetle saysā¦
āitās too easy to hide in theory hereā¦ā
CC-BY-NC-ND scotproof via Flickr
151. At some point
we need to stop theorising
(or hypothesising ā to be pedantic)
and get back in touch with
the real-world, in real-timeā¦
- down into the āNot-knownāā¦
153. Back to the real-world againā¦
before
āComplicatedā
(exploration, improvisation and
invention all happen here)
āSimpleā
āAmbiguousā
edge of
innovation
āNot-knownā
NOW!
certain
uncertain
154. Here, even a dung-beetleā¦
CC-BY-NC-ND kishlc via Flickr
155. can learn how to fly!
CC-BY-NC-ND kishlc via Flickr
165. Where do new
ideas come from?
āAccept the burden
of uncertaintyā¦
be comfortable
with being
uncomfortable.ā
CC-BY-NC-SA grrrl via Flickr
166. Where do new
ideas come from?
āI can tell you things
all day long, but you
have to put your ass
in the seat to really
learn.ā
CC-BY-NC USArmyEurope via Flickr
167. We also come at this
from the other direction
- over from the Simple,
in real-time action,
across the edge of panicā¦
168. Across the edge of panic
before
āSimpleā
(ENACT)
edge of
panic
āNot-knownā
(EXPLORE)
NOW!
certain
uncertain
176. The other learning-loopā¦
before
(we need real-time
feedback loops
between certainty and
uncertainty)
fears
āSimpleā
(ACTION)
edge of
panic
āNot-knownā
(EXPLORE)
options
NOW!
certain
uncertain
182. Making sense of complexity
Use context-maps such as SCAN
to identify
what may or must change
what is or is not certain
how these vary over time
and what to do with each
183. Linking it all together
before
questions
āComplicatedā
edge of
uncertainty
(ANALYSE)
āAmbiguousā
(EXPERIMENT)
answers
rules
edge of
action
āSimpleā
(ENACT)
realitie
s
principles
edge of
innovation
news
fears
edge of
panic
āNot-knownā
(EXPLORE)
options
NOW!
certain
uncertain
184. Common themes in each domain
before
order
unorder
COMPLICATED
AMBIGUOUS
(analyse)
(experiment)
reciprocation
resonance
(algorithms)
recursion
reflexion
(patterns, guidelines)
SIMPLE
plan
actual
NOT-KNOWN
(enact)
(explore)
regulation
rotation
(rules)
reframe
rich-randomness
(principles)
NOW!
certain
uncertain
185. Remember that itās recursive
before
C
C
C
NOW!
S
certain
A
S
N
S
A
A
N
N
uncertain
186. A surgical exampleā¦
before
patient identity
patient
condition
theatre
booking
equipment
plan
verify identity
surgery plan
surgical-staff
availability
consumables
action-records
family
behaviour
pre-op
complications
change of
emergency
theatre-availability
action
NOW!
certain
uncertain
187. A surgical exampleā¦
before
patient identity
theatre
booking
equipment
plan
we need to be certain
about all of these
verify identity
consumables
action-records
NOW!
certain
uncertain
189. A surgical exampleā¦
before
we donāt expect
these to happen,
but we need
contingency-plans
and guiding-principles
for all of them
family
behaviour
pre-op
complications
emergency
action
NOW!
certain
uncertain
190. But to let our brave dung-beetle
have the last wordsā¦
always rememberā¦
CC-BY-NC-ND wisse via Flickr
197. Further information:
Contact:
Tom Graves
Company:
Tetradian Consulting
Email:
tom@tetradian.com
Twitter:
@tetradian ( http://twitter.com/tetradian )
Weblog:
http://weblog.tetradian.com
Slidedecks:
http://www.slideshare.net/tetradian
Publications: http://tetradianbooks.com
Books:
ā¢ The enterprise as story: the role of narrative in enterprisearchitecture (2012)
ā¢ Mapping the enterprise: modelling the enterprise as services
with the Enterprise Canvas (2010)
ā¢ Everyday enterprise-architecture: sensemaking, strategy,
structures and solutions (2010)
ā¢ Doing enterprise-architecture: process and practice in the
real enterprise (2009)