Transcript Document

Dialogue on
Complexity & Design
12 January 2005
Eve Mitleton-Kelly
Director
Complexity Research Programme
London School of Economics, UK
[email protected]
http://www.lse.ac.uk/complexity
Familiar terms
Fractals
Attractors
Paradoxes
Edge of chaos
etc
CHAOS THEORY
Complexity
• Interrelationships
• Connectivity & interdependence
• Multiplicity
• CREATION OF NEW ORDER
Complexity theory
•
•
•
•
•
Context, time, history
Process, meaning, politics, power
Emergence, contingency, feedback
Novelty, change, evolution, transition
Continuity of identity over time
Theories
self-organisation
Natural sciences
Dissipative structures
chemistry-physics (Prigogine)
Generic
Autocatalytic sets
characteristics
evolutionary biology (Kauffman)
of complex
co-evolving
Autopoiesis (self-generation)
systems
biology/cognition (Maturana)
emergence
connectivity
interdependence
feedback
far from equilibrium
space of possibilities
co-evolution
Chaos theory
historicity & time
Social sciences
path-dependence
Increasing returns
economics (B. Arthur)
creation of new
order
Clusters:
1.
•
•
•
•
2.
•
•
3.
•
•
•
•
•
Connectivity & interdependence
Self-organisation
Emergence
Feedback
Co-evolution
Exploration of the space of possibilities
Far from equilibrium & dissipative structures
Historicity & time
Path dependence
Creation of new order
Organisations and a different logic
Connectivity & interdependence
• Networks of relationships with different degrees
of connectivity
– strength of coupling
– epistatic interactions
i.e. the fitness contribution made by one
individual will depend upon related
individuals
• Essential element of feedback
Connectivity
Diversity
Density
Intensity
Quality
of interactions
between human agents
Determine network of relationships
Emergence
• Emergent properties or qualities or
patterns
• Arise from interaction
• Cannot be predicted
Self-organisation
• Spontaneous ‘coming together’
• Not directed or designed by someone
outside the group
• The group decides what needs to be done,
how, when …
• Can be a source of innovation
• Consider what facilitates self-organisation
Feedback
2 mechanisms:
• Reinforcing (amplifying) – a driver for
change – positive feedback
• Balancing (moderating or dampening) creates stability – negative feedback
• Processes not mechanisms
– need time dimension
Feedback Process
not Mechanism
to avoid the machine metaphor
A machine is a system, which we can:
–
–
–
–
understand
design
plan its operation in detail
predict its behaviour
and
– control
A machine:
• Is a complicated system
• With many inter-related parts
• Relies on feedback
• Can be thought of as an object
• Feedback in this context is taken to mean
influence, which changes potential action
and behaviour.
Influence
– Not uniform
– It depends on the degree of connectivity
– Actions and behaviours vary with different
individuals
– With time and context
– Reciprocal
Feedback links the micro and the macro
processes
– The microscopic events and the macroscopic
emergent structures or patterns change and
evolve and in so doing influence each other
through feedback processes.
Small Groups
• Is your understanding of self-organisation and
emergence different from that discussed? In what
way? How do you think about them?
• How do they relate to design? Can you identify
examples of self-organisation and emergence in
design?
• What was the role of feedback?
Cluster 2
• Co-evolution
• Exploration-of-the-space-of
possibilities
Co-evolution
• Reciprocal influence that changes the
interacting entities
• Co-evolution within a social ecosystem
– not just adaptation to the environment
• One domain changes in the context of the
other.
Co-evolution within an ecosystem
Co-evolution in a Social
Ecosystem
• A social ecosystem includes:
»Social
»Cultural
»Technical
»Geographic
»Economic
»Political dimensions
Exploration of the space of
possibilities
• Exploration of new options, different ways of
working and relating
• The search for a single 'optimum' strategy is
neither possible nor desirable, in a turbulent
environment – multiple micro-strategies +
distributed strategies, power, intellectual cap.
• But variety alone is not enough. New connections
or contributions also need to be ‘seen’.
Exaptation
• Often not expensive R&D which produces major
innovations, but ‘seeing’ a novel function, in a
new light.
• “Exaptation is the emergence of a novel function
of a part in a new context. … Major innovations in
evolution are all exaptations. Exaptations are not
predictable” [Kauffman, Complexity and Technology
Conference, London, 11 March 1997]
Adjacent possible
• When searching the space of possibilities,
whether for a new product or a different
way of doing things
• It is not possible to explore all possibilities
• But it is possible to consider change one
step away from what already exists.
Fitness ‘Landscape’
• In the competition for survival, species
attempt to alter their make-up by taking
‘adaptive walks’ to move to higher ‘fitness
points’, where their viability is enhanced.
• Adaptive walks are an optimisation
technique for searching a space of
possibilities.
• Powerful technique – able to search many
parts of the space in parallel (Kauffman)
Fitness ‘Landscape’
N = number of entities or elements in a system
K = degree of connectivity between the entities
• Each entity N makes a fitness contribution which
depends upon that entity and upon K other entities
among the N
• K reflects the rich cross-coupling of the system
• K measures the richness of epistatic interactions
among the components of the system.
[NK model, The Origins of Order, Kauffman, 1993]
Organisational Fitness
Landscapes
• Concept may be applied to evolutionary journey
of an organisation.
• Consider multiple micro-strategies, exploring the
space of possibilities.
• Success of strategies of an organisation is
determined by the strategies of the other entities in
the same ecosystem.
• Inter-coupling of landscapes + richness of
individual interactions – alter the co-evolutionary
dynamics
• A complex co-evolving ecosystem is one of
intricate and multiple intertwined interactions
and relationships.
• Connectivity and interdependence propagate
the effects of actions, decisions and behaviours
throughout the ecosystem.
• Depend on degree of connectivity
Creation of community
Small Groups
• Can you identify examples of co-evolution,
exploration of the space of possibilities,
exaptation and the adjacent possible?
• How would they work as necessary
conditions in the design process?
• How would you employ micro-strategies
and use distributed intelligence?
Cluster 3
•
•
•
•
Far-from-equilibrium
Historicity & time
Path dependence
Creation of new order
• Organisations and a different logic
Far-from-equilibrium
&
Dissipative Structures
Ilya
Prigogine
• Benard cell - example of a
physico-chemical dissipative
structure
• “By applying an external
constraint we do not permit the
system to remain at equilibrium.”
[Nicolis & Prigogine 1989, p10]
Several things have happened:
(a) self-organisation: the water molecules have
spontaneously organised themselves into righthanded and left-handed cells;
(b) from molecular chaos the system has created
order and a structure has emerged;
(c) the handedness or direction of rotation can
neither be predicted nor controlled although we
can predict that the cells will appear;
(d) the system was pushed far-from-equilibrium by
an external constraint or perturbation;
(e) the homogeneity of the molecules at equilibrium
was disturbed and their symmetry was broken.
(f) the particles behaved in a coherent manner,
despite the random thermal motion of each of
them.
This coherence at a macro level characterises emergent
behaviour, which arises from micro-level interactions of
individual elements.
• In classical thermodynamics heat
transfer or dissipation was considered
as waste, but in the Benard cell it has
created new order.
• It is this ability of complex systems to
create new order and coherence,
which is their distinctive feature.
Ilya Prigogine’s contribution
• Reinterpretation of the Second Law of
Thermodynamics.
• Time-irreversible processes are a source of order
• Arrow of time need not be associated with
disorder
• Dissolution into entropy is not a necessary
condition – but “under certain conditions,
entropy itself becomes the progenitor of order.”
Ilya Prigogine’s contribution
• To be more specific, “... under nonequilibrium conditions, at least, entropy
may produce, rather than degrade, order
(and) organisation ... If this is so, then
entropy, too, loses its either/or character.
While certain systems run down, other
systems simultaneously evolve and grow
more coherent.”
[Prigogine & Stengers 1985, p. xxi]
Bifurcation:
• Splitting into alternative solutions.
• “Several solutions are possible for the same
parameter values.
• Chance alone will decide which of these solutions
will be realized. The fact that only one among
many possibilities occurred gives the system a
historical dimension, some sort of “memory” of a
past event that took place at a critical moment and
which will affect its further evolution.”
[Prigogine and Nicolis 1989]
Summary of characteristics
•
•
•
•
•
Self-organisation
Creation of order
Emergence of structure
Coherence
Precise behaviour can neither be predicted nor
controlled
• Far-from-equilibrium – external constraint
• Symmetry breaking
• Bifurcation: several possible solutions
Complex Social Phenomena
• Historical dimension & the role of time
• Chance events, unfolding in time, are
intertwined to generate social phenomena
• Qualitative approach
• Narrative captures the historicity of social
phenomena
Path dependence
• Previous interactions bring about what we
currently experience
• e.g. technological and economic changes
are path dependent
• Increasing returns – Brian Arthur
• The form and direction they take depend
on the particular sequence of events that
preceded them
Why complexity thinking?
• Seeing organisations as complex coevolving systems and by understanding
their CCES characteristics we can facilitate
learning and sustainability.
• We often inadvertently constrain these
characteristics and limit innovation and the
creation of new order.
Change of emphasis
from objects
• to relationships between entities
from control
• to enabling infrastructures
Enabling Infrastructure
Combination of cultural, social and
technical conditions which facilitate ‘x’
Conditions
enable
inhibit
A CCES organisation:
•
•
•
•
•
Facilitates (does not actively inhibit) emergence
Encourages self-organisation
Explores its space-of-possibilities
Facilitates co-evolution
Understands about degrees of connectivity &
interdependence
• Appreciates its distributed intellectual capital
• Fosters a collaborative culture
A CCES organisation:
• Creates variability
large repertoire of
responses
• Able to cope in an unpredictable environment
• Not too organised and not too random
• Emphasises Enabling Infrastructures (not C&C)
• Facilitates the emergence of new order
- new ways of working and relating
- new organisational forms
- generation & sharing of knowledge
Small Groups
• What does ‘design’ mean from a complexity
perspective?
• What difference does it make to our
thinking about the design process
• Is it possible to ‘design’ an organisation?
How?