Consistent definitions of emergence, selforganization and selfassembly, and a hypothesis about the r

1
Consistent definitions of emergence,
self-organization and self-assembly, and a
hypothesis about the role of criticality in
evolution

• Julianne D. Halley
• David A. Winkler

2
Overview of Interaction Task and 3 papers

• Interaction Task Self-organization and natural
  selection
• Paper 1. A simple description of emergence and
  its relation to self-organization
• Paper 2. Towards Consistent Concepts of
  Self- organization and Self Assembly
• Paper 3. Self-organized noisy templates interact
  with natural selection

3
Interaction Task Self-organization and selection

• Biologists have, as yet, no conceptual framework
  in which to study an evolutionary process that
  commingles both self-organization and selection
• Kauffman (1995) At Home in the Universe
• There is a need for new theoretical frameworks
  that allow us to understand how
  self-organization, selection and historical
  accident find their natural places with one
  another
• Solé et al. (1999) TREE 14(4)156

4
Interaction Task Self-organization and
selection co-evolutionary architects
of diversity

• A cross-disciplinary review of the ways in which
  natural
• selection and self-organization interact
• Specifically, we ask
• 1) In your area of scientific enquiry, do
  selection processes and self- organization
  interact? If so, how?
• 2) How do these two processes give rise to
  diversity and how are they affected by it?
• 3) On which scales does diversity arise?
• 4) Diversity could be viewed as an emergent
  property of selection and self-organization.
  What other emergent properties feedback to
  affect these processes?

5
Interaction Task Outcomes

• 1. a cross-disciplinary review of ways in which
  selection and self-organization interact and
  facilitate diversity and complexity evolution
  (aimed at PNAS)
• 2. an integrated (but not necessarily convergent)
  set of papers suitable for publication in a
  complexity journal, as a special issue
• One-day Symposium at Melbourne Uni (Dec 2005)
• Autonomous Agency and the Evolution of
  Diversity
• Workshop in Katoomba (May 2006)
• Selection, Self-Organization and Diversity

6
Interaction Task Self-organization and
selection co-evolutionary architects
of diversity
7
Overview of Interaction Task and 3 papers

• Interaction Task Self-organization and natural
  selection
• Paper 1. A simple description of emergence and
  its relation to self-organization
• Paper 2. Towards Consistent Concepts of
  Self- organization and Self Assembly
• Paper 3. Self-organized noisy templates interact
  with natural selection

8
Emergence IT Our view of emergence

• In some sciences, emergence implies
  non-equilibrium self-organization
• e.g. emergence of collective intelligence in ant
  colonies
• In others, emergence can refer to collective
  equilibrium properties
• e.g. temperature and pressure of ideal gasses

To highlight these fundamentally different
collective properties, we recognise two classes
of emergence
9
Simple and Complex emergence

• Simple emergence implies near equilibrium
  behaviour
• e.g. temperature and pressure of gasses,
  self-assembled materials
• Complex emergence implies non-equilibrium
  self-organization
• e.g. convection cells, collective intelligence

10
Emergence IT Our view of emergence
Onset of self-organization Threshold of complex
emergence
Simple emergence
Complex emergence
simple systems largely linear near
equilibrium computable at component level
complex systems nonlinear non-equilibrium
incomputable at component level
11
Complex systems science transcends boundaries

• Our definitions must transcend many boundaries
• Should be simple enough so that all interested
  parties can participate in discussion over wine

12
Overview of Interaction Task and 3 papers

• Interaction Task Self-organization and natural
  selection
• Paper 1. A simple description of emergence and
  its relation to self-organization
• Paper 2. Towards Consistent Concepts of
  Self- organization and Self Assembly
• Paper 3. Self-organized noisy templates interact
  with natural selection

13
Self-organization and self-assembly

• Both describe decentralised processes that
  produce collective order from small scale dynamic
  interactions
• Terms used interchangeably, or with a different
  emphasis and meaning in each science
• Whitesides and Grzybowski (2002) static and
  dynamic self-assembly
• Yamaguchi et al. (2004) consider SA and
  dissipative structure as equilibrium and
  non-equilibrium forms of pattern formation, both
  of which are SO
• In social insect literature
• only constituents of final structure take part
• qualitative stigmergy SA
• We reviewed definitions of both terms and came to
  the following conclusions

14
What is self-organization?

• Many definitions are wedded to specific
  disciplines
• adaptive, function, goal, efficiency
• Some definitions use other complex terms, such as
  emergence
• Some argue that self and organization must
  first be defined
• No external control or guiding template

15
What is self-organization?

• Most agree that SO requires an energy source
• although magnetization and crystallization have
  been considered classic examples of SO
• Nonlinearity (positive feedback) essential but
  inadequate by itself

Self-organization is a dissipative
non-equilibrium order at macroscopic levels, due
to collective, nonlinear interactions between
multiple microscopic components This order is
influenced by interplays between intrinsic and
extrinsic factors, and decays upon removal of the
energy source
16
What is self-assembly?

• Overused to the point of cliché
• In chemistry, SA often refers to energy
  minimization and thermodynamic equilibrium
• But in nanochemistry, SA often used
  interchangeably with SO
• In the social insect literature, it has been
  argued that SA and SO are conceptually indistinct
  
• Probably because SA is used in two different
  contexts!

17
Self-organization and self-assembly

• Nobel prize winner Lehn suggested that SO and SA
  could be distinguished on a thermodynamic basis.
• Self-organization implies non-equilibrium
• Self-assembly reserved for processes tending
  toward equilibrium
• this distinction is consistent with the early
  uses of both terms
• including that by Nobel prize winner Prigogine in
  his classic 1977 book Self-organization in
  non-equilibrium systems
• we argue that qualitative stigmergy should be
  called agent assisted self-assembly

18
What is self-assembly?
Self-assembly is a non-dissipative structural
order on a macroscopic level that is due to
specific interactions among (usually microscopic)
components This order is encoded in the rules of
interaction and does not require an energy source
19
Overview of Interaction Task and 3 papers

• Interaction Task Self-organization and natural
  selection
• Paper 1. A simple description of emergence and
  its relation to self-organization
• Paper 2. Towards Consistent Concepts of
  Self- organization and Self Assembly
• Paper 3. Self-organized noisy templates interact
  with natural selection

20
Our hypothesis about the interaction between
natural selection and self-organization in
evolution

• Our proposition comprises 2 components
• critical-like dynamics self-organize readily in
  non-equilibrium systems
• critical-like dynamics provide a template upon
  which selection grafts additional features

21
Critical-like dynamics emerge readily in
non-equilibrium systems
In slowly driven non-equilibrium systems,
self-organized criticality can produce
critical-like dynamics
In rapidly driven non-equilibrium systems, rapid
SOC can produce critical-like dynamics
Kinouchi and Prado (1999) PRE 59, 4964 De
Carvalho and Prado (2000) PRL 84, 4006 Halley
et al. (2004) PRE 70, 036118
22
Critical-like dynamics provide a template
uponwhich selection builds further elaborations

• Once formed, critical-like fluctuations
• may be exposed to natural selection
• Are such dynamics detrimental?
• Are they beneficial?
• Are they selectively neutral?

Note that interaction between criticality and
natural selection could modify the criticality
beyond recognition
23
Critical-like dynamics can be modified by natural
selection in two fundamental ways

• Modification of collective properties
• (subcritical, supercritical, oscillations)

Can a particular scale be stabilized?
24
Critical-like dynamics can be modified by natural
selection in two fundamental ways

• Differentiation of individual behaviour

• Is it beneficial for only some individuals to
  participate in the collective process?
• Individual differences reinforced and refined by
  selection

25
Examples
26
Implications

• Evolution of key features of biological systems
  (compartmentalisation, division of labour,
  flexibility and robustness) could be facilitated
  by iterative interaction between SOC-like
  dynamics and natural selection
• Viewed from our perspective, the fractal geometry
  of nature appears anything but coincidental