Emergence

1
Emergence

• John Martin
• For OUSyS AGM 27-1-07

2
Too much theory?

• So far, we have never hired speakers for
  particular topics, so OUSyS presentations have
  depended on unpaid volunteers and their
  enthusiasms
• My topic is, indeed, theoretical, but my view
  is that traditional Systems concepts contain a
  number of subtle traps and biases that have
  practical implications.
• Personally, I feel quite trapped by the classic
  Systems concepts, and the very poor linkage to
  related fields, so this topic is part of a wider
  personal project to unpick these traps

3
EMERGENCE how, when and why the whole is
greater than the sum of its parts (EPSRC
sandpit Oct 06 budget 1.45 M)

• For the first time since the enlightenment
  we have started to understand that there are
  non-causal systems in which some things just
  are. The concept of emergence accepts that
  even with the same starting conditions the same
  pattern would not necessarily repeat.
• Complexity is not the only factor influencing
  emergence, for example human and animal activity
  are also driven by context and instinct,
  climate change is influenced by human activity as
  well as weather patterns, disease transference is
  influenced by technology , social and
  commercial factors as well as human contacts.
  
• Emergence occurs in a very broad range of areas,
  so cross-fertilisation of ideas and approaches is
  essential. From anti-terrorism to zoology, and
  from understanding behaviour to meeting the
  requirements of society the level of uncertainty
  in managing the future is on the increase.
• Topics to be considered in the sandpit might
  include
• Is complexity a metaphor, a reality or both?
• What types of research method are appropriate to
  study emergence in human systems or physical
  systems?
• What is the role of modelling and simulation in
  systems with emergent properties? What are their
  limits?
• How could an understanding of complexity improve
  policy formation and practice in both Government
  and Industry?
• Examples of disciplines and research areas that
  could be potential contributors include
• Anthropology, Art, Behavioural sciences,
  Biology, Chemistry, Cognitive Science, Complexity
  Science, Computer Science, Computer modelling,
  Design, Economics, Engineering, Human Factors,
  Management, Medicine, Meteorology, Mathematics,
  Philosophy, Physics, Psychology, Sociology,
  Statistics, Technology, Zoology

4
Gianfranco Minati (UKSS, 2006)

• Emergence refers to the core theoretical
  problems of the processes by which systems are
  established.
• We are now facing the process by which GST is
  becoming more and more a Theory of Emergence ...
  , searching for suitable models and
  formalizations of its fundamental bases.

5
Explanations and examples
A transcript of the definitions and examples
generated in this exercise is in the file
Ovals.rtf
6
Similarities and differences

• The handout sheet lists some of the potentially
  emergence-related events, properties, etc. that I
  came across in the literature, and you have just
  generated some more!
• Spend 10 minutes working with one or two others
  to see if you can identify any clusters or
  distinctions that seem helpful in making sense of
  this sea of alleged examples
• For each cluster or distinction you come up with,
  name it on an oval and stick it up on the wall.
• Be selective! Just pick out what catches your
  attention

The handout sheet is available as
EmergenceEGs.rtf A transcript of the resulting
categories is in Ovals.rtf
7
So what is emergence?
8
The traditional Systems view
A
B
A is not the same as B A is a whole with
emergent properties
9
Water the many disciplines needed to study its
many aspects, such as

• Atomic structure
• Ice/water/steam state changes
• Compressibility, surface tension, cohesion,
  adhesion, capillarity
• Dynamics of temperature change
• Buoyancy
• Patterns of flow and turbulence
• Tidal action
• Waters role in world climate
• Adapted from Corning, 2002

10
But be warned (!!)

• We're often told that certain wholes are more
  than the sum of their parts. We hear this
  expressed with reverent words like holistic and
  gestalt, whose academic tones suggest that they
  refer to clear and definite ideas.
• But I suspect the actual function of such terms
  is to anesthetize a sense of ignorance.
• We say gestalt when things combine to act in
  ways we can't explain, holistic when we're
  caught off guard by unexpected happenings and
  realize we understand less than we thought we
  did.
• Minsky, 1986, quoted in Ronald, Sipper and
  Capcarrère, 1999

11
contradictory opinions abound

• There is no universally acknowledged definition
  of emergence, nor even a consensus about such
  hoary (even legendary) examples as water.
• And if emergence cannot be defined in concrete
  terms so that you will know it when you see it
  how can it be measured or explained?
• As Jeffrey Goldstein noted in his Emergence
  article, "emergence functions not so much as an
  explanation but rather as a descriptive term
  pointing to the patterns, structures or
  properties that are exhibited on the
  macro-scale.
• Corning, 2002

12
Origins

• The whole is something over and above its
  parts, and not just the sum of them all
• (Aristotle, Book H, 10458-10)
• Every resultant is either a sum or a difference
  of the co-operant forces their sum, when their
  directions are the same their difference, when
  their directions are contrary. Further, every
  resultant is clearly traceable in its components,
  because these are homogeneous and
  commensurable....
• It is otherwise with emergents, when, instead of
  adding measurable motion to measurable motion, or
  things of one kind to other individuals of their
  kind, there is a cooperation of things of unlike
  kinds.... The emergent is unlike its components
  in so far as these are incommensurable, and it
  cannot be reduced to their sum or their
  difference.
• G.H.Lewes (1874-79)

13
Samuel Alexander - 1920

• the emergence of a new quality from any level
  of existence means that at that level there comes
  into being a certain constellation or collocation
  of motions belonging to that level, and this
  collocation possesses a new quality distinctive
  of the higher-complex.
• To adopt the ancient distinction of form and
  matter, the kind of existent from which the new
  quality emerges is the matter which assumes a
  certain complexity of configuration and to this
  pattern or universal corresponds the new emergent
  quality.
• Quoted in Rueger, 2000

14
Subsequent history

• 19th century - teleological ideas about
  evolution Lamarck, Lloyd Morgan, Alexander,
  Broad, Smuts, Lovejoy, etc. Quantitative,
  incremental, change can lead to unpredictable
  qualitative changes, irreducible to their parts.
  Tended to include creative divinity, vitalism,
  etc.
• Early 20th century - quashed by reductionists
  McDougall, Carnap, Russell, etc.
• Goes underground in the 30s Needham, Huxley,
  Novikoff, Tansley (Ecosystem), Lindeman.
• GST in the 50s von Bertalanffy, Boulding,
  Ashby, von Foerster, Ackoff, Beer, and many
  others.
• Complexity theory starting in the 70s Sperry,
  Haken, Prigogine, Santa Fe Institute, Kauffman,
  Holland, etc.
• Adapted from Corning, 2002

15
Alife as a modern influenceE.g. cellular
automata such as Conways Game of Life
(Gardner, 1988)

• There is a two-dimensional rectangular grid of
  cells, such as a checker board.
• A cell's state at a given time is determined by
  the states of its eight neighbouring cells at the
  preceding moment, according to the birth-death
  rules
• A living cell dies if fewer than 2 neighbours are
  alive. (loneliness)
• A dead cell becomes alive if 3 neighbours are
  alive (breeding)
• A living cell dies if more than 3 neighbours are
  alive. (overcrowding).
• Adapted from Bedau, 1997

16
A Game of Life in progress
17
Some rather simple Alife emergence criteria

• Design The system has been constructed by a
  designer, by describing local elementary
  interactions between components (e.g. artificial
  creatures and elements of the environment) in a
  design language.
• Observation The observer is fully aware of the
  design, but describes global behaviors and
  properties of the running system, over a period
  of time, using an observation language.
• Surprise The design language and the observation
  language are distinct, and the causal link
  between the programmed elementary interactions
  and the observed behaviors is non-obvious to the
  observerwho therefore experiences surprise.
• In other words, there is a cognitive dissonance
  between the observer's mental image of the
  system's design and his contemporaneous
  observation of the system's behavior.
• Adapted from Ronald, Sipper and Capcarrère, 1999

18
Some more specific Alife criteria for emergence

• Bedau argues that even though Game of
  Life-type processes are determinate,
  nevertheless the only way we can work out how
  they will turn out is to try them out, and see
  what happens they are beyond any short-cut
  calculation.
• So for him, the definition of an emergent
  property is that its behaviour can only be
  derived by experimental observation.
• Bedau 1997

19
A more mathematical approach

• Rueger explains novelty and irreducibility in
  terms of structural instability, parameters
  that bifurcate and differences in scale (e.g.
  Einstein v. Newton)
• From Rueger, 2000

20
Emergence as a sub-set of synergies

• Joint environmental conditioning
• Information sharing
• Joint decision making
• Risk-sharing
• Mutual catalysis Etc.

• Functional complementarities (Velcro, NaCl)
• Division of labour
• Symbiosis (ruminants)
• Scale (avalanche, Emperor penguins, gregarious
  nesting)

Synergies become emergent (Corning suggests) if
they are based on dissimilar things that generate
qualitatively novel effects. They do NOT depend
(he argues) on a perceiver, or on
self-organisation. Corning, 2002
21
But there are also observer-dependant
explanations

• In this view emergence is a process of
  detection by the observer of the formation of new
  collective properties (different from those of
  the individual components), self-organized by the
  coherent behaviour of interacting components.
• The observer detects properties as new depending
  on the cognitive model used, suitable for
  detecting (i.e., cognitively generating)
  coherence.
• Minati, 2006

22
Goldstein (1999)(in inaugural issue of
Emergence)

• The arising of novel and coherent structures,
  patterns and properties during the process of
  self-organization in complex systems.
• Common characteristics include
• Radical novelty (features not previously observed
  in the system)
• Coherence or correlation (meaning integrated
  wholes that maintain themselves over some period
  of time)
• A global or macro "level" (i.e., a property of
  "wholeness")
• Being the product of a dynamical process (it
  evolves)
• Being "ostensive" (it can be perceived)
• Supervenience (downward causation)
• Quoted in Corning, 2002

23
Some useful distinctions

• Designed vs unpredicted
• Resultant vs emergent properties
• Composition vs emergence
• Weak vs strong
• Diachronic vs synchronic
• Different sources of wholeness or surprise
• Different kinds of component entity

24
Designed vs. unpredicted

• Designed emergence These are the intended
  properties of an assembly that arise from the
  inter-connections we have built into it. E.g. a
  car has emergent properties that are deliberately
  built into its components and how they are
  arranged. If the car breaks down, a mechanic
  knows how to fix it.
• Unpredicted emergence Events or properties may
  emerge unpredictably in a situation, and may
  often be qualitatively unlike other properties in
  the situation.
• In unpredicted emergence, we are, of course,
  free to look (retrospectively) for a system
  that might have generated those
  events/properties, but NB that this kind of
  emergence reflects our reaction to the event,
  not the explanation we adopt for it.

25
Resultant vs emergent properties

• Resultant properties arises simply from combining
  components in predictable ways. So if you add
  three 3 kg weights to make a 9 kg weight, it has
  new properties, but they are not emergent.
• Emergent properties tend to be structural the
  components are arranged so that they construct
  something very unlike the components themselves.
• See Reuger, 2000

26
Composition vs. Emergence

• Processes of composition between elements take
  place, for instance, by reacting, merging or
  diluting, when elements take on new positions or
  new roles in a structure. A typical example is
  given by crystal or molecular structure
  formation. Processes of composition give rise (as
  their result) to new stable or unstable entities
  having properties different from those of the
  components.
• Emergence takes place during and not as a result
  (such as a new state) of the process of
  interaction. In the process of emergence new
  properties are established thanks to the
  continuous process of interacting. This process
  sustains emergence (e.g., swarming).
• Minati, 2006

27
Weak vs. strong emergence

• In weak emergence, it is accepted that the
  causal properties of the emergent property are in
  principle derivable from the causal properties of
  its components and their arrangement. Causality
  is always upwards. This is consistent with
  normal materialist philosophy.
• In strong emergence, the emergent higher level
  properties are said to acquire forms of downward
  causality that can intervene in normal upward
  causality. This is controversial.

28
Strong emergence

• O'Conner wants
• to capture a very strong sense in which an
  emergent's causal influence is irreducible to
  that of the micro-properties on which it
  supervenes.
• It bears its influence in a direct 'downward'
  fashion, in contrast to the operation of a simple
  structural macro-property, whose causal influence
  occurs via the activity of the micro-properties
  which constitute it.
• OConner, (1994) quoted in Bedau, 1997

29
But does it make sense?

• Although strong emergence is logically
  possible, it is uncomfortably like magic. How
  does an irreducible but supervenient downward
  causal power arise, since by definition it cannot
  be due to the aggregation of the micro-level
  potentialities?
• Such causal powers would be quite unlike
  anything within our scientific ken. This
  indicates how they will discomfort reasonable
  forms of materialism. Their mysteriousness will
  only heighten the traditional worry that
  emergence entails illegitimately getting
  something from nothing.
• Bedau, 1997

30
Strong emergence rides again!

• emergent phenomena in the natural world
  involve multilevel systems that interact with
  both lower- and higher-level systems
• Furthermore, these emergent systems in turn
  exert causal influences both upward and downward
  not to mention horizontally. (If determinism is
  stratified, it is also very often "networked.").
• The search for "laws" of emergence, or some
  quantum theory of living systems, is destined to
  fall short of its goal because there is no
  conceivable way that a set of simple laws, or
  one-level determinants, could encompass this
  multilayered "holarchy" and its inescapably
  historical aspect.
• Corning, 2002

31
Supervene (From Wikipaedia article)

• Suppose two objects, X and Y, both share a
  particular set of properties (B), and this
  automatically means that they must therefore also
  share another set of properties (A). Then A
  is super-venient on B (which is sub-venient
  on A).
• E.g. if psychological properties supervene on
  physical properties, then any two persons who are
  physically indistinguishable must also be
  psychologically indistinguishable or
  equivalently, any two persons who are
  psychologically different (e.g., having different
  thoughts), must be physically different
• Supervenience has traditionally been used to
  describe relationships between sets of properties
  in a manner which does not imply a strong
  reductive relationship.
• For example, many hold that economic properties
  supervene on physical properties, in that if two
  worlds were exactly the same physically, they
  would also be the same economically.
• However, this does not entail that economics can
  be reduced in any straightforward way to physics.
  
• Thus, supervenience allows one to hold that
  "high-level phenonema" (like those of economics,
  psychology, or aesthetics) depend, ultimately, on
  physics, without assuming that one can study
  those high-level phenomena using means
  appropriate to physics.
• Concept introduced by the US philosopher, Donald
  Davidson

32
Diachronic and synchronic

• Diachronic emergence This refers to surprising
  new events that appear over time. So if your
  computer crashes, that could be described as
  diachronic emergence.
• Synchronic emergence This is the traditional
  notion, where emergent properties or states
  at high levels are built on lower level
  arrangements of components i.e. the high and
  low levels are present at the same time.
• See Rueger, 2000

33
Different sources of wholeness or surprise

• From wholes that have an independent reality
• Natural processes e.g. wholes that are the
  result of statistical attractors in complex
  random structures. Often very sensitive to
  starting conditions
• Self-organization i.e. where control loops
  are self-maintaining
• Enactment i.e. wholes that we have designed
  and built so that they are discrete objects
• Embedding i.e. we tend to break up reality in
  ways that will provide structures that are
  economically or socially optimal for us
• Reifying the surprising i.e. processes that
  are important to us but are not understood tend
  to be labelled as nouns (e.g. a disaster)
• Embodiment i.e. human neurology processes or
  categorises information in this way e.g.
  chunking, linguistic labels, and Lakoffs
  basic categories
• to wholes that exist only in our minds

34
Different kinds of component entity

• E.g. Minatis notions of Multiple Systems (MS)
  and Collective beings (CB)
• An MS is a set of systems established by the
  same elements interacting in different ways
  i.e. having multiple simultaneous roles (e.g.
  models with multiple memory systems).
• A CB is an MS established by agents possessing
  the same (natural or artificial) cognitive
  system
• Minati, 2006

35
Two closing questions

• Did any of those distinctions echo the
  distinctions you made?
• A discussion-provoker Could brain surgeons
  operate on their own brains (presumably an
  example of downward causality)?