Einingalíkön og hermun










Key Concepts
  Causation: Upward, downward  
  Diachronic, Synchronic  
  Emergent complexity / simplicity  











Emergence: What is it?
  Many definitions The concept comes into play in many disciplines
  Simple version "The whole is more than the sum of the parts"
  Psychological approach A system's level of emergence can be measured by how difficult it is, on average, to imagine how it works
  Information theory Complexity of a system is the amount of information necessary to describe the system
  Darley (1992)

"Emergent systems are those in which even perfect
knowledge and understanding [of their constituents] may give us no predictive information [of their behavior]."

Hence: "A true emergent phenomenon is one for
which the optimal means of prediction is simulation."

  Continuity "...there is no discontinuous separation between emergence and non-emergence. Emergence is purely the result of a phase change in the amount of computation necessary for optimal prediction of certain phenomena." (Darley 1992)
  "Dissipative structures" (Prigogine) From thermodynamics: dissipation of energy, increases entropy (See Second Law of Thermodynamics.)
  Energy flows into the system

Influx of energy into the system, dissipation of energy out of the system: export of energy (heat) increases entropy in environment; retained energy used to decreases internal entropy

















Emergence: Is it Real?
  In short, yes Numerous examples exist of the existence and operation of emergence

Balance automatically reached between supply & demand - "invisible hand" (Adam Smith)

  Ecologies Adjust automatically to variations in species, temperature, etc.
  Thinking Emerges (mostly) out of interactions between neurons
  Life Emerges out of interactions between organic molecules











Features of Analysis
  Elements ... and their number
  Interactions ... and their strength
  Formation/Operation ... and their timescales
  Diversity / Variability  
  Environment ... and its demands
  Activity ... and its objectives









(from: Thórisson, K. R. (2008). Modeling Multimodal Communication as a Complex System. To be published in I. Wachsmuth, M. Lenzen, G. Knoblich (eds.), Springer Lecture Series in Computer Scienc: Selected Writings in Embodied Communication. New York: Springer.)







When the system reaches a point in its evolution it can go in two or more directions; the directions are impossible to predict without going (at least) one level down in description.
  Phase transition e.g. from ice to water to gas














  Belousov-Zhabotinsky reaction A chemical "clock"; cyclical chemical reactions
  Belousov-Zhabotinsky Video (3x speed)


  Simulations of chemical reactions

Often modeled as cellular automata

  Example simulations
  Robot example "millipede" video
  Intuitive proof Ant farm excavation (video)
  Cellular automata The "flyer", Conway's "Life" (applet)
  RU software for exploring cellular automata Vélaldin Emergence Engine (link) by H. Th. Thórisson

















Types of Emergence
  Emergent Complexity

Complexity from simplicity: A system composed of simple parts exhibits complex features; the behavior of the whole is complex

Example: Life

  Emergent Simplicity

Simplicity from complexity: When only a few interaction rules result in ordered complexity. On a small scale the system is complex (e.g. the ecosystem on the Earth) but on the large scale the system behaves simply (Earth's orbit)

Example: The solar system

  Level of description How we view a system is dependent on the level of description we choose















Philosophical Stance
  Weak emergentism "Properties of a system are made up of material parts only." All substance-dualistic positions are rejected: Materialism.
    Diachronic emergentism "Properties of a system is impossible to predict from its basic elements"
    Synchronic emergentism "Properties of a system are not reductively explainable from its basic elements"













Diachronic example
  100x100 grid 2 colors allowed to change, green and brown
  5 rules

When green

  • Turn brown:If there are more than 20 green patches around and lifetime exceeds 30
    Turn brown: If there are less than 12 green patches around and lifetime exceeds 20
    Turn brown: If number of green patches around equal 25
    Turn brown: If lifetime under any circumstance exceeds 60

When brown

  • Turn green: If there are more than 8 green patches around and their lifetime combined exceeds 80 and there are more than 10 brown patches around
  Starting position  










(notice: extra gray and red cells in this example do not change simulation)












Emergence: Complexity from Simplicity
  Video, Part 1 NOVA
  Video, Part 2 NOVA