Changing%20the%20Rules%20of%20the%20Game

1
Changing the Rules of the Game

• Dr. Marco A. Janssen
• Department of Spatial Economics

2
Research questions

• How do rules emerge, get selected and be
  remembered in social ecological systems?
• What can we learn from (computational models of)
  immune systems and language development?

3
Contents

• Puzzles from empirical studies of common pool
  resources.
• Immune system
• Language development
• Methodology
• Modeling self-organization of institutions
• Discussion

4
Common Pool Resources

• Are used by multiple-users
• For which joint use involves subtractability,
  that is, use by one user will subtract benefits
  from another users enjoyment of the resource
• It is difficult to exclude users

5
Management of CPRs

• Economic Theory predicts Nash equilibrium and
  overharvesting
• Solutions to derive cooperative solution
• Government will manage the resource
• A market will be created
• Laboratory experiments and field studies show an
  alternative self-organization of institutions.

6
Factors important for self-organization

• Type of communication
• Building up mutual trust relationships
• Rules how to monitor and sanction defined by the
  local users and implemented by local users
• Memory of successful solutions by taboos,
  rituals, religions, etc.

7
Immune System

• Distributed system which is able to detect and
  eliminate invasions of pathogens.
• Detection self vs non-self
• Response generation antibodies
• Memory storing successful responses

8
Pathogens

• Bacteria
• Parasites
• Viruses
• Fungi

9
Detection
10
Recognition
11
Response

• - Continue generation of new cells.
• - Replication of cells which bind lots of
  pathogens Antibodies
• - Antibodies neutralize pathogens

12
Impact of Memory
13
Artificial Immune Systems

• Distributed systems for information processes.
• Origin
• study of immune systems
• bio-algorithms
• genetic algorithms
• neural networks

14
Language development

• Different perspectives on language.
• Universal grammar/language
• Genetic transmission
• Localized hard-wired neurological structures
  crickets and songbirds
• Higher animals learn language gradually training
  parameters of neural network

15
Complex adaptive system approach

• Language
• result of local interactions of language users
• self-organizing process
• agents benefit from being understood (fitness)
• clustering of agent with same language/dialect

16
Methodology

• Games
• game theory for institutions, repeated games with
  prisoners dilemma
• language games, imitation games
• evolution of grammar fitness related to mutual
  understanding

17
Vowels
18
Emergence of vowels by adaptive imitation games
(De Boer, 2000)
19
Methodology (II)

• Networks
• Neural networks learning by finding the right
  connection strengths
• Immune networks maintaining immune memory,
  spreading information over other parts of the
  network.
• Social networks.

20
Methodology (III)

• Evolutionary Computation
• Genetic and evolutionary algorithms
• fitness
• selection
• mutation
• (cross-over)

21
Modeling self-organization of institutions

• Coding rules
• Creating rules
• Selecting rules
• Remembering rules

22
Coding rules

• Grammar of Institutions (Crawford and Ostrom,
  1995)
• Rules are build up from 5 components
• Attributes (characteristics of the agents)
• Deontic may/must/must not
• Aim action of the agent
• Conditions when, where and how
• Or else sanctions when not following a rule

23
Creation of Rules

• Mutations and cross-over
• Immune systems constant generation of new
  lymphocytes
• Language interaction with other groups and with
  new experiences
• Computer led to new words (e-mail internet) and
  new meanings (windows mouse)
• Social groups jargon of scientists

24
Genetic Libraries
25
Selection of Rules
Rules Constitutional Collective
Operational
Levels of analysis
Constitutional Collective Operational choice cho
ice choice
Processes Formulation Policy-making Appropriatio
n Governance Management Provision Adjudicatio
n Adjudication Monitoring Modification Enfor
cement
26
Selection of rules (II)

• Criteria for success
• Social networks
• Mutual trust relationships
• Recognition of trustworthy others (reputation,
  symbols, indirect reciprocity)

27
Remembering Rules

• Law, universities, taboos, rituals, religions
• Reinforcement and disturbances
• Resilience
• Redundancy

28

29

Coverage of antigen space by antibodies
30
Fitness versus redundancy
(Hightower et al, 1995)
31
Fitness related to redundancy
(Hightower et al, 1995)
32
Training the system

• Allow small disturbances to maintain training of
  the strength of the network, the diversity and
  functional redundancy

33
Discussion

• Empirical evidence for self-organization of
  institutions.
• Formal models may help to explain observations.
• But how to formally model how rules emerge, get
  selected and be remembered?

34
Discussion (II)

• We may learn from similarities and differences
  between institutions, immune systems, and
  language development.
• Computational tools exists to simulate immune
  systems and language development
• Toward computational laboratories for
  social-ecological systems.