Advanced Computational Modeling of Social Systems Lars-Erik - - PowerPoint PPT Presentation

Lars-Erik Cederman and Luc Girardin Center for Comparative and International Studies (CIS) Swiss Federal Institute of Technology Zurich (ETH) http://www.icr.ethz.ch/teaching/compmodels

Advanced Computational Modeling

• f Social Systems

2

Presentations

• 07.06.2005 III
• 14.06.2005 III
• 21.06.2005 III
• 28.06.2005 II
• Nicolas Bürkler (4)
• Thomas Fent
• Stephan Gammeter (2)
• Jean-Luc Geering (1)
• Jeremy Hackney (2)
• Lutz Krebs (4)
• Benjamin Lutz (1)
• Charles Mitchell (3)
• Gloria Müller (3)
• Carsten Murawski (1)
• Jonas Nart (3)
• Nils Weidmann (2)

3

Computational Models of Social Forms (AJS 2005)

• Putting agent-based modeling in a social-

theoretic context:

– Sociological process theory offers theoretical guidance for future modeling – Computational modeling helps process theorists craft better, more precise theories

• Focus on research problems at the macro-

level

4

Outline

Computational Modeling Sociological Process Theory Epistemology Ontology Nomothetic theory Generative theory Nomothetic modeling Generative modeling Variable-based theory Configurative theory Variable-based modeling Agent-based modeling

5

Sociological process theory

• Simmel: sociations or

Vergesellschaftung

• Chicago school: Mead et al
• Elias’ “figurations”
• Barth’s formal anthropology
• Emirbayer’s “relationism”,

Giddens’ “structuration”, Archer’s “morphogenetic” approach, Fararo’s “generative theory”

Georg Simmel George Herbert Mead Norbert Elias

6

The epistemology of process theory

Compuational Modeling Sociological Process Theory Epistemology Ontology Nomothetic theory Generative theory Nomothetic modeling Generative modeling Variable-based theory Configurative theory Variable-based modeling Agent-based modeling

7

Toward generative theory

• Covering laws ==> uncovering mechanisms
• Positivism ==> scientific realism
• Abductive inference instead of induction or

deduction:

– making the puzzling less puzzling – metaphors, analogies, and simple models

8

The ontology of process theory

Computational Modeling Sociological Process Theory Epistemology Ontology Nomothetic theory Generative theory Nomothetic modeling Generative modeling Variable-based theory Configurative theory Variable-based modeling Agent-based modeling

9

Toward configurative theory

• A social form is a configuration of social

interactions and actors together with the structures in which they are embedded.

• Variables merely measure dimensions of

social forms, not the forms themselves.

• Social forms always have an extension in

time and (geographic/abstract) space

10

The epistemology of computational modeling

Computational Modeling Sociological Process Theory Epistemology Ontology Nomothetic theory Generative theory Nomothetic modeling Generative modeling Variable-based theory Configurative theory Variable-based modeling Agent-based modeling

11

Toward generative modeling

• In Micromotives and Macrobehavior,

Schelling invites the reader “to try to figure

• ut what intentions, or modes of behavior,
• f separate individuals could lead to the

pattern we observed.”

• Axelrod: “a third way of doing science”
• Epstein: “If you didn’t grow it, you didn’t

explain it.”

12

Emergence

• Can be exhibited by generative processes
• Emergent properties stem from the

irreducibility of complex systems to the properties of their constituent parts

• Usually due to non-linear interactions

among system components

• Bottom-up vs. intrinsic emergence

13

The ontology of computational modeling

Computational Modeling Sociological Process Theory Epistemology Ontology Nomothetic theory Generative theory Nomothetic modeling Generative modeling Variable-based theory Configurative theory Variable-based modeling Agent-based modeling

14

Toward agent-based modeling

• ABM features explicit representations of

social forms.

• Whereas variable-based modeling

expresses causal relations among variables, ABM represents interactions among the actors directly.

• Object-orientation facilitates the task of

representing social forms.

15

Modeling Emergence

Four types of endogenous social forms:

• Behavioral interaction configurations
• Property configurations
• Interactive networks
• Actor structures

16

Emergent interaction patterns

actor actor actor actor actor actor actor actor actor

• Models of “emergent
• rder” producing

configurations

• Axelrod (1984, chap. 8):

“The structure of cooperation”

17

Emergent property configurations

• Models of “emergent

structure” constituted as property configruations

• Example: Schelling’s

segregation model; Carley 1991; Axelrod 1997

• See Macy & Willer 2002

for further references

actor actor actor actor actor actor actor actor actor actor actor actor actor actor actor actor actor actor

18

Emergent networks

• Most computational

models treat networks as exogenous

• Recent exceptions:

– Albert and Barabási’s scale- free networks – Economics and evolutionary game theory: e.g. Skyrms and Pemantle

frequency degree d d-α

19

Emergent actors

• Computational models

normally assume the actors to be given

• Some exceptions:

– Axelrod’s model of new political actors – Axtell’s firm-size model – Geopolitical models in the Bremer & Mihalka tradition

• Emergence?

20

Conclusions

• ABM follows in the epistemological and
• ntological footsteps of sociological

process theory

• Future challenges of ABM: endogenization
• f networks and actor structures
• Future challenges of process theory: clearer

specification of generative theory that allows for systematic empirical anchoring.

21

The Luhmann Simulator

• Prof. Dr. Uwe

Schimank Fern Universität Hagen