COMPLEXITY AND SOCIAL SCIENCES: POSSIBLE METHODOLOGYCAL APPROACHES
1 like296 views
This document discusses potential methodological approaches for complexity sciences and social sciences, including mathematical modeling, computer simulation, and social measurement techniques. It outlines some key features of complex systems like emergence and nonlinearity. Challenges are noted, such as the difficulty of obtaining large, reliable time series data for social systems and the inability of static models to capture changing social structures. The document advocates developing adaptive simulation models to explore possible social dynamics while acknowledging they are simplifications of real systems.
More Related Content
Similar to COMPLEXITY AND SOCIAL SCIENCES: POSSIBLE METHODOLOGYCAL APPROACHES (20)
COMPLEXITY AND SOCIAL SCIENCES: POSSIBLE METHODOLOGYCAL APPROACHES
- 1. COMPLEXITY AND SOCIAL SCIENCES: POSSIBLE METHODOLOGYCAL APPROACHES Herman Resende Santos - MSc
- 3. ComplexityScienceandtheSocialWorld- WillMedd EncyclopediaofSocialMeasurement(2005) • ‘‘understanding the emergent patterns of system behavior and in developing models that can be applied to a whole host of phenomena” • ‘‘complexity social scientists’’ “want to show how social science can contribute to complexity science and how complexity science, applied and developed for the social world, can illuminate real processes of social dynamics”. 3
- 4. Complex systems key features • 1. Self-organized emergence • 2. Nonlinearity • 3. Self-organized criticality • 4. Evolutionary dynamics • 5. Attractor states - trajectories that tend to be restricted to (attracted to) particular parts of phase space (abstract multidimensional space with coordinates for all the variables of the system) • 6. Interdependent systems - patterns of system dynamics which are repeated through different scales (systems within systems) 4
- 5. Social measurement Quali • ethnographic • participant- or non-participant observation • content analysis Quanti • Mathematics • Simulation 5 bifurcation diagrams non-linear modeling computational models genetic programming cellular automata programming distributed artificial intelligence evolutionary game theory parallel distributed cognition artificial life network analysis sociocybernetics
- 6. Mathematical tools 1° General equilibrium theories – Equations to define a dynamical system - Systems of differential equations describing the behavior of simple homogenous social actors. Change occurs as a result of perturbations and leads from one equilibrium state to another. • Weakness: the assumption of equilibrium is forced by the mathematics, not by the observation of social behavior • 2° Statistical theory - indirect method for discovering causal relationships and observe regularities in social phenomena 6
- 7. Computer technology - solve nonlinear equations numerically (by trial and error) rather than analytically (finding a formula to a solution). Objective: • to examine graphic representations of time series data in order to expose ‘‘attractors,’’ or forms of underlying order, pattern, and structure. Phase space • point attractors - stable equilibrium • cyclical attractors - stable periodic oscillation between the same points • strange attractors - figure-eight pattern. 7 Nonlinear Mathematics
- 8. Three particular difficulties • 1° - such analysis requires large amounts of time series data that are consistent and reliable, which is difficult to achieve in the social sciences • 2º - some of the emergent patterns are as much a function of the nonlinear equation as they are of the particular data being explored • 3° - equations assumed deterministic relationships in their structure, meaning that the structure of the equation does not evolve; this is problematic for social analysis because underlying social structures may change in ways that the equations supposedly representing them cannot • Models could not provide objective viewpoints of the social world, and would therefore become by necessity tools in the social world 8
- 9. Simulation • Artificial societies - specialized type of simulation model that typically employs an object-oriented, agent-based system architecture. • Interests - self-organized patterns that emerge in running the simulation, including the emergence of individual behaviors of the agents. 9
- 10. Critics • Simulations are always too simplistic —simulation requires a simplification of the agents to rule-based behaviors, of the agents interacting in particular ways, and of what is seen as important in the environment (what is important?) • Simulations - may fail to capture a key aspect of the social world (human beings may change their very behavior in unpredictable ways as a consequence of the emergent system) Simulation does not claim to be the real system, but instead provides a method for exploring possible dynamics that might be looked for in the real system 10
- 11. • Nonlinear mathematical models to the social world - establishing determinate sets of relations to represent complex social interactions • Developing simulation models -If it is the case that those boundaries are dynamic - continued adjustment / continually - adaptive to changing internal and external environments. Challenges 11
- 12. October, 2015 Thanks! Herman Resende Santos - MSc [email protected]