1 of 15

ERIC D. BEINHOCKER (2013)

Reflexivity, complexity, and the nature

of social science

GIVI VESHAPELI

2 of 15

Introduction

I, however, believe that there is at least one philosophical problem in which all thinking men

are interested ... the problem of understanding the world – including ourselves, and our

knowledge, as part of the world.

– Karl Popper, The Logic of Scientific Discovery

3 of 15

Introduction

I, however, believe that there is at least one philosophical problem in which all thinking men

are interested ... the problem of understanding the world – including ourselves, and our

knowledge, as part of the world.

– Karl Popper, The Logic of Scientific Discovery

However,

4 of 15

Introduction

I, however, believe that there is at least one philosophical problem in which all thinking men

are interested ... the problem of understanding the world – including ourselves, and our

knowledge, as part of the world.

– Karl Popper, The Logic of Scientific Discovery

However,

Popper was wrong about one thing – not all ‘thinking men’ seem

to be interested in this problem

Example: Economists

5 of 15

    • A Process where an observer is also a participant in a system and there is a two-way feedback between the participant/observer and the system
    • In order for a system to be “reflexive” it must have the following six elements

​

    • Enviroment
    • Agent
    • Goal
    • Cognitive function
    • Manipulative function
    • Internal model

​

​

​

What are reflexive systems?

6 of 15

    • A Process where an observer is also a participant in a system and there is a two-way feedback between the participant/observer and the system
    • In order for a system to be “reflexive” it must have the following six elements

​

    • Enviroment
    • Agent
    • Goal
    • Cognitive function
    • Manipulative function
    • Internal model

​

​

​

Necessary, not sufficient

What are reflexive systems?

7 of 15

    • Internal decision model of the agents are not fixed and can change in response to the interaction between the enviroment

The system has:

    • interactive complexity due to multiple interactions between heterogeneous agents
    • dynamic complexity due to nonlinearity in feedbacks in the system

Internal model updating

Complexity

Distinguishing characteristics: internal model updating and complexity

8 of 15

    • Constructing an accurate internal model which will enable an agent to successfully pursue its goals runs into fundamental limits to knowledge issues

​

    • Difficulty discovering the correct model from finite data
    • Lack of knowledge of initial conditions and parameters
    • Inability to predict with finite computing time
    • No ‘God’s-eye view

Limits to knowledge and fallibility

almost any real-world economic situation – the true underlying model will not be given a priori to the agents, knowledge of initial conditions and parameters will be limited, data will be finite and noisy, computing power will be finite, and the agents doing the observing will also be participants in the system

9 of 15

    • Reflexive systems are inherently contingent and time bound, they do not have timeless universal laws, and one cannot make reliable predictions in such systems.
    • By reflexive system we mean not only social, but natural sciences as well (e.g. existence of atoms after big bang; time, evolution)
    • The goal of science is not the discovery of timeless, universal laws – they don’t exist. Rather, it is to explain time bound and contingent regularities in an empirically testable way.
    • Reflexivity destroys the ability to make “determinate predictions”

​

​

​

Implications for social science

10 of 15

A spectrum of complexity

11 of 15

What defines the epistemological challenge of understanding particular phenomenon?

A spectrum of complexity

12 of 15

    • fallible models, limits to knowledge, and the inherent indeterminacy of reflexive systems do not necessarily imply that all is lost and that we cannot gain insight into such systems.

Example from Biological, Human and Artificial systems: They all manage to fuction despite fallibility

Evolution, good enough models, and muddling through

“Good enough” approach

‘Essentially, all models are wrong, but some are useful’ (Box, 1987, p. 424).

13 of 15

Model-dependent realism: reconciling Soros and Popper

    • Reflexivity is not only consistent with an evolutionary view of epistemology, but with so called ‘model-dependent realism’
    • Clear boundaries of where should the discussion stop (does it matter?)
    • Agents do not have direct access to any objective reality – there is no ideal of the detached, objective observer

​

​

​

​

14 of 15

Thank you

15 of 15