Transcript second order science: logic, strategies, methods

SECOND ORDER SCIENCE:
LOGIC, STRATEGIES, METHODS
Stuart A. Umpleby
Department of Management
The George Washington University
Washington, DC
Context
• Hundreds of years of scientific research and
results
• What have we learned?
• How can we improve the production of new
knowledge?
The basic idea
• Science can be expanded by adding the
observer to what is observed
• We could shift our thinking from viewing
science as creating descriptions of systems to
viewing science as an active part of social
systems
• We would think about the co-evolution of
theories and society
How we think about social science
• It is easy to accept that theories affect society
• We create theories because we hope that
people will accept them, act on them and
society will operate better
• However, when we do science, we pretend
that theories do not affect what is studied
• We act this way because we think this is the
way to do science
• But we can expand our conception of science
Three conceptions of Science 2
• Meta-science: combining the results of many
studies in order to create more holistic, more
useable knowledge
• Take advantage of the internet to cooperate in
scientific research and sharing of data
• Take account of the observer and examine the
co-evolution of theories and phenomena,
particularly in the social sciences
Four models underlying science
• Linear causality – many statistical methods are
available, e.g., correlation and multiple
regression
• Circular causality – inherent in regulation,
system dynamics models
• Self-organization – Adam Smith, Charles
Darwin, complex adaptive systems
• Reflexivity – von Foerster, Soros, Lefebvre
Examples from several fields
•
•
•
•
•
Management
Political Science
Sociology
Economics
Science policy
Including the observer is not
completely new
• The field of management is largely a collection
of methods
– How to improve a manufacturing process
– How to conduct a strategic planning exercise
• But little attention has been paid to
connecting this form of knowledge to the
philosophy of science
• Reflexivity theory makes this possible
Second order knowledge in
management
• In management there is a large literature on
how to do things, on how a manager should
act
• The most useful literature has often been
created by consultants rather than academics
• But most professors of management attempt
to emulate the physical sciences
• Often people in business say that academic
studies of business are not particularly helpful
Should knowledge in the field of management be
constructed in the form of theories or methods?
Theories
Is there a difference between the
natural sciences and the social
sciences?
Yes
No
Popper’s doctrine of
the unity of method
Should we reject the
philosophy of science?
Yes
What should take its place?
How should knowledge be
constructed?
No
Expand the philosophy of
science to include knowing
subjects
Methods
Should methods be for the use of
individuals or groups?
Individuals
Groups
“Think like this”
“Act like this”
The nature of knowledge in
management
• Management methods are inherently reflexive
• Managers observe what is happening, decide
what action is needed, participate in the
organization to implement the decision,
observe the effects of the decision, etc.
• But most dissertations in management seek to
find linear relationships between variables
Consider political science
• People have a tendency to develop an idea
and then believe that it will solve all problems
• Socialism, capitalism, market fundamentalism
• An alternative is to view “reforms as
experiments”
• Accept fallibility of our knowledge
• Try an idea or an institution and evaluate it
(e.g., Campbell’s “experimenting society”)
Sociology of knowledge
• Our position in society influences what we
believe
• If we are rich, society is working well
• If we are poor, society is not working well
• Culture, religion, and educational background
all shape what we see and how we think
• Hence, there is an interaction between society
and what we think
The nature of knowledge in
economics
• If management is a field in which useful
knowledge is inherently reflexive, economics
is the social science field that has been most
successful in imitating classical science
• Quantification has been easier in economics
than other social sciences
• Economics is now defined primarily by its
methods rather than its subject matter
Equilibrium theory in economics
• Physicists at first dealt with small numbers of
objects – planets, pendulla, billiard balls
• But gases involve millions of particles
• So, to create thermodynamics physicists
looked at gross parameters – pressure,
volume, temperature
• Economists adopted this model – imports,
exports, savings, GDP
Behavioral economics
• Equilibrium theory in economics uses the
concept of self-organization – elements of a
system go to their stable, equilibrial states
• When there is a disturbance (e.g., a new
technology or new legislation), people seek to
return to equilibrium
• Several assumptions – people are rational;
they seek to maximize personal profit; they
have equal and complete information
Progress in recent years
• Behavioral economics has challenged the
assumptions underlying equilibrium theory
• Several people have won Nobel Prizes by
showing that people are not always rational,
that they do not always seek to maximize
profit, that they do not have equal
information, and they do not have complete
information
Additional assumptions in economics
• There is another, more philosophical, set of
assumptions that economists have not yet
addressed
• The classical philosophy of science assumed
– The observer is not included in what is observed
– Theories do not affect what is observed
• Do these assumptions fit the social sciences,
or do we create theories in an effort to change
social systems?
Effects of the financial crisis
• We are returning to Keynesian economics to
solve problems created by over-reliance on
markets
• Reliance on markets, begun by Thatcher and
Reagan, has swung back to a belief that
markets are imperfect
• Swings between these two theories act like a
control system for the economy
The loss of history in economics
• The history of economic thought is no longer
taught in economics curricula
• “There is not time enough to teach the old
ideas, since there is so much new stuff to
teach”
• The old theories, which were verbal rather
than quantitative, are seen as inferior to
newer ideas
Need for second order theory
• Current economic theories are elements in the
regulation of an economy
• A second order theory is needed to describe
how first order theories are to be used
• A current economic theory can be viewed as a
partial description or a temporary description
or a context dependent description
• A second order theory of economics would be
a cybernetic theory, a theory of control by
using ideas
Correcting underlying assumptions
• Behavioral economics is an effort to correct
specific assumptions underlying equilibrium
theory or the idea of an economy as a selforganizing system
• Reflexivity theory challenges assumptions
about science -- that theories do not affect the
operation of an economy and that older
theories can be discarded
Second order science
• The idea of second order science is more
relevant for the social sciences than the
physical sciences
• But science itself is a social process
• And even physical theories are used within a
social context
• Cybernetics is a theory of information and
regulation in somewhat the same way that
physics is a theory of matter and energy
How to do second order science
• An academic discipline requires methods
which can be practiced
• What are some methods for doing second
order science?
Several methods for second order
science
• The influence cycle is similar to the credit
cycle and the leverage cycle
• Interaction between ideas and society, an
historical approach
• Examine social activities in terms of ideas,
groups, events, and variables
• A comparative approach
• An historical approach
Amount of
credit
+
Lending
activity
+
+
+
+
Collateral Debt
values
service
Economic +
stimulus
_
The credit cycle
-
Lending
activity
+
Amount of down-payment
required
+
Collateral Debt
values
service
Economic +
stimulus
_
The leverage cycle
+
opportunities for
innovation
+
deregulation of +
economic
business activity
growth
+
belief in the +
value of markets
_
The influence cycle
+
abuse of asymmetric
information
Ideas
Variables
Groups
Events
A reflexive theory operates at two levels
Other social sciences
• Political science has comparative politics
• Sociology has the sociology of knowledge
• Management science has a variety of
management methods
• Knowledge can be structured as methods as
well as theories
Why practice second order science?
• It is more realistic
• We cease pretending that social systems are
like physical systems
• New scientific theories and methods will be
invented
• With a larger conception of science we will be
able to explain and to influence processes we
could not otherwise
Reasons not to do second order
science
• Giving up a claim of objectivity means giving
up a claim to special authority
• The informal fallacies caution against errors in
language, emotion and thought
– Circular reasoning, language
– Ad hominem fallacy, emotion
– Fallacy of accent, thought
• Self-reference leads to paradox and hence
logical inconsistency
Coping with self-reference
• In practical affairs we routinely deal with the
logical problem of self-reference
• The Fifth Amendment prohibits selfincrimination
• We let juries decide whom and what to
believe
• We seek to eliminate conflicts of interest
• When we do science we try to construct
logically consistent theories
Why should we not be deterred?
• In practical affairs we have learned to cope
with self-reference
• We commonly see ourselves as participants in
a multi-player game
• We lose nothing by adding a new dimension
to science
• The science we practiced before we can still
practice
The Correspondence Principle
• Proposed by Niels Bohr when developing the
quantum theory
• Any new theory should reduce to the old
theory to which it corresponds for those cases
in which the old theory is known to hold
• A new dimension is required
• By applying the correspondence to the
philosophy of science, we change not just one
theory or one field but all of science
New philosophy of science
Old philosophy of science
Amount of attention paid to the
observer
An Application of the Correspondence Principle
Acknowledging fallability
• When we do science, we pretend we are
objective observers
• Social scientists seek to claim the authority of
science
• We could instead claim the authority of selfawareness and say that science is a method
for achieving agreement
• We would need to rely more on judgment and
less on statistics
The effect on science policy
• Science policy today is mostly an exercise in
budgeting
• An alternative approach is to study the history
of science (an intellectual exposition)
• Study science as a social activity – history,
philosophy, cognitive science, sociology of
science.
• What practices have been helpful? How has
the most innovative science been produced?
Logic
• Include the observer in what is observed
• In the social sciences theories can change the
phenomenon observed
• Knowledge can be organized as methods as
well as theories (instructions to participants)
• Add the dimension of time to resolve
problems involving self-reference
Strategies
• Study cognition and incorporate what is
learned into our understanding of knowledge
and epistemology (Heinz von Foerster)
• Study high performing research teams
(Hollingsworth, Mueller)
• Create methods for doing second order
science
Methods
• Model cycles in preferred theories
• View first order theories as being part of social
control processes
• Chart the interaction between ideas and
society (e.g., shoelace model)
• Create multi-disciplinary descriptions using
ideas, groups, events, and variables
• Do cross-cultural studies
A key question: Do human activities
change systems?
• Human beings change social systems by
changing laws and theories
• As technology improves, human beings are
even changing the natural environment – soil,
fish, climate
• We are learning to think about ourselves as
participants in the systems we study
• But to do that we need to change our
conception of science
Contact Information
Prof. Stuart Umpleby
Department of Management
School of Business
George Washington University
Washington, DC 20052 USA
www.gwu.edu/~umpleby
[email protected]
Presented at the World Multi-conference on
Systemics, Cybernetics and Informatics
Orlando, Florida
July 19-22, 2011