A reply to the whole of Chris Auld's message to the list would be too
long for this forum. I propose therefore to address the overall point
in this message and then his points about the humbug production function
in a following email. Others might I will include in the subject
field "abss or ace" so that those who are not interested in the debate
can filter the contributions to their trash folders.
The essay under discussion was written as a public lecture for (mainly)
members of the management research community. It was intended as a
polemic (though I do not take that as licence to be misleading or
personally abusive). I will take up Chris Auld's claim that I was
misleading in a following email. The claim that I was in effect
personally abusive I address in the context of the point and purpose of
the paper, to which I now turn.
Many in the management research community -- and certainly in my
audience -- conflate "economic modelling" and "mathematical modelling"
which they then reject for its lack of relevance. There is also an
argument in this field to the effect that there is no reality -- this
being a fairly extreme form of critical theory and post modernism. My
objective was to argue that agent based social simulation (ABSS)
modelling is one means of describing social processes, institutions and
how these are influenced/determined by individual behaviour and
interaction among individuals. It is complementary to, for example,
ethnographic descriptions even when the ethnographer is motivated by
critical theory and sees the social institutions as an analogy of a
literary text.
There are, of course, varieties of ABSS. Representational ABSS includes
the Stanford VDT model describing actual industrial design processes by
means of a critical path model and information processing representation
of the organisation; Rouchier's models of "Potlatch" and the relations
among nomads and sedentary populations in the North Cameroon, Doran's
EOS models of palaeolithic social change, the various models of the
Anisazi, my own models of the development of mental models by
entrepreneurs in transition economies or critical incident management in
environmentally sensitive industries. Foundational ABSS is NOT intended
to describe actual institutions or processes but rather to use logical
formalisms to clarify issues that are important in any representational
social analysis such as trust, belief, rights, duties, obligations and
so on.
What is interesting about ABSS and distinguishes it from, for example,
agent based computational economics (ACE), is that a wide range of
techniques is used to represent actors, institutions and social
processes in representational ABSS and these techniques are chosen or
developed by the modeller in light of the systems being described and
the issues being addressed. In ACE, agents are typically represented by
optimising algorithms and, when they are not, some kind of evolutionary
optimising algorithm such as genetic algorithms or replicator dynamics
is used to prove the existence of (or actually compute) an optimal
equilibrium result. Evidently, if it isn't optmisation then it isn't
economics.
In general terms, I have argued for many years that economics, including
its computational variant, ACE, (re)specifies issues and environments to
render them amenable to analysis with a narrow range of techniques. The
social sciences more generally (when they haven't been influenced by
economics), including their computational variant, ABSS, seek
descriptors and analytical techniques suited to the issues and
environments to be studied.
Now for Chris Auld's charge that I made essentially personal attacks on
economists. I think this is based on the sentence from the abstract:
"This third way rejects economics as bad science and intellectually
dishonest". Those who are able to distinguish between social structures
and the individuals who populate societies will also be able to
distinguish between an attack on a set of norms and beliefs and an
attack on the individuals who conform to those norms and hold those
beliefs. Some individuals are more responsible for particular norms and
beliefs than others -- sometimes to the point of culpability. Others
are simply misguided. I claim that the social structures of the
economics community are dysfunctional in the predominance given to
technique over substance and the inability to discard concepts and
models or theories even when they are shown to be logical nonsense in
any social and physical conditions we have ever observed. This
dysfunctionality is what I called bad science and intellectual
dishonesty.
Every use of the word economist or economists was, I believe, clearly
referring to the community of economists. My claim is that, by acting
in accordance with the highest standards of excellence and honesty
within the economics community, the behaviour of economists collectively
is what a reasonable scientist from outside that community would call
bad science and intellectual dishonesty.
Whether others share my assessment will depend, I imagine, on the
evidence presented. Chris' claim that my evidence is misleading will be
taken up in another email.
--
Professor Scott Moss
Director
Centre for Policy Modelling
Manchester Metropolitan University
Aytoun Building
Manchester M1 3GH
UNITED KINGDOM
telephone: +44 (0)161 247 3886
fax: +44 (0)161 247 6802
http://www.cpm.mmu.ac.uk/~scott
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|