[harnden 941110]
as i said in my self-introduction, i am seeking to pursue a thesis
topic that applies pct in an area of my own (newly-minted) expertise:
the modelling of human ecology. the following text is an unformatted
version of a preliminary proposal i've sent to the concerned parties.
i present it here for review. i sincerely hope that i have not
misrepresented the essence of pct, and await all pointed critique
that will clarify my understanding, and guide me towards operating
as an effective representative of the paradigm.
TEXT FOLLOWS:
···
-------------------------------------------------------------------------
Informal Draft Proposal
To Model the Complex Dynamics of
Human Behavior and the Global System
Eric Harnden
American University Physics Dept.
4400 Mass. Ave., NW
Wash., D.C. 20016-8058
Email: harnden@physics.american.edu
Voice: (202) 885-2746
Fax: (202) 885-2723
This document is a preliminary proposal of work to be done toward
completion of the thesis component of the M.A. in International
Development (ID) at the American University s School of International
Service (SIS), in conjunction with an intended internship at the World
Resources Institute (WRI). It is presented for review to Jan Rotmans of
the National Institute of Public Health and Environmental Protection
(RIVM) in Bilthoven, the Netherlands; to Radoslaw Zapert of the World
Resources Institute in Washington, D.C.; and to Prof. John Richardson of
the SIS. Upon completion of review and receipt of commentaries,
including the possible redirection of the basic topic, a final proposal will
be drafted in accordance with the guidelines set forth by the SIS.
The convention in global modeling practice is to disaggregate
purely physical biogeospheric systems from human systems such as
agricultural activity, industrial production, or political policy, whose
variables and boundaries are not as easily defined. Nonetheless, these and
other components of the human sphere are part of the global system. This
is true now more than ever, as the mass of our population, and its attendant
demands on resources, more directly and profoundly alters the physical
makeup of our planet. Further, even our best efforts at global modeling
must almost by definition be fraught with uncertainty. This uncertainty
arises not only from the enormous difficulties in attempting to formulate
and validate an intelligible schematic of something as complex as the
entire Earth - a kind of uncertainty which, it might be argued, remains
largely empirical. It also arises from the fact that the model is
unavoidably a product of our perceptions; of the Earth, of its nature, of our
relations to it, and of the model itself. In effect, the structure of the model
and our interactions with it are maps of our sense of the world, which is
manifested in our behavior towards it. Therefore, in order for a global
model to be complete, so far as that term is valid within the context of an
extraordinarily high-order conceptualization, it must include a component
that models the human factor - the behavioral system which interacts with
the world and changes it, and by so doing, changes itself.
I propose to model such a behavioral system, and to link it to the
TARGETS integrated assessment model as a steering mechanism.
TARGETS, under construction at RIVM, is a global model built on
modules specifically addressing various elements of global change, such
as land use, water cycles, and atmospheric conditions, as well as human
population dynamics, energy demands, and health and welfare indices.
These modules are classified as representing pressures on the system,
dynamic conditions of the system, and observable effects. These layers of
the integrated model are driven by steering models which determine
investment strategies and the use of other policy instruments in response to
the observed effects. My model would operate either as an alternative (or
parallel) steering module, or more likely as a meta-module, driving the
policy determinants and intercepting the observables. In keeping with the
design philosophy at work in TARGETS, the initial model would be
highly aggregated, perhaps representing nothing more than the interaction
of a hypothetical user with the TARGETS system. Further elaboration
would result in a disaggregated model, a hierarchical system in which
larger-scale behavioral features arise as a result of the complex
interactions of many actors.
For the structural basis of this model, I propose to implement
Perceptual Control Theory (PCT), the theory of behavior put forward by
William T. Powers in his book, Behavior, The Control of Perception
(N.Y.: Aldine De Gruyter, 1973). In contrast to behaviorist perspectives
that view action as a response to stimulus (the S-R system), PCT identifies
behavior as the goal-seeking activity of a purposive system. Rather than
the dependent, controlled variable in the loop between organism and
environment, behavior is the independent variable, controlling the
organism s perception of difference between external states of affairs and
internally-held goal states. The theory provides an explicit causal
mechanism for the occurrence of behavior, without recourse to statistical
inference to determine the relations between inputs and outputs.
Moreover, the mechanism is adaptive, as goal states can be modified in the
presence of extreme perceived error - as occurs, for instance, in the
resolution of cognitive dissonance. A further elaboration of the theory,
Hierarchical PCT (HPCT), accounts for reorganization of the internal
feedback structures which perceive and act, under control of higher-order
goal-oriented control systems. Such reorganization is the basis of that
adaptive process known as learning.
The potential adaptability of the behavioral system is of primary
importance in the design of a steering model since, as has been pointed out
elsewhere in the TARGETS project, not only do the perspectives that
people bring to their understanding of the world and the modeling process
differ, but these perspectives change. I hope to show how these
differences and changes can be viewed, not as due to external non-
behavioral processes, but as a result of the act of perception itself. Next, I
expect to demonstrate that the HPCT viewpoint can provide some insight
into how even well-intentioned policy decisions - that is, behaviors which
seek to control the world toward what the actor would presumably think of
as positive ends - lead to perverse results, such as environmental changes
which in fact are contradictory to the behaving system s own best
interests. Finally, I suspect that the linkage of a fully-realized HPCT
model to a world model of the scope of TARGETS will result in complex
dynamic interactions, analysis of which may show that the path toward
sustainability is counterintuitive and to some degree unforseeable, and that
the notion of sustainability itself must account for conditions of
continuous disequilibria.
I expect to have the first stage of this work, a running PCT model
linked to the first, highly aggregated version of TARGETS, completed in
the Fall of 1995, for presentation as the basis of my Master s thesis.
Subsequent elaborations will then proceed if and as directed by WRI and
RIVM. The model will be constructed in the M programming language, in
accordance with the standards and practices of RIVM.
<--- Entropy Always Wins, And I Like To Be On The Winning Side --->
Eric Harnden (Ronin) harnden@physics.american.edu (202)885-2746