Page 24 -

2 Historical Introduction 17

“the future of the social sciences is contingent upon identifying techniques to
simultaneously link a multitude of relatively trivial conceptual structures, producing
realistic outcomes when no premise alone is powerful enough to determine the state
of the system at any moment” (Federico et al. 1981, p. 518). This is certainly a
prediction which came true in the decades to come, as agent-based modelling in
its various modern approaches is more or less correctly described with Federico’s
and Figliozzi’s words. Nevertheless their “classification of computer simulation
studies of psychosocial or sociotechnical systems” (Federico et al. 1981, p. 515)
with its double dichotomy of operational and theoretical nature and micro and
macro scope is no longer in line with current classifications. Putting, for instance,
Abelson’s and Bernstein’s study (see above for details) in the box of operational
(as contrasted to theoretical) macro simulation studies seems strange as this study
connects microbehaviour to macrostructures and does not only look at the macro
level. The same is true for other studies that fall in this cell of Federico’s and
Figliozzi’s cross table. The reason for this is that what they call “microtheoretical
computer simulation studies” is restricted to behaviour in small groups, thus “micro”
does no refer to the individual level of social systems (as it usually does today) but
to small systems such as Hare’s (1961) five person group.

2.3 Computer Simulation in Its Own Right

The Simulmatics Corporation already mentioned in the previous subsection did not
only work in the context of election campaigning, but later on also as a consulting
agency in other political fields. Crisiscom is another example of an early forerunner
of current simulation models of negotiation and decision-making processes. At the
same time, it is an early example of a simulation not aimed at prediction but at “our
understanding of the process of deterrence by exploring how far the behaviour of
political decision makers in crisis can be explained by psychological mechanisms”
(Ithiel de Pool and Kessler 1965, p. 31). Crisiscom dealt with messages of the type
“actor one is related to actor two”, where the set of relations was restricted to just
two relations: affect and salience. In some way, Crisiscom could also be used as
part of a gaming simulation in which one or more of the actors were represented by
human players, whereas the others were represented by the computer programme—
thus in a way it can also be classified as a predecessor of participatory simulation
(see Chap. 11).
The 1970s and 1980s saw a number of new approaches to simulate abstract
social processes, and most of them now were computer simulations in its own right,
as—in terms of Thomas Ostrom—they used the “third symbol system” (Ostrom
1988, p. 384) directly without using it as a machine to manipulate symbols of the
second symbol system, mathematics, but directly translating their ideas from the
first symbol system, natural language, into higher level programming languages.
Although this was already true for Herbert Simon’s Logic Theorist, the General
Problem Solver and other early artificial intelligence programmes, the direct use of