Reviewed by
Hans de Haan
DRIFT-Dutch Research Institute for Transitions, Faculty of Social Sciences, Erasmus University Rotterdam
The title "Simulation: Pragmatic Construction of Reality" suggests that the editors were ready to take on a task of reviewing simulation at large, and in fact this is what they did. In this edited book, which is the 25th Sociology of the Sciences Yearbook, the state and status of simulation in the broadest sense are explored in twelve chapters. The emergent question that is raised and addressed by this book is if simulation has grown to become a mode of scientific inquiry in itself rather than merely being a tool in existing scientific practice. Although many of the contributors to this volume have a broad conception of what simulation is, in general computer simulation is meant. And, as in the last decades the possibilities of computation in science have grown enormously, this emergent question is very relevant. After an introductory chapter this book is divided in three main parts. The first part called "Imitating Models" discusses the most familiar aspect of simulation, actual models in four chapters. In the second part "Layers of Integration" again four chapters treat composite models in many guises. And the last part "Social Practice" features three chapters that address the form and role the process of modelling takes in several contexts. Let us skim through these three parts and see the question emerge.
In the first part the reader is confronted with chapters on modelling in such diverse fields as nanotechnology (The Shape of Molecules to Come, by Johnson), linguistics (From Representation to Production, by Knuuttila), ecology (Foundations for the Simulation of Ecosystems, by Hauhs and Lange) supplemented with a chapter on how the ubiquity of modelling changes our way of interpreting reality (Models, Models Everywhere, by Ihde). In these chapters the image arises of simulation as a method changing the way the scientist approaches his subject. In the chapter on nanotechnology, for instance, is pointed out how the in silico experiment leads the traditional experimentation, as opposed to the simulation as a check with theory of the times of yore. Especially the chapter on linguistics is interesting if one is to appreciate the transition that simulation apparently is going through. The discussion of a grammar parser as a model that is not representing but rather producing a reality to be interpreted, is the basis of a thorough philosophical discussion on the epistemology of simulation. This in contrast to the chapter on modelling in ecology, where the authors advocate an interactive modelling paradigm and seek to point out that philosophy of science ought to take a closer look at this.
The second part is all about composite models and the challenges that integration of models pose. The first chapter discuses how model integration moved from hierarchical to network type integration and how meteorology and climate research took the lead in this (From Hierarchical to Network-Like Integration, by Küppers and Lenhard). In the next chapter model integration is treated from the point of view of how to interpret these composite models (The Difference Between Answering a 'Why' Question and a 'How Much' Question, by Boumans). The author analyses this challenge in terms of black, gray and white box models. The next chapter discusses the problems in coming to a composite model (Struggle Between Specificity and Generality, by Mattila), by discussing the development of a simulation platform for infectious disease models. In my opinion this article had better be placed in the last part on social practice. The last chapter in this part discussed how the integration of different levels of scale can be dealt with (Handshaking Your Way to the Top, by Winsberg). The example in this chapter is again nanoscience and clearly exposes how such problems can be overcome by means of handshaking algorithms, which is especially interesting since the author treats in a clear way how the underlying theories that need to be connected all present their problems for such algorithms.
In the last part the reader arrives at sociology of the sciences proper. Three chapters treat the social practice of simulation. By studying the case of particle accelerator physics, where simulation is a valued companion of the physical experiment, the first chapter tries to identify the role of simulation in actual laboratories (Locating the Dry Lab on the Lab Map, by Merz). When stakes are high and uncertainty is great doing science is postnormal, I believe the adagium was. The challenges for simulation and modelling in such settings are explored in the next chapter (Simulation Uncertainty and the Challenge of Postnormal Science, by Petersen). The closing chapter of this book addresses simulation as a research technology and uses the context of the emergence of the C++ language to do so by exploiting the parallels of the use of this simulation language and simulation as such.
I stated in the beginning that this book raises an emergent question. The question whether simulation is becoming a mode of scientific inquiry in itself. The question is implicitly raised and the reader is left with the question itself rather than an answer to it. Although the book features some interesting chapters I can not help but find it lacking a sort of unity. The editors have clearly tried to provide direction in the division in the aforementioned three parts and have partly succeeded in this. Another problem I felt this book has is that its scope is rather limited, social simulation is left untreated for instance. Another shortcoming is that technical aspects of simulation and their influence on the 'pragmatic construction of reality' hardly enter the picture. How the modelling method (ABMs, system dynamics, etc.) influences the way of interpretation seems to me the missing chapter.
My opinion is that the book left me with some interesting open questions, which is in itself good, but that if one was to purchase this volume one should do so for the individual chapters, since this whole is not much more than the sum of its parts.
Return to Contents
of this issue
© Copyright Journal of Artificial Societies and Social Simulation, 2007