Review: Turkle et al, Simulation and its Discontents

Book review: Sherry Turkle et al, Simulation and its Discontents (Cambridge MA: MIT Press, 2009).

In what promises to be an important book, Sherry Turkle examines the growth of computer-aided simulation in education and practice—and the possible limitations, shortcomings, and costs of that ascendency. Her focus is almost entirely the fields of design and natural science, with architecture, biology, chemistry, and physics accounting for the bulk of her cases. In discussions with scholars in the field, Turkle finds a disquietening sense among many that something has been lost. Simulation and virtual modeling, while an enormous boon to design and scientific research, may also be corrosive of indepndent and original thought, conceal built-in biases and distortions, and risk confusing the virtual realities with the real thing.

In architecture, for example, computer-aided design has facilitated the production of building plans, but with a tendency to use the default options provided by software packages and a consequent risk of loss of imagination. Some younger designers and architects, her respondents warn, may also have lost touch (literally) with the very materials out of which their creations are built, erroneously assuming that the virtual version is a full and faithful representation of how the material functions, looks, and feels in the real world.  Similarly, scientists interviewed by Turkle warn that students may become so dependent on software to construct virtual proteins, molecules, or other models of reality that they fail to question the assumptions built into the modeling software, and the implications for this for the validity of what they are doing.

Unfortunately, while full of fascinating reflections, the book never fully delivers on its promise. While its points are important, they are made repeatedly, and often with the same repetitive examples. There is little effort to branch out beyond her primary case studies, or to explore the broader social (as opposed to technical and scientific) implications of the issues raised. Moreover, Turkle’s primary contribution is remarkably short, a mere 101 pages. It is followed by four case studies by other authors. Two of these—on computer-aided design in an architectural firm (Yanni Loukissas), and on computerized modeling of protein folds (Natasha Myers)—hew a closer to the theme of the book. The other two cases, however, really examine remote sensing more than they do with the issue of simulation. Of course, there are some connections. A robotic rover on Mars (William Clancey), or a remotely-controlled submersible in the ocean (Stefan Helmreich) are rendering representations of reality to their operators through cameras and sensors , which only approximate their actual environment. In general, however, this seems to be an interesting but rather different set of issues. All human perception, after all, is representation, with our own senses transmitting their own biologically-limited perception of the environment around us.

Can one extrapolate from the issues raised in this volume to the field of social simulation, especially as it relates to questions of development, crisis management, and conflict resolution/peacebuilding? There are, of course, some important differences. Much of the simulation done in this field, such as the World Bank’s Carana simulation, or UNHCR staff training exercises, are human-moderated. This may alter the participants’ experiences in important ways. Social realities are complex and the relationship between variables are poorly understood, which—to the extent this is understood a priori by participants—perhaps provides a degree of inoculation against overconfidence in simulation output or findings. (Of course, the reverse is also true: those running simulations may have difficulty convincing experienced participants that such an gross abstraction from reality is a useful learning tool.)

Nonetheless, much of what Turkle points to does have its echoes in the social sciences too. Computerized simulations may be especially problematic, in that they “hide behind the hood” critical assumptions and simplifications about the way the world works, while usually requiring players to select from a preset menu of policy responses. As they grow in audiovisual sophistication, moreover, computer-based simulations might also grow more alluring without a corresponding improvment in the human-designed theoretical presuppositions that underpin them. Even absent the sophistication of computer-mediation simulation, however, one needs to be careful about “lessons” that are derived more as an artifact of simulation dynamics than of the real world that they are meant to approximate. Many international relations simulations, for example, seem to implicitly privilege realist models of politics and high politics/military interaction because of the competitive frame of participants, the absence of real world constratints and socialization, and the difficulties of including the many complex process of low politics within a playable, time-limited game. Practicality and “realism” therefore run at cross purposes, and in ways that may teach unintended and inappropriate lessons.

Because of this, post-simulation debriefs become essential, since they allow an opportunity to discuss where the simulation exercise and real life might differ, and why that is so. Indeed, such episodes can become a teaching lesson in and of themselves, encouraging students to sharpen their critical thinking skills as they reflect on their simulated experiences.