Complex Technical Systems (review)

Some reviews assess the merit of a colleague’s work for an audience of peers; others appreciate a book’s status as historical artifact. My review of Complex Technical Systems, a collection of papers presented at a conference at Sigtuna in December 1994, belongs to the latter category. Although no historians participated in the event, the topics covered there will interest historians of technology and industrial policy.

Sponsored jointly by the Swedish Council for Planning and Coordination of Research (FRN), the Swedish National Board for Industrial and Technical Development (with the pithy acronym NUTEK), and Linköping University’s Department of Technology and Social Change, the Sigtuna conference explored the nature of complexity and system building. A deindustrializing, post-Maastricht Europe provides the largely unacknowledged backdrop here. In the face of the social and economic upheavals unleashed by the end of the Cold War, the FRN and NUTEK appear hopeful that interdisciplinary systems analyses and planning methods may help ensure Sweden’s place in the new world order. However, few contributors to this volume (all but two of whom are Swedes) share their optimism.

That only three of the ten substantive papers muster much enthusiasm for systems thinking is significant in itself. The optimists address three major pitfalls of systems thinking. The first is the philosophical tension between holism and reductionism intrinsic to all systems approaches. For example, Anders Törne, a computer scientist, believes that a “complex system can only be comprehended by humans through decomposition and abstraction” (p. 114) (i.e., reductionism), whereupon software can be designed in components. Putting Humpty Dumpty together (i.e., holism) is admittedly more problematic: “only the simplest possible synthesis tools” are available to assemble and test programs. Ultimately, the software designer must revert to empirical testing. Anders Karlqvist, a systems analyst at the Royal Institute of Technology also associated with the Santa Fe Institute, is rather more cheery about this empiricism. After sharply criticizing Newtonian attempts at reducing nature to a set of unified laws, Karlqvist recommends the use of chaos theory to resolve complexity. Of course, chaos theory merely exchanges Newton’s reductionism for Heisenberg’s: instead of relying on scientific laws to perform the (top-down) synthesis, Karlqvist recommends the (bottom-up) empirical approach of mathematical modeling. All that limits us from conquering complexity, Karlqvist suggests, is our computer technology.

The second pitfall of systems thinking is its undisciplined use of metaphor and its justification through the unceasing evocation of “complexity” and novelty. As social scientist Bernward Joerges points out, “there is no clear connection between constructions of complexity in the engineering sciences, in social systems theory or in evolutionary biology” (p. 69). Joerges deftly dissects the “discourse of complexity,” identifying weaknesses in boundary definition and conceptualization. He notes that the utility of the Hughesian systems approach to social scientists stems from Thomas Hughes’s descriptive use of systems terminology to historicize and elucidate the thinking of systems builders. The current confusion among social scientists, Joerges charges, arises from their attempts to use this terminology analytically.

The third pitfall of systems thinking is its historical tendency toward undemocratic—and even antidemocratic—centrism. Too often the language of systems has aided and abetted expertise in hijacking public discourse. The legitimacy of interests is displaced; to systems thinkers, conflicts are the result of incomplete information or misunderstandings. Thus mechanical engineer Björn Karlsson outlines an elegant linear programming model of Sweden’s combined heat and power production, distribution, and consumption. Within the system, Karlsson assumes a “perfect market . . . in the sense that everybody is aware of the cost of satisfying the energy demand and takes any possible measure at any time to reduce the total cost” (p. 76). But as sociologist Jane Summerton shows, the “perfect market” assumption conceals a pitched political battle over regulation, concentration, and international collusion. Fellow social scientists Lars Ingelstam, Torsten Hägerstrand, and Langdon Winner issue similar, if more philosophical, warnings. By no means do the social scientists monopolize such criticisms...