Thinking Systematically: Thomas and Agatha Hughes, Systems, Experts, and Computers

Judging from the 2001 promotional pamphlet of the Operations Research Center at MIT, competence in systems analysis is an essential component in the training of the present-day manager. The ubiquity of this attitude toward systems thinking demands an understanding of the origins and importance of this phenomenon. As historians of science approaching this field from the outside, we see great promise in the historiographical questions that the study of the origins and development of systems thinking raises. Central to these are issues of narrative, agency, and causation. Fortunately, several historians of technology have begun to explore these areas anew, and we would like to take this opportunity to reflect on one important such attempt, the recent volume edited by Thomas and Agatha Hughes, Systems, Experts and Computers. ¹

Historiography, like the construction and operation of technological systems, involves the marshaling of heterogeneous components into a working whole. Thomas Hughes and his late wife Agatha assembled in this volume a heterogeneous group of scholars who take diverse approaches to examine the complicated trajectory of systems and systems thinking from World War II to the 1970s. The fruits of a conference held at the Dibner Institute in Cambridge, Massachusetts, in 1996, Systems, Experts, and Computers sketches the pervasiveness of systems thinking both within and without the engineering sciences during that period.

As is frequently the case when dealing with a heterogeneous concept analyzed with great historical acumen, the "systems approach" is rarely explicit in the volume. The Hugheses, in their introduction, define it as follows: "Practitioners and proponents embrace a holistic vision. They focus on the interconnections among subsystems and components, taking special note of the interfaces among the various parts. What is significant is that system builders include heterogeneous components, such as mechanical, electrical, and organizational parts, in a single system" (2). What is meant by a "system" is fairly clear: a system is built to carry out definite, purposeful sets of actions and conceived as a collection of interrelated heterogeneous components, each of which interacts directly or indirectly with every other component, with no subset of components uncoupled from any other. And, as emphasized by the Hugheses, systems designers and builders, instead of conceptualizing the whole by assembling analyses of the parts, develop their conceptions of the heterogeneous parts by disjoining their conception of the whole. ²

By the late 1950s, although one could point to numerous examples of systems, the subject matter of "systems science" was never exactly defined, nor was there a consensus among its practitioners as to what it should be. It was an outgrowth of Norbert Weiner's cybernetics that dealt with the art of control. Early on, cybernetics split into two parts: automatic control theory and its applications, and systems science. But whereas control theory was a fairly well defined subject, a heterogeneous assembly of subjects came to be subsumed under systems science, ranging from information theory to the mathematics of fuzzy sets and problems of productions. Nonetheless, even though its subject matter was not well defined the aims of systems science were fairly well delineated. The hope was to formulate a general theory of structures that would encompass such heterogeneous components as engineering systems, biological systems, urban systems, societal systems, and so on, and combine it with a suitable theory of control.

The first problem to be resolved in the pursuit of this goal was to identify the common features (if any) of these different systems and formulate a mathematical description in terms of the variables assigned to these features. That aim may well have been too ambitious. The second was to describe efficiently the nature of the interactions between the constituents that make up the system. Broadly speaking, these can be characterized as "energetic" and "informational." Under energetic interactions are subsumed those that are amenable to description in terms of physico-chemical processes (mechanical, electromagnetic, acoustic, and so on...