Wholesome Design for Wicked Problems

The system theorist Horst Rittel coined the term "wicked problem" in the early 1970's as a corrective to the rationalist approach to planning and design of large-scale systems. The late 1960's and early 1970's were a heyday for rationalist planning, which can be summarized as the process of fully and explicitly laying out goals, assumptions, and constraints of a problem situation, generating and evaluating alternative solutions, and expecting that the preferred solution will emerge clearly, backed by good reasons. This approach grew, among other things, from the rise of digital computation, the activist Federal mood of the 1960's, and the prospect of bypassing bitter political struggles over such things as urban extensions of the Interstate highway system. Among its applications were low-cost public housing projects, flood control initiatives, moon missions, and Vietnam war strategy. The hope was that objective, data-driven analytical approaches would provide a broadly applicable toolkit for solving large-scale social and environmental problems.

Rittel saw that this hope was doomed, because the problem situations in view could not be defined in agreed, unchanging ways. These problems are intrinsically ill-defined, and attempts to define them are already actions which reshape the problem and commit the analyst to a course of problem-solving which omits legitimate alternatives—and there is no way of escaping this. An extreme example is the Israel-Palestine question: a storm of protest and counter-protest greets every attempt to say what "the question" is, much less propose answers. But much milder situations exhibit wickedness. Consider low-cost housing, middle-school math curriculum, or the length of the salmon-fishing season. Indeed, every building project and most social action, down to the smallest scale, has elements of wickedness. In each case, the interests in play, the intangibility of key values, and the elusiveness of key information mean that a commonly agreed base for problem solving is not objectively available.

There are, however, ways forward. The first is to shift the goal of action on significant problems from "solution" to "intervention." Instead of seeking just the right moves to eliminate a problem once and for all, one should recognize that any actions occur in an ongoing process, in which further actions will be needed later on. This is not to accept injustice or suffering quietly. If there are ways to eliminate smallpox as a public health problem, and the vaccination campaigns of the 1970's proved there were, one should pursue them by all means. But one should realize that smallpox will remain a part of the global health situation in some way. (And so it has, most recently as a potential bioterror weapon.) The intervention mentality recognizes that situations tend to continue, even if their form changes radically.

There is a natural fit between the wicked-problem mentality and the DESIGN STANCE (see the pattern of that name). While design has often aimed at closed, once-and-for-all solutions, the multi-factor, iterative, imagination-based process of generating designs is very congenial to what is needed for intervening in wicked problems. Design naturally generates multiple possibilities before settling on one proposal; design naturally engages in a sort of dialogue with the problem situation, in which drawings or other representations of the design idea reveal consequences or relationships which call for changes in the design idea, and vice versa. The precise definition of the problem evolves alongside the ideas for interventions until they converge on action.

A second way to work with the wickedness of significant problems is to admit the significant actors to the design process. A typical wicked problem is shaped and reshaped by multiple actors whose influence cannot be closed out. The long maneuvering over the reconstruction of Ground Zero in New York City is a classic example. Both the rationalist tradition in architecture, engineering, public policy, and most other fields and what could be called the now-we-need-a-genius tradition in those same fields have relied on an expert (or sometimes a team of experts) generating a solution in isolation. But the multiple actors in wicked problems can not only obstruct such a solution, they can change the problem’s definition while the solution is being generated.

A third step is to design "loose-fit" actions. Instead of tailoring an intervention tightly to the understood conditions of a problem, for example choosing sealed windows and central air conditioning for an office building, one should allow for uses, costs, and regulations to change in unforeseen ways, for example drastic escalation in energy costs. Architecture is one area where one can now see that the rationalist, optimizing mood of the 1970's (and after) has saddled businesses and communities with rapidly obsolescing buildings of many kinds, but the need for loose-fit designs or plans occurs in any area with wicked problems.

A powerful example of successful handling of a wicked problem is passive solar design of buildings. The problem area emerged from the oil crises of the 1970’s, as activists’ strong desires to make use of solar energy as a renewable, free alternative to oil encountered technical difficulties such as low efficiency and daily and seasonal variability, economic challenges such as high first cost and low availability of experienced suppliers, and political/cultural resistances including vested interests, suspicion of unfamiliar technologies, and opposition to ideas perceived as counter-cultural. This mix of difficulties and disparate actors is typically wicked.

Slowly, over two decades, the problem and goals shifted (e.g. from replacing oil heat to reducing the need for it), experimentation revealed unforeseen directions of development (e.g. building orientation and control of overheating became more critical than total window area), knowledge from traditional practices as well as from engineering measurement began to accumulate, and designers found more and more ways to blend solar performance with standard building functions. A growing consensus on good practice included building code officials and contractors, as well as solar enthusiasts and academics. Discovery, development, and a degree of controversy continue, but the U.S. is now at a point where passive solar guidelines are widely available and used, at times in such routine guise as to be invisible.