Prototyping considered dangerous

In this paper, we argue that prototypes can hinder, rather than aid effective communication. The dangers are: (1) prototypes may contain hidden assumptions which might surface too late; (2) obtaining feedback in the context of use is prohibitively expensive and rarely done; and (3) partly as a consequence of these problems, prototypes cause a focus on displays and other surface features of computers, not on the more difficult problem of how people function in their environments to solve problems. We propose design intent which specifies how a system will fit in and interact with the environment in which it is placed and expectation agents which monitor the system in use and detect uses counter to the intent as ways to alleviate these dangers. the “Tower of Babel” problem (Brooks, 1975), communication between diverse groups of people is required. The second problem, fluctuating and conflicting requirements, arises primarily because computer systems fit into larger environments which are not themselves static. While requirements sometimes change because people involved in the development effort rethink earlier decisions, more typically, they change to reflect the fact the environments are changed. Because the environments into which systems are to be placed continue to evolve, problem understanding must proceed continuously and simultaneously with other system development activities. New requirements are likely to emerge as understanding matures. Similarly, the third common problem, communication bottlenecks and breakdowns, does not typically occur because one group neglects to communicate with another. Rather, they occur because one group is not able to effectively communicate with another. The underlying problem is that many people are required to develop a system, but each has their own view of the system, does not recognize how many others have different views of the system, and cannot communicate effectively with many of the others. Our solution to this “symmetry of ignorance” problem (Rittel, 1984) is to provide a common artifact through which many people can communicate. While this is the solution that prototyping intends to implement, we will argue later that it can often fail to do so. 3 FLAWED IMPLEMENTATION OF NOBLE