be valuable in the future, are no longer worth pursuing. A report that goes too far in the opposite direction risks giving its readers the impression that problems requiring immediate attention admit of no policy solutions. Definitive results are seldom produced by an immature field, but the interim findings available may suggest approaches that are, at least, better than doing nothing.

This problem is not simply applicable to summaries of the knowledge of health care. All levels of generalization, from the most catholic conceptions to detailed assertions of scientific research, experience this difficulty. This discussion is concerned with the generalizations of a system for the classification of the consequences of disease. Exploring the expected and unexpected results of system building will illustrate the way generalizations create imbalances that are sometimes redressed.

From the seventeenth century onward, Western thinkers have had an ongoing preoccupation with the creation of systems—systems designed to define, refine, and reform philosophy, science, theology, manufacturing, and so forth. The impetus behind most systems is to find fundamental principles of order, simplicity, or efficiency where there seemed to be overwhelming disorder, complexity, or inefficiency. In short, the immediate goal of making things more systematic, with the ultimate goal of making them easier, has motivated many rigorous inquiries in the West.

Yet the rigor of such inquiry owes as much to provocation as it does to the desire for ease.⁷ Systems, especially complex conceptual systems, can be understood as subtle but powerful forms of intellectual provocation. Our standards for accepting systems are generally low. They are, in fact, much lower than the standards most system designers set for themselves—and often much lower than the claims made by the system designer. With the expectation of being more reasonable, straightforward, and efficient in understanding, predicting, and controlling the systematized subject, Western man has been willing to accept the fundamental principles of systems. Such popular acceptance, however, can be a source of frustration for individuals who disagree with the principles of a given system and may eventually provoke a loud and forceful dissent, even as the system is perceived as a source of convenience by the majority of the public. When these dissents unmask discrepancies between the system's performance and its creator's claims, inquiries to judge the principles at the foundation of the system may be set in motion. Systems therefore may unintentionally balance the force of their generalizations by provoking their detractors to expose their deficiencies.