In general, the more an organization depends for its survival on innovation and adaptation to a changing environment, the more essential its interest in organizational learning. On the other hand, formal organizations also have a powerful interest in the stability and predictability of organizational life. An organization is a cooperative system in which individuals depend on the predictability of one another’s responses. Managers must rely on the predictable behavior of their subordinates. Surprise, which is essential to learning, is inimical to smooth organizational functioning. Thus organizations evolve systems of error detection and correction whose function is to maintain the constancy of variables critical to organizational life. They are “dynamically conservative.”
Significant organizational learning – learning which involves significant change in underlying values and knowledge structure – is always the subject of an organizational predicament. It is necessary to effective organizational adaptation, but it disrupts the constancies on which manageable organizational life depends …
Reflection-in-action is both a consequence and cause of surprise. When a member of a bureaucracy embarks on a course of reflective practice, allowing himself to experience confusion and uncertainty, subjecting his frames and theories to conscious criticism and change, he may increase his capacity to contribute to significant organizational learning, but he also becomes, by the same token, a danger to the stable system of rules and procedures within which he is expected to deliver his technical expertise.
Thus ordinary bureaucracies tend to resist a professional’s attempt to move from technical expertise to reflective practice. And conversely, an organization suited to reflective practice would have features very different from those of familiar bureaucratic settings.
(Shön, 1995, p. 327ff)