118 5: Th eories of Public Management
operation. Th en, most remarkable of all, they go ahead and do what their calcu-
lations suggest they should. (1995, 293)Th ese fi ndings are a kind of leadership version of the best practices research
found in the NPM literature, and like that research, they are grounded in direct
observations that describe or account for singular events that are diffi cult to rep-
licate. Th is is theory built on soft , qualitative fi ndings, diffi cult to replicate and
equally diffi cult as a body of knowledge or data from which one might draw de-
fensible fi ndings or conclusions. Th is research is unquestionably a series of fairly
clear snapshots of what has happened and how it happened in individual cases,
recognizing the biases of the camera operator.
Perhaps a better way to view this framework is taken from cultural anthropol-
ogy, where it is assumed, fi rst, that “reality” is a social construction rather than
an objective thing or phenomenon that is the same for all observers; and, second,
that organizations are a system of socially constructed and cognitively ordered
meanings (Lynn 1996; Boisot 1986; Weick 1995). Th is form of knowledge is tacit,
understood, oft en unspoken, but generally shared and accepted; it is also under-
stood to be inherently vague, ambiguous, and uncertain, but is nevertheless oft en
an important guide for behavior. “Codifi ed knowledge, in contrast, is more im-
personal, associated less with proper socialization or experience than with skill in
abstract thinking or linear reasoning” (Lynn 1996, 112).
Laurence E. Lynn Jr. makes informative observations on this form of manage-
rial theory:
Some types of knowledge are easier to come by, because they are more linear and
impersonal than others. Considerations of deference and trust do not intrude.
We may call such knowledge “scientifi c” or “technocratic.” Uncodifi ed, undif-
fused knowledge, e.g., the tacit, intuitive knowledge of a wise and experienced
manager who communicates primarily in face-to-face forums and by the ex-
ample of behavior, is diffi cult to master in any conventional sense because con-
siderations other than the literal and the logical are involved. We may call such
diffi cult-to-acquire mastery “artistic.”
For codifi ed, technocratic knowledge, the university course, the supervised
workshop, and the problem set or other formal assignment, all emphasizing
individual goal achievement, may be the most effi cient approaches to learning.
For tacit, artistic knowledge, mentorship, apprenticeship, and internship tend-
ing toward socialization—and perhaps prolonged on-the-job experience—are
likely to be the most effi cient ways to promote learning. Where both types of
knowledge, or their subtle integration, are required for success, vicarious ex-
perience in simulated work situations featuring both social and intellectual
demands—for example, case discussions requiring the application of a mix
of analytic frameworks—may be the appropriate approach to learning (1996,
113).