A preliminary typology of organizational learning: synthesizing the literature

Journal of Management, Fall, 1996 by Danny Miller

Organizational learning processes are being explored with increasing interest and vigour. But it remains unclear just what learning is, how it takes place, and when, where and why it occurs. Part of the problem is that learning, as portrayed in the literature, is a haphazard and eclectic notion. Researchers lump together processes that are strikingly different in their causes, effects, and domains. The thesis of this article is that we can only begin to understand learning after we have made some essential distinctions among its many varieties.

Based on two fundamental contrasts among the dominant organizational paradigms: voluntarism vs. determinism, and methodical vs. emergent thought and action, we have identified six common modes of learning. These manifest contrasting cognitive and socio-political processes, engage different parties, and have varying spheres of influence. We shall argue that the nature, contexts and typical outcomes of the different learning modes will be very distinct.

What is Organizational Learning?

The definition of learning remains somewhat obscure, in part because the process has been described so differently in the literature. Some authors view learning as a change in behavior in response to a stimulus (Cyert & March, 1963; March, 1989). But this seems mostly to be a description of reaction or adjustment, which may be blind, automatic, and productive of no new knowledge. Other scholars suggest that learning requires some conscious acquisition of knowledge or insight on the part of organization members (Argyris & Schon, 1978; Hedberg, 1981; Huber, 1991). But were this knowledge to be unrelated to organizational action or decision making, it would be relevant only to individual learning, not to that of organizations.

Fiol and Lyles (1985) noted that "organization learning means the process of improving actions through better knowledge and understanding." We agree that it is important to include both cognitive and behavioral elements in a definition of organizational learning. However, we are uncomfortable with any normative overtones. Given that some kinds of learning may ultimately do more harm than good, and that learning can help one objective or manager at the expense of others, we propose a different definition: organizational learning is the acquisition of new knowledge by actors who are able and willing to apply that knowledge in making decisions or influencing others in the organization.

Learning is to be distinguished from decision making. The former increases organizational knowledge, the latter need not. Learning may in fact occur long before, or long after, action is taken. Of course, methods of decision making may well influence learning processes, and vice versa.

Dimensions of Contrast: Two Paradigmatic Distinctions

Some of the most influential paradigms of organizational research can be contrasted along two important dimensions. The first dimension is the extent to which human or organizational action is constrained by cognitive, political and resource factors. Compare, for example, voluntaristic theories of business strategy that envision few such constraints, with more deterministic bureaucratic or ecological theories that anticipate many such limitations. A second dimension is the mode of administrative thought and action. Compare economic theories that emphasize methodical and intentional behavior, with institutional theories that envision emergent and spontaneous behavior far less driven by technical or economic norms.

Voluntarism vs. Determinism: The voluntarism-determinism axis gauges the extent to which people and their institutions are deemed intelligent and autonomous actors rather than entities severely restricted in cognition and action (Astley & Van de Ven, 1983; Berger & Luckmann, 1966; Burrell & Morgan, 1979; Giddens, 1979; Hrebiniak & Joyce, 1985). Strategic choice theorists, for example, adopt voluntaristic perspectives, allowing much latitude for free choice by decision makers (e.g. Andrews, 1971; Ansoff, 1979; Porter, 1985). Bureaucratic and neo-bureaucratic theorists are more deterministic and view cognition as being channeled and behavior as being constrained (Crozier, 1964; March & Simon, 1958; Cyert & March, 1963). Population ecologists see almost no latitude for strategic choice - or at least for strategic choice that increases the chances of organizational survival (Hannan & Freeman, 1977).(1)

These paradigmatic distinctions suggest an important difference in the way organizations can learn. We may contrast voluntaristic learning that is relatively free in thought and action, with that which is bound by cognitive (March & Simon, 1958), resource (Cyert & March, 1963), political (Lindblom, 1959) or ideological strictures (Berger & Luckmann, 1966). Global analysis performed by top managers is a relatively unconstrained kind of learning. Ideally, it surfaces critical assumptions about an organization and its markets, evaluates a wide range of strategic options, and affords much latitude for taking action (Ansoff, 1980). Some scholars, however, see firms as acting and learning in a more incremental way by performing small experiments that only slightly modify existing methods or offerings (Cyert & March, 1963; March & Olsen, 1976). A still more constrained kind of learning takes place within existing routines and programs: for example, employees may learn how to adjust or modify their equipment based on feedback from their quality control monitors (Nelson & Winter, 1982).(2)