Abstract

In this essay, we aim at interpreting the recursive aspects of the reproduc¬tion of organizational knowledge building on a cognitive definition of routine. Within this framework, routines are developed, "à la Simon", around the concept of representation for action. Thus, we can characterize routines as sets of rules, and show how their birth is better described by the interplay between the heteronomous and autonomous regulation processes generating behaviors, rather than the usual dichotomy between designed and emerging processes. Within this theoretical proposal, we underline the dual nature of routines. They are the result of the intertwined regulation processes, on one hand, and, on the other, they become themselves sources of regulation limiting and channelling future action. Thus, they structure the process of genesis of subsequent routines. It is clear, within this framework, that routines can produce a certain degree of stability in the behaviors of subjects and, at the same time, induce change by means of the synthesis between the different regulation processes.