This issue consists of four articles on the theme of knowledge, learning, and diffusion. They are quite diverse, relying on historical analysis, laboratory experiments, network analysis of judges, and surveys of managers. In this methodological diversity, they represent well the objective of the European Management Review to address important topics and theories by a variety of approaches.
The article by Rukmini Rao and Linda Argote investigates an important issue: is knowledge encoded in people's heads or in structure, that is division of labor. They use the simple but innovative device of varying two experimental conditions: turnover rates and organizational structure. Their results indicate clearly that structure influences the retention of knowledge and thus challenge theories that purport that organizational knowledge is only the aggregation of what individuals know. The article is a rich analysis of data drawn from multiple media, including videotapes of the participants' interactions.
The EMR had the lucky coincidence to receive an article from Emmanue Lazeaga, Claire Lemercier, and Lise Mounier who supply an interesting variation on the finding by Rao and Argote. They focus on a rather fascinating field site: the Commercial Court of Paris. In their study, structure is a constantly shifting variable that reflects the preferences of judges to seek advice from each other. Their results indicate that 'remembering' is conducted by a rather stable elite that is nevertheless oscillating in its coherence over time. A nice aspect of their analysis is the use of dynamic versus network analysis, drawing upon the work of Tom Snijders at Groningen. Their conclusions reinforce the observation often lost in laboratory or case-studies: knowledge is influenced not simply by organizational structure, but by social structure that is generated by local rules of personal exchange.
The study by Denis Bayart is a historical analysis of a series of innovations that lead to a revolution in manufacturing and in business 'engineering': the concept of statistical control. The quandary that the scientists of Bell Laboratory faced was that the theory of probability as developed in the epistemic cultures of science did not comply with the conditions under which it would be applied. As sampling was expensive, quality control had to take into account the trade-off between acquiring certainty and the cost of doing so. And equally important, because sampling was an intervention by outsiders into the domain of manufacturing, there was a conflict of identities and power. Thus, the evolution of statistical control had to invent not only new techniques in response to the acquisition of new technical facts, but also in response to the social facts. It is this dialogue between fact and theory that characterizes the history of statistical control, and, Bayart claims by adduction, characterizes how applied research in general proceeds.
One of Bayart's points is that the dialogue of fact and theory also entails an institutional dimension. David Strang and Mary Still choose this institutional level as the domain of their study of the effects of ambiguity on benchmarking. Whereas the applied physicists of Bell Laboratories sought iteratively to eliminate variation in the manufacturing process, the managers engaged in benchmarking have the far harder time of trying to decide a priori what they appear to observe in another firm would be useful for their firm. A far more penetrating claim is that when managers cannot easily figure out what they do well or badly, they look at other firms to see what they are doing. Strang and Still employ both qualitative and quantitative methods to arrive at a constructionist view of imitation and diffusion. Set against an understanding of the multiplicity of logics at play, diffusion is an impoverished concept that hides the sense-making activities of rational agents trying to identify a means-end relation in a messy world.
I announced in my editor's letter in the previous issue that the representation of knowledge might at times be better encapsulated through looking at the project rather than the paper as the artefact. To start this series, I asked Robin Wensley, who is the Director of the Advanced Institute of Management in the UK, to explain what AIM does and to focus on one particular stream of research. Wensley focuses on AIM's many studies on management practices. He stresses the importance of theory engaging practice. One can claim that Bayart's dialogue of fact and theory is another way to describe Wensley's discussion. It could well be at the level of the project this dialogue is more transparent than in the tidy presentations of results found in articles. A project is far more hypothetical than an article can dare to be. A healthy epistemic community is no doubt one that understands the provisional nature of knowledge, and hence the value of future research.
A dialogue of theory and fact is a good proposal for a journal and for those that believe in a vibrant accumulation of knowledge. I thank the authors of these articles for their contribution, and the assistance of reviewers to make them better. Now it's up to the reader to do the rest.
