December 26, 2006: "Clinical" IT

I would like to continue my thoughts on the question of whether or not there is a suitable academic foundation for "service science" as inspired by my reading of Decision Support Systems: An Organizational Perspective, by Peter G. W. Keen and Michael S. Scott Morton. This time I would like to begin with a question posed to me last October by a friend working at Accenture: He wanted to know what Accenture could do to make IT organizations better at what they do. My point of departure for answering this question involved an analogy from medical practice, viewing the IT organization as responsible for the "health" of the organization. I argued that IT developers needed to take a clinical approach to "taking the history of the patient," who comes in with some set of possibly pathological symptoms. I then extended the analogy by arguing that enterprises need "health maintenance," rather than "illness treatment." Ultimately, this conversation did not progress, possibly because the analogy was too much of a departure from the "normal practices" (as in "normal science") of both Accenture and its customers.

Imagine my surprise then in discovering that Keen and Scott Morton proposed a "clinical" approach to the development of decision support systems! What was particularly interesting was that they recognized two different levels of diagnosis that need to be performed by IT developers. The primary level of diagnosis relates to what I called "taking the history," identifying what needs to be changed and how IT can facilitate that change. However, they insisted that it is also important to diagnose symptoms of resistance to change, because, if the resistance is not "treated," it is not going to matter very much how effective the proposed solution is. The forces of resistance can undermine even the best of ideas, no matter how well they are implemented!

I then noticed in the Bibliography that Keen had been advocating this clinical stance since 1975, when he wrote a Sloan School Working Paper on the subject. That means that the idea has been around for over 30 years but has never really "taken" in the world of enterprise software. I suspect one reason for this is that this kind of thinking does not fit into the specializations found in most academic curricula. One does not go to business school (or, for that matter, computer science departments) to learn about "the socio-technology of diagnosis," let alone the intellectual skills required for such diagnosis, such as an understanding of the subjective (as well as objective) motives behind speech acts or the use of narrative as a tool for "thinking in time." As a result, Keen's insights seem to have faded into obscurity.

Meanwhile, the history of attempts to make IT useful to the enterprise continues to repeat itself. As Marx said, what is tragedy the first time around becomes farce with the next iteration. However, he did not say anything about any subsequent repetitions!