Tuesday, June 28, 2011

The Sage Kings of Antiquity

The ancient Chinese philosopher Mozi (Mo Tzu) identified three criteria for judging a theory
  1. Origin - reference to the sage kings of antiquity
  2. Validity - reference to the evidence ("the eyes and ears of the people")
  3. Applicability - whether it brings benefit to the enterprise and the people
Mozi called this his "three-prong method". [Wikipedia: Mozi]

Enterprise architecture frameworks can and should be judged against the second and third of Mozi's criteria, and I should certainly like to see more effort and rigour in presenting the evidence to support a wide range of so-called principles and methods.

But what about Mozi's first criterion? Much of the current thinking and practice of enterprise architects can be traced back to a bunch of methodology gurus from the 1980s. A complete list would be impossible, but it would include people like Tom de Marco, Clive Finkelstein, Michael Jackson, James Martin and Ed Yourdon, and with John Zachman just gettting in at the end of the decade. (I think Zachman has more in common with these methodology gurus than with the OO/Patterns gurus - Booch, Jacobson, Rumbaugh, Wirfs-Brock et al - who followed them.)

The term "enterprise architecture" wasn't used in the 1980s, and appears to have been coined by Steven Spewak in the early 1990s. In those days, people generally talked about Business Systems Planning (BSP) and Information Systems Planning (ISP), and a lot of the early thinking came out of IBM (where both James Martin and John Zachman had worked). Information Systems Planning was the top level of the Information Engineering pyramid, produced by Finkelstein, Martin and others.

Why should 21st century enterprise architects care what the sage kings of methodological antiquity thought about enterprise systems? Firstly because the sage kings have given us a huge legacy of principles and practices, much of which are still in use today. Secondly because some of the assumptions made by the sage kings may now look obsolescent, not just because of technology change but because of emerging ideas of complex systems organization and management, and we need to critically review and refresh our principles and practices to ensure we are not still bound by these assumptions. And thirdly because there may be some principles and practices that have fallen into disuse and deserve to be revived.

Given that the empirical evidence for enterprise architecture is fairly weak, anecdotal and inconclusive, we are still more dependent than we might like on the authority of experts - whether this be semi-anonymous committees (such as TOGAF) or famous consultants (such as Zachman). And are the experts consciously or unconsciously recycling the wisdom of the sage kings?

So how much difference is there between today's enterprise architecture frameworks and the structured enterprise methodologies of the 1980s, such as information engineering? And how confident are we that we have kept the good bits and discarded the bad bits?

Declaration: I worked for James Martin Associates from 1986 onwards.

No comments: