HBR & Complexity

I suppose mainstream applied social science journals like Harvard Business Review feel they must stay grounded in a definition of science that emphasizes the "logic of the facts" (though, my opinion is that social sciences are more logic than facts....i.e., the inferences often run way ahead of the data).

However, real-world needs are usually beyond what a thick-data/thin-logic approach (e.g., Brahe's data + Kepler's equations) will yield...so, there's a tension between social science that's narrow enough to be relatively "hard", and social science that's useful to the practitioner applying it to a real world decision context.

These thoughts were triggered by HBR publishing two complexity-centric articles (at least) in the last few months (the Cynefin article in Nov 2007, and "Strategy as a Wicked Problem" this month (May 2008)).

I suspect part of the reason is highlighted in the second paragraph: "[Companies] can't develop models of the increasingly complex environment in which they operate."  Those of us who have been looking at the changing IT landscape (e.g., SOA, Web 2.0, etc.) over the past few years would second this statement.  It's as true for innovation and adaptability/agility in operations as it is for strategy. 

The problem is that management science has been epistemologically grounded in a reductionist approach to knowledge since it began to be formalized by Taylor in the early 1900's.  Modeling (ideally via math or simulation) has been a basic analytical tool since those early days.  Anyone who deals with complexity knows the limits of modeling...and an understanding of those limits seems to be dawning on the business world.

Like Snowden, the author recognizes that wicked problems are not simply bigger...they're ontologically  ("being-ness") different.

The author defines 5 characteristics of wicked problems in strategy formulation:

• Involves many stakeholders with different values & priorities
• Roots are complex & tangled
• Difficult to grasp; changes when addressed
• Unprecedented
• No "right" answer

And, the author's suggestions are consistent w/ Snowden's discussion of the Complex domain:

• Lots of collaboration (Cynefin: Probe-sense-respond (vs. sense-analyze-respond))
• Define corporate identity (Snowden & Weick have both written at length on this aspect of complexity.  Known/Knowable work does not stress corporate DNA in the same way Complex/Chaotic work does.  This is especially "wicked" when you consider that identity not only helps us shape our environments, but our environment also helps shape our identity)
• Focus on action (again, Probe-sense-respond...and use boundaries & attractors (along with weak signal detection) to shape a complex context).  In the PPG example, Pareto analysis & scenario analysis are used as probes...though I'm not sure Snowden would be very enthusiastic about these two approaches to probing (at least in decision-rich contexts).
• Adopt a "feed-forward" orientation (again, boundaries & attractors...and, the author even says "scan the environment for weak signals")

Overall, a good article on an infrequently discussed topic.  I hope the trend of occasionally addressing the topic continues.