January 1, 2006

...Learn TDD with Codemanship

Reductionism vs. Complexity

More lively discussions about process management and agility have caused me - complete with Jan 1st headache - to get up early and think some more about the whole debate.

It's always dangerous to make sweeping generalisations - so here goes nothing:

Is the business world now divided into two distinct schools of thought?

On the one hand we have various management theories that are built on the ability to understand complex systems by breaking them down into simpler subsystems, and understanding the cause-and-effect relationships between them. By doing this, we believe we are able to get an understanding of the behaviour of the whole as being exactly the sum of its parts. This is reductionism, leading to determinism. Phenomena like throughput, efficiency, sales, profit, quality can be engineered in much the same way that we can engineer the blue ball into the corner pocket by hitting it at exactly the right angle, speed and spin with the cue ball. This is Linear Management Theory - or "classical management theory". Management theories that seem to fit into this category include the Theory of Constraints, Total Quality Management, Business Process Management and others of those ilk.

On the other hand we have a handful of relatively new management theories that are built on the shiny new science of complexity. No matter how much we understand about the simple components of a complex system - like a business, for example - it offers no understanding of the whole, and offers no ability to predict or engineer outcomes by predicting or manipulating the behaviour of its component parts. This is what I'm going to start calling Nonlinear Management Theory, or "modern management theory", because it is new, risky and unpredictable. Nonlinear management theory offers a stark contrast to classical management theories because it offers no comfort blanket of predictability or simple cause-and-effect.

Just as with classical and modern physics, there are times when Linear Management will work, and times when it won't. Understanding the behaviour of fundemantal particles is little help in predicting the winner of the 2:30 at Kempton Park, for example. Winning races is an emergent phenomenon. No amount of understanding of the individual horses, riders, the track, the weather, or the quantum states of the particles that make them up, will allow us to know with any degree of certainty whether Mr Fiddles will win by a nose. I suspect profit is also an emergent phenomenon. Software is probably another. When outcomes are emergent, I wonder if Linear or Nonlinear Management best applies.

A key question that's now wracking my addled mind is what this has to do with "agility" - whatever that may mean. My instinct is to look to the two camps, and propose that "lean" sits more in the Linear Management spectrum, and "agile" sits more in the Nonlinear Management spectrum. More specifically, if "agility" is the ability to adapt to discontinuous change, and if discontinuous change is a product of complexity - which it generally appears to be - then the two might be closely related.

I can imagine this going down like a tonne of bricks with the many (including myself up to now) who have subscribed to the notion that "lean" and "agile" are kind of, sort of the same thing (after a fashion).

This at least brings me a little closer to a working definition of agility - it is a product of Nonlinear Management.

Of course, this all may turn out to be utter rubbish, as most sweeping generalisations tend to do :-)
Posted 15 years, 8 months ago on January 1, 2006