Defining a System

[cloughley@telegnosis.com.au]

SCheek’s posting of 10 March describing the characteristics Deming gave to a system (aim, theory & components of a system) is an new insight for me and it raises a couple of issues.

Now that we seem to be reaching some consensus on what a system is, perhaps we ought to try and identify different types of systems by these characteristics. I have tried to use them to categorize some types which initially came to mind, although they are probably neither exclusive nor exhaustive:

Pure chance systems
This categorizes systems like dice games, biological systems etc which produce statistically variable outputs conforming to the normal distribution. The theory is basic statistical theory.

Mechanical systems 
This categorizes systems like motor vehicle engines, factory processes etc where the aim is to produce a steady and consistent output. The theory is based on the laws of physics and SPC. The use of control charts can be used to identify assignable causes of defects etc. 

Complex pure chance plus human intervention systems
This categorises systems like horse races, poker games, business organisations and the stock market which have both pure chance characteristics as well as extensive human interventions and produce outputs which for some of the time resemble pure chance systems but at other times produce surprises or apparently chaotic behavior. The various component aims can be either cooperative or non-cooperative. A variety of often conflicting theories exist.

Non-cooperative systems can end up in a Nash Equilibrium, where the outcome though stable is less than optimal. The components would prefer almost anything to this non-optimal outcome but cannot reach a better bargain unless they drop their adversarial positions and work together on a common aim. This is not a trivial task.

In my opinion, it would seem Deming was addressing this issue in his work. (can anybody confirm/deny this hypothesis?) He advocated the aims of a business system should be defined in terms of how life would be better for everyone and not in terms of numbers etc. and with the aid of a common management theory/methods, aim to create a cooperative system of components whereby an optimal outcome could be achieved. 

In my experience, what seems to be happening during the last few years in many organisations is the belief that Deming theory has been superceded by more recent theories like “shareholder value” and the “balanced scorecard” which can destroy the trust holding the organisation in the cooperative, optimal mode and subsequently cause a revertion to a traditional adversarial, stable and non-optimal operation.


Regards,

George Cloughley


Message posting through the Clemson CQI Web Server.