Virtuous Growth Cycle LO10707

Julian Macnamara (100317.2417@CompuServe.COM)
26 Oct 96 11:03:35 EDT

Replying to LO10682

Andrew Rowe asks:

...big snip...
> How would you explain SIMPLY to students brought up very much in the
> traditional management paradigm (if I could term most of the other courses
> offered by the depatment) how Stacey's work improves upon traditional
> business strategy? Indeed, how can chaos have an order?

> Secondly, it has been suggested that the main thrust of Stacey's work
> seems to resonate closely with Adam Smith: that organization strategy is
> best left to the 'invisible hand' of the informal organization, rather
> than Stacey actually providing a detailled answer to business strategy in
> a chaotic world. I'm not too sure about this criticism, what do you
> think?

Andrew, thanks for posing a couple of awesome questions.

First off, my LO10582 references the work of Chris Langton which
demonstrates the transition from "order" to "complexity" to "chaos" in
cellular automata and that this is a function of an intensity factor which
links each of the auotmata to its environment. Richard Tanner Pascale
demonstrated something similar in "Managing On The Edge" in real
organisations but, personally, I disagee with aspects of his explanation.
(An interesting aspect of Chris Langton's work is that it confirms
Wolram's class IV laws which a physicist would recognise as second-order
phase transitions).

I would suggest that people need to be clear that 'chaos' per se is merely
a shorthand way of saying that random behaviour can occur in situations
which appear to be ruled by exact and unbreakable laws. If your
undergraduates are of a mathematical bent, then the definition that should
appeal to them is "stochastic behaviour occurring in a deterministic
system".

In this sense, chaotic behaviour is a potential property of any system
that incorporates positive feedback and where the future depends on the
past, ie, any iterative system. Many systems are iterative and Weick would
argue that organisations are complex iterative systems, starting right at
the fundamental level of interaction between two or more people within the
organisation.

In terms of explaining it, some starters might include:

- Setting up a simple microphone and loudspeaker system to demonstrate
acoustic feedback. Here the volume control is the "intensity" factor. This
produces even more spectacular results if you do it with a couple of video
cameras.

- Use a computer simulation. I have a couple that you're welcome to. One
demonstrates how "news can spread like wildfire", the other is a
production planning simulation and shows how stable demand will almost
always produce fluctuations in the distribution of demand; under certain
circumstances, which it is hypothesised occur frequently, the distribution
in demand can appear to be almost random - even though the system is
actually responding to very simple rules; in the light of the above,
accurate models of demand based on demand history are difficult to
develop; even where an accurate model can be developed, its value as a
predictor of future demand is limited; in situations that are
characterised by stable underlying demand and fluctuating distribution in
demand, level production scheduling will produce at least as good as, if
not better, results than variable production scheduling in terms of
service levels and finished goods inventory.

- Get your undergrads to consider an orchestra and a soccer team and
consider: What is their most important critical success factor? How do
they prepare? Why do they prepare in thisway? What happens during the
concert / match? How do they measure their success? What stops 'chaotic'
behaviour occuring? The answers to these questions are not as obvious as
they appear. Key points are that an orchestra represents a very good
example of scientific management a la Frederick Taylor (a much maligned
man) and operates in the way that it does because the required outcome is
already known. The football team knows what its trying to achieve but not
how to achieve it.

- Presumably your traditional management colleagues use Ansoff's matrix.
Get the undergrads to review Ansoff's later thinking on turbulent
environments and the importance of weak signal analysis. A turbulent
environment is one in which growth doesn't extrapolate, profitability
doesn't follow growth., historical strategies are suspect and the future
is highly uncertain.

- Use a chess analogy.In chess, the principles are easy to understand. The
game operates on two levels - the physical and the mental. Moves
communicate a strategy which an opponent responds to. Thus, chess is a
system which involves feedback. An optimal solution is theoretically
possible, but unlikely, given that there are 10 to the 120 potential
moves. Great chess players are noted for their ability to spot patterns.
The measurement system does not tell the whole story - particularly in the
early stages of a game. Taking an holistic view of the arrangement of the
pieces helps to predict the outcome. At the same time, cause and effect
are not always immediately obvious. The true implications of a particular
move may not be apparent until some moves later. The principles of
downsizing, delayering, decluttering and reengineering usually lead to
sudden death if they are not carried out with sensitivity and finesse in
the pursuit of a wider, overarching, vision. In chess, there is no win-win
solution.

I think that one of the problems with Stacey's work is that he takes a
somewhat extreme position in a number of areas - at times he seems to be
describing anarchy, rather than chaos, and the more interesting region, in
any event is "complexity". His views on requisite variety are not
congruent with those of Ansoff.

In itself, chaos theory may, or may not, be particularly important as far
as the external environment is concerned. However, everybody is familiar
with the concept of hindsight which suggests that, in a number of
situations, it is recognised that events cannot be predicted.

According to the late Christopher Lorenz (of Financial Times fame):
rather than indulging in a costly (and ultimately fruitless) search for
reliable new decision support tools, managers would do better to rely on
the proven scenario approach - and, of course, on their own judgement .
Others, including Handy, and Davidson and Rees-Mogg, are not so
dismissive. Drucker has commented that: in any system as complex as the
economy of a developed country, the statistically insignificant events,
the events at the margin, are likely to be the decisive events, over the
short range at least. By definition they can be neither anticipated nor
prevented .

With regard to your second question, I was not aware that people had
compared Stacey's work to Adam Smith's. My immediate reaction to this was
that Smith's work is the anti-thesis to Stacey's. At the same time, an
economy is a non-linear system that displays both second and third-order
nonlinearities. The second-order nonlinearities drives the economy toward
equilibrium and stability. The third-order nonlinearities drives the
economy away from equilibrium. Thinking about this, I can't help but feel
that perhaps, these days, Smith would talk about an 'invisible network of
feedback loops' rather than an 'invisible hand' and one of his assumptions
about the 'invisible hand' was that people were seeking to achieve the
best good for all.

I need to think more about this - it's such an interesting line of
enquiry. Also, I'm not sure how interested other members of the LO are in
the failings of neo-classical economics so I'll answer this in more detail
off-line. In this respect, it would be helpful if you could send me a fax
number. (If anybody would like a copy of this off-line post, please don't
hesitate to ask).

Thanks for raising the questions.

Best wishes.

Julian

Julian Macnamara
Glandore Associates
100317,2417@compuserve.com
<http://ourworld.compuserve.com/homepages/glandore>

-- 

Julian Macnamara <100317.2417@CompuServe.COM>

Learning-org -- An Internet Dialog on Learning Organizations For info: <rkarash@karash.com> -or- <http://world.std.com/~lo/>