I did not intend the blanket statement that "the learning curve has
nothing to say about innovation". I don't agree with that statement
and, without checking whether I actually made it, I'll just blush and
make clearer what I'm addressing.
I am a fan of Abernathy and use his work to demonstrate that
production should be about learning rather than the mechanistic
approaches that production people tend to use.
The following quote was offerred by Phillip:
> " According to Abernathy and Wayne (1974), the risks associated with a
> strategy of achieving competitive advantage by moving down the learning
> curve come from the fact that the 'conditions stimulating innovation are
> different from those favouring efficient, high-volume, established
> operations.'
This was also a main point of Juran who talked about the distinction
between "breakthrough processes" and "control processes" and the
difference in approach and in mangement that was required for each.
For reasons that are unknown to me (I suspect mechanistic and
reductionist ones) the control process was taken on mightily by TQM
but the breakthrough hardly at all in popular press or in action.
>The risks are greatest for firms that are most successful in
> reducing price and capturing market share. The successful company must
> capture a larger and larger market share to continue to double cumulative
> output.
The assumptions in the above statement are valid _if the parameters
of a market, production process, etc are accepted as given_ but they
are far from universally valid. What is required is to see that
"output" is not the source of the learning curve. Not in the theory
itself. Certainly not in a current formulation. The key element is
"experience" or, a more modern and accurate statement would be,
"information". (We are all physicalists at a deep level by gift and
curse of our culture.)
That is, we need to deconstruct or unpack "output". Output is not
such an obvious thing when examined. Whose output? Which chunks of
output? What combinations of outputs? The fallacy that has been
obscured is that people assume output to mean the final, finished
product. But that is an arbitrary unit that, if selected, will lead
to the conclusion you state about growth and the need capture market
share. (It pays to unpack "market" in the same way to discover the
same fallacy.)
There is no end to learning. There is no end to innovation. There
is no end to creativity and knowledge creation. There is no _the_
learning curve. There are many learning curves interacting
continually and coming into being and disappearing. My regular
conflict with those who talk about learning is the narrow
interpretations and limited mechanistic processes they attach to it.
(Not to accuse you of being one of those because I don't know.)
> At the same time, the changes it has undergone in perfecting
> the production of a standardized product at ever-lower costs reduce the
> ability to respond to the competition on this basis." (Applebaum and Batt,
> 1994).
>
The following claim about what I've overlooked I hope is now revealed
as something that I've addressed directly by introducing distinctions
(such as "chunking") which is the source of escape from the paradox
of learning curve advantage also being the achille's heel of an
organisation. The pursuit of ever lower costs _at the same time_ as
the pursuit of new distinctions, directions and processes is the
challenge. The basis for the successful pursuit, I'll wager, is that
the elements of experience, information or learning can be worked
with as "building blocks" and "mechanisms" - to borrow terms from
John Holland - or as "core competencies" to borrow less precise terms
from Hamel and Prahalad.
In biology, the main source of the new is recombination of existing
(mainly) successful "production" strategies - like genes. I think
that the main source of new ideas and directions is also a result of
recombinations and that the successful (like 3M) have figured out the
chunking and the mechanisms for this process. (I consider the
process to be a linguistic one of deconstruction and distinction
among other linguistic phenomena.)
And I completely agree with the following statement:
> This last point is, I think, a key one overlooked by Michael. The learning
> curve approach is not just about people learning, it is also about
> installing production technologies based on this learning. In a workplace
> most learning is incrementally based on the existing context.
My model of growth and development of a company is slightly
different than the following one:
> A company is
> steered down a particular capital investment strategy by its conceptual
> model. Once stuck with that technology it is very difficult to switch
> tracks.
The beginning has to do with individual, company and cultural forces
that create a particular model. Then action takes over and much of
the model is violated, forgotten and otherwise becomes a minor
factor. Complex adaptation and emergence take over and the original
model becomes some identity with (frequently) little obvious or
traceable relationship to the original model. (IBM made weighing
machines. How does their switch to computers fit your model?)
Incremental change does continually build on the platform provided by
what went before. And each change creates new rigidities in the
physical and linguistic interaction of elements.
However, while acknowledging the intellectual investment (actually
far above the capital investment in my opinion), I'd say that what it
takes is the ability to create distinctions and to generate
processes and mechanisms that create "automatic" recombination of the
"gene pool".
And for those who are still with us, I'm finally drawn reluctantly
into the capitalist growth conversation.
> Applebaum and Batt's point is also highly apposite when we consider the
> recent discussion thread on whether or not it is necessary to grow in a
> capitalist system. The lesson here is that it is not the capitalist model
> which compels growth, but the adoption of the learning curve as a
> fundamental strategic concept.
The capitalist model says, take your money where the return on
investment is. It says _nothing_ about growth. One validation for
this statement is that economists of the capitalist (or any other)
persuasion have nothing to say about growth. It is somewhere between
impossible and very difficult to find any serious economic writing
about it.
It may be that learning curve theory - at the unsophisticated level
of common experience - has suggested growth as the answer. Wrongly
but understandably in a materialist/physicalist view of things.
I would speculate that what has caused the "growth is the answer"
syndrome in our culture (leaving out my regular deeper cosmology
answer) is two things. One is that we can justify more for ourselves
if the whole is bigger - and I am referring to things like executive
compensation and short term self-interest. The other is that we have
built systems that do not take account of the whole and enable me to
use resources that I don't have to pay for and that leads to the
"tragedy of the commons" on a grand scale. One of the ignored
features of the tragedy of the commons is that, when the land is
free, then the users attempt to grow to get the greatest possible
share of it without concern for balance, usefulness or future costs.
-- Michael McMaster Michael@kbddean.demon.co.uk