Musings About Knowledge Work LO28643

From: Fred Nickols (
Date: 05/28/02

I've been pondering some issues related to (among other things) knowledge
work, CoPs, learning organizations, self management and productivity to
name a few.

As some of you know, I've been in pursuit of knowledge work and knowledge
workers for more than 30 years now, specifically, the problem of making
knowledge work productive. I was set off on my chase by Peter Drucker's
first serious treatment of the advent of knowledge work in a 1968 book
titled The Age of Discontinuity. He dealt with it in detail again in his
monumental 1973 book, Management: Tasks, Responsibilities, Practices.

It seems to me that much of the managerial, executive and organizational
energy that goes in to applications related to knowledge work, CoPs,
learning organizations and especially self management, ties to the ultimate
aim of improving productivity. It also seems to me that much of this
energy is misdirected.

The goal is (or ought to be) one of making the work more productive. The
goal is NOT one of making the worker more productive. This latter goal
leads inevitably to attempts at making people work harder. Any gains in
productivity achieved via this route are hard won, they cost dearly and,
more important, they are at best only temporary. Continuing down this path
creates a downward spiral that leads to antipathy on the part of the
workers, antagonism between the work force and management and, on occasion,
to the demise of the organization. However, even though Drucker pointed
out that the knowledge worker cannot be supervised (in the traditional
sense) and that the knowledge worker must be self-managing to be effective,
simply tossing the burden of management into the laps of the workers and
proclaiming an era of self management is an equally quick route to disaster.

Drucker noted this and observed that the task was that of making the work,
not the worker, more productive. In turn, he noted that work is a process
and that it has a result. To make work productive, then, one must focus on
the desired results and the process(es) that do or don't achieve them. In
this regard, the knowledge worker, if he or she is indeed to be self
managing and to take responsibility for improving the productivity of his
or her own work, must acquire no small capability with respect to the
design and engineering of work processes.

At this point, lots of issues crop up. To what extent should results be
set down for knowledge workers to achieve and to what extent should they be
specifying those results? The dividing line is not easily drawn. How does
one provide "direction" to people who are managing themselves? How is
individual effort to be integrated? Whose responsibility is that? These
questions, and countless more in a similar vein, are not new to many
members of this list and dozens more just like them could be easily added
to the list. I'm not trying to present nor and I trying to assemble a
comprehensive list of issues; instead, I'm about to suggest that many such
questions are much more selectively applicable than many seem to think. In
other words, all the hoopla and folderol surrounding the shift to knowledge
work and the consequent calls for self managing individuals and teams
might, in many cases, be much ado about nothing. Here's why.

One key aspect of the shift from manual work to knowledge work is a shift
in the locus of working, that is, a shift from the muscles to the
brain. Most of us now use our heads more than we do our hands or our
backs. This shift from overt to covert activity, in turn, signals that
working has moved out of sight and is no longer accessible for the purposes
of direct supervision. That, alone, is cause for alarm in some circles.

But, I don't think that that's the core issue when it comes to making
knowledge work more productive. Instead, I think the core issue is the
distinction between work that consists of prefigured routines and work that
consists of configured routines. This, distinction, also drawn by Drucker,
refers, in the first case, to work that has been designed for people to do
and, in the second case, to work that is designed by the people who do it.

I've long sorted so-called "mind work" into three piles: information work,
knowledge work, and intelligence work. Briefly, the three are
distinguishable as follows:

         Information work consists of unambiguous information processing
operations (e.g., processing a claim, a loan application, an insurance
application and so on). The basic inputs and outputs are information and
the processes can be prefigured. In other words, the results and the
processes can be specified in advance. This kind of work can be performed
by people or by computers and, in most ways, it can be (and usually is)
managed much the same way that manual work has been managed for quite a
long time.

         Knowledge work entails the application of knowledge in the course
of producing results that manifest themselves apart from any information
being processed. Applying knowledge to produce results requires
intervening, acting to change things with some purpose or outcome in
mind. Some of this work comes close to being prefigurable (e.g., much of
what is known as technical troubleshooting or fault isolation can be taught
and applied in a very systematic manner). However, much of this work
remains pretty much a mystery (e.g., we know far less about medical
diagnoses than we do technical troubleshooting). Somewhere in the middle
of this kind of work lies the problem solving associated with managerial
work. So, at the lower end of complexity and ambiguity, knowledge work
comes close to be prefigurable. At the higher end of the scale, it is the
kind of work that has to be configured, i.e., designed to fit the situation
at hand.

         Intelligence work produces new knowledge and
information. Researchers do this kind of work; so do intelligence
analysts. Intelligence work is almost always of the configured variety.

At one end of this continuum of mind work, the information work end, self
management is not only not necessary, it is quite probably undesirable and
likely to prove costly and counter-productive. For work that can be
prefigured, conventional or traditional management methods and techniques
will work quite well. At the other end, the intelligence work end,
configured work is the rule and very different management methods and
techniques are called for. Even so, simply relinquishing or surrendering
the control over work to the worker is not necessary nor is it
desirable. Between these two extremes, of course, is a vast gray area of
situations that must be dealt with on a case-by-case basis.

It is also the case that, as we come to know more and more about a
particular category of work, that it can be converted from configured to
prefigured work, thus introducing greater standardization, quality and
reliability and, ultimately, driving down the cost of doing the work in

My reason for being on this list, as well as on the CoP list and some
others, isn't because I have an interest in learning organizations or CoPs
per se but because I see learning organizations and CoPs as avenues leading
toward the enhanced ability of people in organizations to improve the
productivity of work, in many cases, their own work but also the work of

This, then, is an attempt to steer some of the conversation in that direction.


Fred Nickols
"Assistance at a Distance"


Fred Nickols <>

Learning-org -- Hosted by Rick Karash <> Public Dialog on Learning Organizations -- <>

"Learning-org" and the format of our message identifiers (LO1234, etc.) are trademarks of Richard Karash.