Impact of Groupware (was Re: cultural anthropology)

Noel Dickover (
Wed, 11 Dec 1996 13:51:45 -0500

In article <58f7m5$>,
> The impact of groupware on bureaucracies and the bureaucratic culture
> is probably worth some investigation. We all know that the control
> and manipulation of information in a bureaucracy is one of the sources
> of power. The setting of policy, procedural standards and their
> enforcement is another. I believe that groupware threatens both due
> to the virtually uncontrolled lateral communications that it makes
> possible (far more than EMail), and the setting of de facto standards
> that result from this communication. This weakens hierarchy, hence
> the resistence to groupware in some organizations (my own employer
> included.) Some analogies to consider could be the attempt in the
> former Eastern Block to control FAX and photocopiers to diminish the
> spread of samizdat publications

Saw this post earlier, didn't get a chance to react. Hope I'm not too

Certainly viewing buracracies from a power-based control perspective is
valuable, and leads to certain insights. If you state as a given, that
the purpose of buracracies is to maintain power, then your follow-on
reasoning is congruent and insightful. If your intention is to looking
for courses of action to implement groupware, or improve communication, I
don't think this viewpoint will yield good results.

I find that viewing buracracies from a mechanistic perspective leads to
more usable insights of the current situation and useful action towards
improving situations. Morgan's Images of Organization (Sage, 1986) has an
excellent write-up of this perspective. I hope you won't mind a little
background of this perspective, which will help explain my following
points. If you do, just skip three paragraphs down.

Buracracies were created prior to the advent of computers, and most
often, were embedded in stable operating environments. This means that
the concepts and presuppositions of how the world worked (for their
business or area of expertise) were mostly unchanging. Because of this,
if a "machine" was created that could handle the more or less unchanging
complexity, it would be able meet the goals of the organization
(profitability, country management, religious indoctrination, etc.).

Mechanistic organizations are built around the idea of creating
specialized parts, related in very specific, unchanging ways. For
instance, accounting is separated from contracts, which is separate from
operations, which is separte from logistics, etc. Policies and
procedures are put in place to regulate the interaction between the
groups. Only those near the top of the organization have access to all
the information necessary to coordinate the various parts. This
information has been "filtered" through both specialized forms and middle
management synthesis to provide (hopefully) only that information which
top management cares about.

This type of communication structure allows the organization to maintain
"control," in Ashby's law of requisite variety sense, over the operating
environment; meaning management could effectively meet their
organizational goals. When a change does occur, the response of the
organization is not to reorganize the machine, but simply to "add-on" a
new function. This is akin to "adding" an air-conditioner to a car
instead of changing the way the engine works.

This type of communication and control structure is effective when
dealing with an unchanging operating environment. The problems, and the
pathologies occur when a mechanistic organization is embedded in a
dramatically changing operating environment. Too often, based on changes
in the operating environment, certain specialized parts of the
organization lose their purpose in life, and have to fight for survival.
A good example of this today can be found in government printing and
storage offices. When a part of the organization is concerned with
maintaining its survival, it will become more political, more information
hoarding and more power hungry than is already the case in bureaucracies.

Previously, it was not possible to have strong horizontal communications
in an organization. Groupware, intranets with back-end databases, and
document imaging makes horizontal communications infinitely more
possible. In for-profit businesses, horizontal communications will
become necessary, as dramatic changes in many markets have allowed new,
smaller, more agile companies to come into the competition. Because
these new companies are already employing new technologies, they will not
find aversion to them when they are bigger. Those larger companies are
already experiencing serious changes in how they do business. Not
because they want to improve communication, but because of economic
implications of not doing so. They need to reduce overhead. This is
usually done by reducing middle management. To do so, top management
must essentially give up control for most of the operational information
to lower level people. Groupware and other types of technology makes
this possible.

The thing that has to change in bureaucracies is the way information is
viewed. Currently, information is gathered individually and stored
individually. Information that is desseminated can lead to its use
against you. Network drives are seldom used for sharable information.
As long as there are defensive mechanisms embedded in everyday patterns
of communication that are centered around how information is used,
groupware will have a rough time. I have found that Chris Argyris'
approach works best for this type of situation, although it is far fro
easy to implement. This course is sort of like swimming upstream. You
are trying to change strongly embedded negative feedback systems. Unless
there is clear economic reasons for doing so, don't bother.

The other course is to attempt to incorporate groupware within the
system. Meaning, if you understand the power centers (political
perspective, also in Morgan, 1986), couch groupware in a way that will
enable one of the high-level types to see personal benefit in his (I
assume it's a he if we are high enough in an organization of this type)
stature. Controlling the groupware interactions, and recording the
information obtained (easily done with groupware) in a searchable format
would indeed provide a new level of information access for that manager


Noel Dickover
Cybernetics and organizational change guy at LLD