andyo@masscomp.ccur.com (Andy Oram) (07/11/90)
I'd like to offer some ideas here about hidden sources of resistance to
groupware (a common topic in this group). If the ideas have been explored
elsewhere, please let me know.
Hardly anybody is complaining about single-user systems
-------------------------------------------------------
Personal computers and workstations have been spreading like wildfire
for almost ten years. Despite occasional gripes about incompatible
file formats or similar hassles, most people are satisfied that they
can:
o Write and print a beautiful document, even though the recipients
cannot reformat or edit the information.
o Generate great tables and graphs from raw statistical data, even
though the recipients cannot translate the output back into raw
form to experiment with it or build on it.
And so on. We must thus look for reasons people DON'T want the
advantages of shared computer data and resources.
The competitive advantage of non-sharable resources
---------------------------------------------------
Imagine a typical professional use of computers: a consultant
bidding on some kind of project. The sequence of activities is
something like:
Analyze
Requirements -------> Write up/Format
Develop bid ---------------> Proposal
Your own data ------->
If your analysis tools let you develop a cost faster, more accurately,
more flexibly, or just more impressive to outsiders, you have a
competitive advantage over other bidders. You definitely do not want
the buyer, much less your competitors, to play around with your raw
data or to examine your tools closely.
Similarly, on just the psychological side, your written proposal is
much more impressive as a whiz-ding laser-printed document than as
pure text that the buyer can handle electronically.
The same incentives to hoard exist within a single organization, too.
See the well-known book by Shoshana Zuboff [1]. So we're going to
have to confont the incentives head-on.
Consensus as the antithesis of competitive advantage
----------------------------------------------------
If I am not being presumptuous, it seems to me that the
collaborative-computing/groupware community wants to change
fundamental assumptions about typical activities. We'd like to model
the activities as shared access to a common set of data and tools:
Requirements ---> |
|
Bidder 1's data ---> |
| Buyer's analysis --->
Bidder 2's data ---> | --> Bidder 1's analysis ---> etc.
| Bidder 2's analysis --->
. |
. ---> |
. |
where the buyer and bidders come to a consensus on what requirements
mean, what data is relevant, what tools are valid, etc. I think we
all know how threatening such a model is. But beyond selfishness,
people avoid the model because it implies a lot more work
communicating and transferring interpretive skills. I'm not sure the
proponents of collaborative computing have figured out how much extra
effort this is.
Intuitive, interpretive knowledge is difficult to manage
--------------------------------------------------------
It's easy to train someone to use computer tools. The hard part is to
understand their output. But this is a prerequisite for productive
and responsible use of shared resources: experienced users must
transfer their skills to everybody else who shares the data base.
Since these skills, of course, are fuzzy and intuitive, a lot of
people don't know how to transfer them. Nor is there any clear
endpoint or quantitative standard for the process.
Conclusion: we must press for responsible understanding of real-life domains
-----------------------------------------------------------------------------
Here we encounter an old, familiar issue: the complexity of the
interactions between technology and social organization. The
prerequisites for sharing computer resources are the same as for any
honest low-tech actitivity.
If we want people to work together, not to be mystified and misled,
and to make the best use of their resources, they must communicate
their beliefs and methods to each other. Probably, this means a
recursive kind of explaining and listening, intensively
person-to-person and unformalizable.
This training in real-life issues is more important than the
particular tools or data structures used to make a collaborative
environment -- but of course, the proper choices of tools and data
structures can make the process go better. In any case, collaborative
computing is not just a peer-to-peer free-for-all. It is a means of
transferring knowledge and building a cohesive community.
Reference
----------
Zuboff, Shoshana, In the Age of the Smart Machine: The Future of Work
and Power, Basic Books, New York, NY, 1984.
Andy Oram
Organization: ccc (Concurrent Computer Corporation)
Digital Communications Route: andyo@westford.ccur.com.{harvard,uunet,petsd}
{harvard,uunet,petsd}!masscomp!andyo
Analog Communications Route: (508) 392-2865