Page 163 -

- - - - - -

132 Chapter 4 Design for collaboration and communication

Table A: Menu items for initiating a new conversation.
Request Sender wants receiver to do something.
Offer Sender offers to do something, pending acceptance.
Promise Sender promises to do something (request is implicit).
What if Opens a joint exploration of a space of possibilities.
Inform Sender provides information.
Question A request for information.
Note A simple exchange of messages (as in ordinary E-mail).
Figure 4.1 4 Menu items for initiating a conversation.

Thus, the Coordinator was designed to provide a straightforward conversa-
tional structure, allowing users to make clear the status of their work and, like-
wise, to be clear about the status of others' work in terms of various commitments.
To reiterate, a core rationale for developing this system was to try to improve
people's ability to communicate more effectively. Earlier research had shown
how communication could be improved if participants were able to distinguish
among the kinds of commitments people make in conversation and also the time
scales for achieving them. These findings suggested to Winograd and Flores that
they might achieve their goal by designing a communication system that enabled
users to develop a better awareness of the value of using "speech acts." Users
would do this by being explicit about their intentions in their email messages to
one another.
Normally, the application of a theory backed up with empirical research is re-
garded as a fairly innocuous and systematic way of informing system design. How-
ever, in this instance it opened up a very large can of worms. Much of the research
community at the time was incensed by the assumptions made by Winograd and
Flores in applying speech act theory to the design of the Coordinator System.
Many heated debates ensued, often politically charged. A major concern was the
extent to which the system prescribed how people should communicate. It was
pointed out that asking users to specify explicitly the nature of their implicit speech
acts was contrary to what they normally do in conversations. Forcing people to
communicate in such an artificial way was regarded as highly undesirable. While
some people may be very blatant about what they want doing, when they want it
done by, and what they are prepared to do, most people tend to use more subtle
and indirect forms of communication to advance their collaborations with others.
The problem that Winograd and Flores came up against was people's resistance to
radically change their way of communicating.
Indeed, many of the people who tried using the Coordinator System in their
work organizations either abandoned it or resorted to using only the free-form
message facility, which had no explicit demands associated with it. In these con-