Michael McMaster wrote,
>I would say, at a different scale, that there are *only* incomplete
>conversations. None has a clear beginning nor end and none has an
>author that is the sole source. A conversation is an emergent
>process that requires simultaneous co-emergence (to qualify for my
>operational definition) and will be always incomplete on that
>account. It will also be incomplete in that understandings,
>information and assessment of status will all be different for each
>participant at different times.
I'm probably missing your point about scales, but I beg to differ. In the
conversations for action model proposed by Winograd & Flores there is a
very explicit declaration of satisfaction and hence completion of the
conversation. A network of eternally incomplete conversations would lead
to "organizational anxiety" and ineffectiveness
-----
Host's Note: I'll jump in on this one... Maybe, when we look down from
30,000 feet at our organizations, we'll see a web of connected
communications -- I'm reminded of watching videos of converstions with the
sound turned off, played back at high speed. What looked and sounded like
a meeting at normal speed, looks like a rhythmic dance, each person making
physical moves and the others responding. Maybe from 30,000 feet, our
organizations would look like they're suffering rom anxiety and
ineffectiveness. I've certainly seen a fair amount of this in the
organizations I've had a chance to observe.
On the more direct point, I'm very curious about how the language/action
models are useful. They make sense to me looking backwards at
conversations, and it does make sense to me to be clear about what is a
request and to be clear in saying "OK, I'm satisfied with that" but in my
limited attempts to apply them, they have always felt too rigid, too
constraining, too inhibiting.
-- Rick Karash, rkarash@karash.com, host for Learning-org
-----
-- R. Reichard emergent@sirius.comLearning-org -- An Internet Dialog on Learning Organizations For info: <rkarash@karash.com> -or- <http://world.std.com/~lo/>