Here's a post from Jim Coplien <cope@research.bell-labs.com> on the
patterns discussion list that resonates with this thread (although it
doesn't directly address the question):
There's a group that's met several times here and there around the
world, most recently called the Villeneuve Group, which takes many
of its cues from the Santa Fe Institute. There's a lot of Alexander
influence in the group, and it's an unbelievably eclectic group. I
was with them last time they met, and they were toying with the idea
that my organizational patterns {see next paragraph -- Dwig} might
be the strange attractors that give rise to organizational
structure. The group firmly believes, as I do, that most real
organizational dynamics are formally chaotic, and that any real
structure (at least, structure that's important) is emergent.
That was my turn-on to organizational patterns -- our studies of
even the best ISO9000-certified organizations showed only weak
correspondence between their documented processes and empirical
practice. Furthermore, those organizations showed highly variable
behavior across multiple process enactments. I started asking: so
what is it in these organizations that's stable? That led to
several years of organizational study using sociometric techniques,
and the organizational patterns
(http://www.bell-labs.com/~cope/Patterns/Process/) are, to a large
degree, documentation of the structural patterns we found in the
sociograms -- structure, geometry. That, of course, corresponds
directly to three-space, as there are close ties between social
coupling and office layout in professional organizations (Thomas
Allen's research at Sloan School in the 1960s and 1970s). And the
organization structure -- the instrumental structure, that is, not
the thing on the org chart -- maps onto the patterns in the code
{in the physical buildings, people, etc. -- Dwig}
(Conway, "How do committees invent," Datamation 14(4), April, 1968).
This whole lesson ties into organizational learning, and goes back
to your context question. Is there reality outside of perception?
I myself am in the Berkelean "esse is percipi" tradition, "to be
is to be perceived"... I guess you can interpret that either in the
sense you presented in your mail, where you can only see the patterns
if you adopt the right context to bring them into your mind (or
if you translate time into space, for example); or you can say
that adopting a new perception framework actually changes
reality ("Software Development as Reality Construction," Resier
et al, a delicious book). The latter is a good analogy for
organizational learning. Most powerful patterns are awfully
simple. Most organizations know most of the powerful patterns
("have an architect," "don't do geographically distributed
development," "adding people to a late project makes it later")
but they simply don't follow them. This bespeaks some blindness
on the part of the organization to its own reality. It takes a
major problem, a failure (all that stuff we've been talking about
in anti-patterns) to uncover the organization's blindness to itself.
That opens opportunities for learning, for being able to "see"
the patterns. It's an ages-old process (think of it in terms of
Aristotelean tragedy...)
Don Dwiggins "All models are false,
SEI Information Technology but some are useful"
dwig@earthlink.net -- George Box, "Statistics for Experiments"
--Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>