Hi Peter,
I suggest the game is completely changed by the realisation that individuals
and groups are unable, due to biological limitations, to understand and
predict the dynamic behaviours of situations involving multiple feedback
loops.
The lack of this ability to predict the dynamic behaviours over time of
multiple aspects of design situations and their outcomes is a limiting
factor on the competence of design activity.
Where this lack of ability is found (and it seems to be true for
multifeedback loop situations) then, stakeholders/ constituencies' belief
in agreement on a design problem and strategies for solution, is not an
indicator of competence in design or of the identification of a satisfactory
design solution.
I suggest the above facts if true (and they seem to be) result in the
collapse of the underlying assumptions of most of the methodological,
theoretic and philosophical foundations in the literature relating to
complex design. Ashby, Meadows and Forrester being exceptions.
Best wishes,
Terry
===
Dr Terence Love FDRS, AMIMechE, PMACM, MISI
[log in to unmask] Mob: +61 434 975 848
Senior lecturer,
Researcher, Social Program Evaluation Research Unit
Dept of Psychology and Social Sciences
Edith Cowan University, Western Australia
Senior Lecturer, Dept of Design
Curtin University, Western Australia
Honorary Fellow, Institute of Entrepreneurship and Enterprise Development
Management School, Lancaster University, UK
===
-----Original Message-----
From: PhD-Design - This list is for discussion of PhD studies and related
research in Design [mailto:[log in to unmask]] On Behalf Of Peter
Jones | Redesign
Sent: 11 February 2012 22:21
To: Dr Terence Love
Subject: Re: Wicked Problems
Ken, thanks for changing the title, and the replies Terry, Ken, et al -
Methodology recapitulates ontology.
Without extending the background any further, I will indicate that in our
Strategic Foresight & Innovation program we teach the Meadows system
dynamics approaches, Gharajedaghi/Ackoff school of social systems and
idealized design, and Christakis/Warfield's methodology of dialogic design
science. I am also starting to develop lessons in service systems (also a
"science") as formulated by authors in ISSS and IBM (current president David
Ing is a Toronto collaborator of mine). I know the strengths and weaknesses
of the different schools of thought.
Systems research has always, always been about philosophy. Norbert Wiener
developed Cybernetics as a response to his information theories being
ignored in favor of Shannon and Weaver in his time. His Human Use of Human
Beings was pure ethical philosophy, he was essentially making the same
warnings as Jacques Ellul. Ross Ashby's work on control theory absolutely
defines the Christakis SDD process, except socially, not as an reductionist
system approach but a change-driven approach. West Churchman's work on
inquiring systems describes 5 inquiry or reasoning processes. Traditional
system dynamics is very much analytic-deductive, whereby feedback loops have
to be identified, operationalized, measured before controlled. This is but
one inquiry system. SDD is constructed to draw on all 5 (Churchman's 5th or
systems approach is inquiry of multiple perspectives and toward human
progress).
Again, methods are not theory, and the social theories of human behavior and
interaction draw from different philosophical stances, but can use similar
methods. But the ontological position one has toward design absolutely
determines the kinds of problem systems you see and you're willing to take
on. There are two points that clarify the difference:
Donella Meadows is our classic system dynamics references, yes? She would
agree with your position on feedback loops, but then also say it doesn't
matter for intervention. Her famous paper on the 12 Places to Intervene in a
System: http://en.wikipedia.org/wiki/Twelve_leverage_points basically shows
that the top four most powerful change drivers (again, "design") are social
decisions that must be made by stakeholders, not technical factors
identified in analysis and modeling. The most significant is power to
transcend paradigms. Social systems design aims at these fundamental
agreements, because even if you get feedback loops "right" you never get
them all, and they change by the time you've analyzed your runs. Look at
Afghanistan:
http://www.comw.org/wordpress/dsr/wp-content/uploads/2009/12/afghanistan-130
0.jpg
Two, in social design we follow Hasan Ozbekhan's principle. It's another
ontological position - that it is unethical to design change programs
without the complete engagement of the stakeholders in the problem system.
Ozbekhan published only a few papers, but he's considered the originator of
normative planning, a process we teach in SDD engagements and OCADU. Even
before Rittel defined wicked problems Ozbekhan wrote about problematiques -
in particular the global problematique of 49 continuous critical problems
that evolve and overlap, obscuring initial causes and demonstrating
misleading effects. These 49 are still with us today. Christakis published a
retrospective inquiry on the problematique that tells much of this story:
http://www.globalagoras.com/files/ANCretrospective.pdf
OK, three. If we elevate our discussion above methodology, to ontology and
paradigms, we'll find some divergence, which is fine. It's always been there
in the systems and design fields. I am a constructivist/interpretivist in
most research, because of the messiness of human-centered problems. But in
designing social systems design that stance and the selection of methods
depends on the problem and its goals. Change outcomes are not manageable by
predictive or reductionist methods, as the Club of Rome discovered with the
Limits to Growth. If research is oriented to Understanding (as dialogic
approaches are by their very nature), they are also in conflict with
positivist or reductionist mindsets. These are not irreconcilable, but our
epistemological cultures determine more of our scientific (and design)
process and method than our desire to resolve action in wicked problem
systems in the best ways "for that situation." Again, an agreement.
In our classes, students learn the applicability of methods by doing. from
the learner problems selected, and let the problems determine their planning
and methods. (If they choose system dynamics methods, we have them do causal
loop mapping, not system modeling, it’s an interdisciplinary program with
mixed backgrounds and no time for building math models in class).
Thanks for the opportunity to share some of these ideas.
Peter
Peter Jones, Ph.D.
Associate Professor, Faculty of Design
Sr. Fellow, Strategic Innovation Lab (sLab)
OCAD University
205 Richmond Street West, Toronto, Canada M5V 1V6
http://designdialogues.com
|
