Systems Methodology
A Holistic Language
of Interaction And Design
Seeing Through Chaos and Understanding
Complexities[1]
February, 2004
Jamshid Gharajedaghi
What is going on? Nobody seems to know. Some are winning while others are losing both for the wrong reasons. A looming prospect of randomness is creating a nasty feeling of insecurity. Even the successful ones are not sure of themselves any more. Games keep changing, but despite all the rhetoric to the contrary, a dominant management culture, by default, keeps reproducing the same non-solutions all over again. We see the world as increasingly more complex and chaotic because we use inadequate concepts to explain it. When we understand something, we no longer see it as chaotic or complex. Maybe playing the new game requires learning a new language.
During
the last 50 years we have been confronted with a dual shift of paradigm. Not only there has been a shift in our
understanding of the nature of the beast from a mindless mechanical system to a multi-minded socio-cultural system. But, there has also been a profound shift in our way of
knowing: from analytical thinking,
the science of dealing with independent sets of variables, to systems thinking, the art and science of handling
interdependent sets of
variables. While the analytical approach has remained essentially intact for
nearly four hundred years, systems thinking has gone through three generations
of change: from Operations Research to Cybernetics and finally Interactive
Design. This evolution has been a
response to challenges of socio-cultural systems. Systems methodology had to
deal not only with the imperative of interdependency and the complexities of
self-organizing systems, but also with the question of purposeful behavior of
multi-minded systems.
The
realities of highly developed social systems are fundamentally different from
those of other forms of living systems.
Members of societies that have outgrown the secure, unifying web of a
paternalistic culture display real choice. Unless we understand the
implications of multi-level purposeful behavior, the multi-minded beast will
out-maneuver any attempt to tame it. On the other hand, contrary to a widely
held belief, the popular notion of a multi-disciplinary approach is not a systems
approach. In fact, the ability to
synthesize separate findings into a coherent whole seems far more critical than
the ability to generate information from different perspectives. Without a
well-defined synthesizing method, however, the process of discovery using a
multi-discipline approach would be an experience as frustrating as that of the
blind men trying to identify an elephant.
Positioned at a different part of the elephant, each of the blind men
reported his findings from his respective position, as ItÕs a snake; ItÕs a pillar; ItÕs a fan; itÕs a
spear!
Consider the futility of trying to make sense of the
whole by using the above story without the prior conception of Òelephant.Ó But
I am sure you experienced no frustration in sorting out the distorted
information and putting it in perspective because the storyteller had already
told us that the subject is an elephant.
It seems we need a preconceived notion of the whole before we can glean
order out of chaos.
A different
version of the same story, found in Persian literature, narrated by Molana
Jalaledin Molavi (Rumi) captures the level of complexity produced when we have
no preconceived notion of the subject.
The story is about a group of men who encounter a strange object in
complete darkness. Since the
storyteller is in the dark himself, he cannot provide a clue about the
object. Here, all efforts to
identify the object by touching its different parts prove fruitless until
someone arrives with a light. The
light, which in this context is a metaphor for methodology, enables them all to
see the whole at last.
RumiÕs version of the story
means that the ability to see the whole somehow requires an enabling light in
the form of an operational systems methodology. Rumi, in his mystical wisdom
proposed that to get the light one needs to tune oneself with the
universe. Yet the operational
meaning of tuning, for our purpose here, is that one should be able to make
his/her underlying assumptions about the nature of the socio-cultural systems
explicitly known and verifiable to oneself. Then one needs to develop an iterative search process that would help to evolve the
initial assumptions until a satisfactory vision of the whole is produced.
Systems Methodology
The version
of systems methodology presented in this paper is a holistic language of
interaction and design developed to face the dilemma of social systems where
the whole is becoming more and more interdependent while the parts display
choice and behave independently.
The methodology gives us a way to see through chaos and understand complexities.
The foundation
of this exciting conception is the interaction of four elements of systems
thinking:
á
Holistic
Thinking, iteration of structure, function and process
á
Operational
Thinking, understanding chaos and complexity
á
Systems
Theories, a socio-cultural view
á
Interactive
Design, creating a feasible whole with infeasible parts
Figure one:
The Four Foundations of Systems Methodology
1: Holistic Thinking
Structure, Function, Process and Context
Analytical thinking assumes that
understanding structure
is sufficient to understand a system. For synthetic thinking function is the key for seeing the whole. The behaviorist, on the other hand,
looks to the process,
the how question, for the necessary answer to define the whole. Each one has
been used as the core concept of a different inquiring system producing a
tremendous amount of information and knowledge.
á
Analysis has been the essence of classical science. The scientific method assumes that the
whole is nothing but the sum of the parts, and thus understanding the structure
is both necessary and sufficient to understanding the whole.
á
Synthesis has been the main instrument of the
functional approach. By defining a system by its outcome, synthesis puts the
subject in the context of the larger system of which it is a part, and then
studies the effects it produces in its environment.
á
Process orientation, on the other hand, has long
been the focus of behavioral science.
It basically deals with the how question.
On
more familiar and practical territory, we could perhaps say that the classical
school of management, with its input orientation, deals basically with
structure. The neo-classical
school, with its notion of "management by objective," is concerned
with functions. And the total
quality movement, with its concern for control, is preoccupied with the
process.
It is my contention that structure,
function, process represent three aspects of the same thing and with the
containing environment they form a complementary set. Together they define the whole or make the understanding of
the whole possible. Structure defines components and their relationships;
function defines the outcomes or results produced; process explicitly defines
the sequence of activities and the know-how required to produce the outcome.
A
set of interdependent variables forms a circular relationship. Each variable co-produces the others
and in turn is co-produced by the others. Which one comes first is irrelevant because none can exist
without the others. They have to happen at the same time. To fail to see the significance of
these interdependencies is to leave out the most important aspect of the
challenge. Therefore, to handle them holistically requires understanding each
variable in relation to the others in the set at the same time. This demands an iterative inquiry
Figure Two: Iterative Process of Inquiry for
Understanding Complexity
Iteration is the key for understanding
complexity. An iterative process
of applying simple rules is at the core of natureÕs mysterious ability to
produce complex phenomena so effortlessly. Iterations of
structure, function, and process in a given context would examine assumptions
and properties of each element in its own right, then in relationship with other
members of the set. Subsequent iterations would establish validity of the
assumptions and successively produce an integrated design. See Figure Two, below.
The
principle of iterative inquiry is reinforced by Singerian experimentalism: there is no fundamental truth;
realities first have to be assumed in order to be learned; truth is not the
starting but the end point of an inquiry. Successive iterations would yield a greater
understanding and more closely approximate the nature of the whole.
2:
Operational
Thinking
Understanding Chaos and Complexity
Complexity is a relative term. It depends on the number and the nature of interactions among the variables
involved. Open loop systems with linear, independent variables are considered
simpler than interdependent variables forming non-linear closed loops with a
delayed response. Key words in the
above statement are closed loop, nonlinear, and delayed response.
The first step for understanding complexity is to appreciate the
iterative and thus dynamic nature of closed loop systems and their
counterintuitive behavior. Consider
the following two simple examples:
1) A saving account in a bank earning simple
10% interest reflects an open loop behavior. Both yearly earnings and the
amount of principal remain constant and total sum would increase at a slow p
Figure Three:
Open Loop System
2)
However, if the savings in the bank were to earn 10%
compound interest, it would represent a closed loop behavior and the money in
the saving account will grow exponentially, doubling
3) seven years. The initial principal of $10,000 would amount to $1,280,000 if left there for 56 years. Compare this amount with $66000 that would be accumulated by simple every interest.
Figure
Four: Closed Loop System
Now, if
the interest rate in the above example would varied according to market
conditions then
we will be facing a nonlinear system. Please
note that in closed loop thinking linear and nonlinear refer to the rate of
change, not the state of a system.
Figure
Five: Linear vs. Nonlinear System
Now let
us look at the dynamic behavior of a simple negative feedback loop (goal
seeking).
Figure 6: Goal-Seeking Behavior
Please
note the counter-intuitive impact (oscillation) of introducing a delay function
to our simple negative feedback loop.
Figure 6,1 counterintuitive impact of delay (oscillation
Now let us consider a
common phenomenon known as positive feedback loop. We know that
it will result in an exponential
growth curve.
Figure Seven: Positive Feedback Loop Producing Exponential Growth Curve
However,
if we just add the impact of carrying capacity, or the imperative of market
potential, and superimpose the reality of a delay function to our simple
positive feedback loop. We would
create a monster, the infamous Òmulti-loop nonlinear feedback system.Ó This is the system that according to
chaos theory produces chaotic behavior.
It explains the collapse of Dotcoms, fiasco of Enron, and faith of
thousands of corporations who pursue a blind short-term growth strategy with no
regard for the limitations imposed by carrying capacity of the system and/or
its environment. The overshoot
and collapse scenario
reflects the cases where the growth strategy has an additional negative impact
on the carrying capacity of the system.
Impact of carrying capacity
Figure Eight: Impact of Carrying Capacity on the Behavior of a System
The point of
emphasis is that the interaction of multiple feedback loops is the prime source
for generating chaos and complexity. Understanding this dynamics is the
essential step to get a handle on the notion of interdependency and
counter-intuitive behavior of social systems.[2]
Unfortunately our
cognitive ability has evolved around assumptions of unidirectional causality or
open loop thinking. It has primarily been concerned with independent variables.
Therefore, we do experience extreme difficulties in visualizing the behavior of
interdependent variables or the outcome of closed loop systems.
According to
Barry Richmond, creator of the i-think model, ÒThe way we think is outdated. As a result, the way
we act creates problems, and then we are ill-equipped to address them because
of the way we think.Ó
Thinking consists
of two activities: constructing mental models and simulating them in order to
draw conclusions and make decisions.
We certainly need help in both accounts.
Apparently our
highly regarded mathematical tools are not doing the job. Otherwise how can we
explain the sorry fact that we have been applying the same set of non-solutions
to the crucial social problems such as drugs, poverty, crime, illiteracy and
maldistribution of wealth for most of the last fifty years?
Stephen Wolfram,
in his book, New Kind Of Science
(2002) has a critical observation:
ÒThe idea of
describing behavior in terms of mathematical equations works well where the
behavior is fairly simple. It almost inevitably fails whenever the behavior is
more complex. Indeed, there are
many common phenomena about which theoretical science has had remarkably very
little to say. Degree of difficulty encounter in mathematical representation of
a phenomenon increases exponentially by the degree of its complexity.Ó (Chapter
one page 3)
He then goes on
to demonstrate how systems too complex for traditional mathematics could yet
obey simple operational rules. He
also shows how remarkably simple iterative computer programs capture the
essential characteristics of complex phenomena.
Operational
Thinking[3] is an ingenious way to overcome the difficulties encountered
in constructing and simulating complex mental models. Relying solely on
mathematical representation for dealing with complex phenomenon has been a
practical nightmare. Combining operational thinking with more manageable forms
of mathematical representation, programs such as i-think software have made it practical to get a
handle on multi-loop nonlinear feedback systems.
It is important
to note that although
multi-loop nonlinear
feedback systems exhibit chaotic behavior, there is an order in this chaos.
Such systems seem to be attracted to a particular pattern of behavior. By operational thinking we can discover
this pattern and recognize the ÒSecond Order MachineÓ (the attractor in action)
that is locking the system to its existing pattern.
The Second-Order
Machine--an implicit set of organizing principles, residing at the core of
organizationÔs collective memory--is most resilient stuff. The triumphant resurgence of old
patterns of behavior despite the concerted efforts of change agents is an
uninterrupted saga of despair.
Unless the implications of these principles (the attractors in action)
are made explicit and dismantled, the nature of the beast will outlive the
temporary effects of interventions.
The pattern recognition is critical for understanding and changing the
undesirable behavior.
To recap,
remember that mapping the dynamic behavior of a system is to capture the
interaction of positive and negative feedback loops. This interaction, in essence, defines the critical
interdependencies among the variables involved. The mapping process would help us change the default setting
and overcome the shortcomings of our cognitive abilities.
3: Systems Theories
A
Socio-Cultural View
I have argued extensively elsewhere (Jamshid
Gharajedaghi 1999) that five systems principles of openness, purposefulness,
multidimensionality, emergent property, and counter-intuitiveness, along with five systems dimensions
define the essential characteristics and the behavior of a socio-cultural
system
Openness means that the behavior of open (living)
systems can be understood only in the context of their environment. Therefore
no problems or solutions can be entertained free of context. However, a
tendency to define the problem in terms of the solution, and a strong
preference for the context-free solution, that is tried and true, keep
producing the same non-solution all over again. Open (living) systems exhibit a tendency toward a
predefined order. Left alone they reproduce themselves. Cultural codes are the
social equivalent of biological DNA.
Self-organization by default will invariably reproduce the existing
order.
Purposefulness. Why
people do what they do is the matter of purpose, that of choice. And the choice has rational, emotional,
and cultural dimensions.
Rational choice is the domain of self-interest, or
the interest of the decision maker, not the observer. A rational choice is not necessarily a wise choice. It reflects only the perceived interest
of the decision maker at the time.
The emotional choice is the domain of beauty and excitement. We do lots of things because they are
exciting or, more precisely, because they are challenging. If the excitement of a good challenge
were not part of our decision criteria, life would be a bore. In other words, setting and seeking
attainable goals is a banal existence.
Culture defines both the cognitive and the
normative behavior of the collectivity.
Just like a high-level computer language that provides default
parameters when the programmer fails to choose one, the culture provides
default values when actors fail to choose one explicitly.
Multidimensionality is probably one of the most potent
principles of systems thinking. It
is the ability to see complementary relations in opposing tendencies. The
mutual interdependence of opposing tendencies is characterized by an Òand Ò instead of an "or" relationship. Unfortunately, for the
majority of cultures, a fallacy has dominated the treatment of opposing
tendencies as a duality in a zero-sum game. Everything seems to come in a pair of opposites:
collectivity/individuality; security/freedom; modernity/tradition,
order/complexity; art/science and so on.
They are cast in such a way that a win for one is invariably associated
with a loss for the other. If X is true then NX cannot be true. This represents an "or" relationship.
Multidimensionality states that lose/lose and win/win as well as
win/lose are possibilities and it denies the fallacy that if x is good more x
is even better.
CounterÐintuitiveness. Social dynamics stand on a level of complexity beyond the
reach of the analytical approach. Counter-intuitiveness means that actions
intended to produce a desired outcome may, in fact, generate opposite results.
Things can get worse before getting better, or vice versa. One can win or lose for the wrong
reason.
To appreciate the nature of
counter-intuitive behavior, one needs to understand the practical consequences
of the following assertions:
á
Cause and
effect may be separated in time and space.
á
Cause and
effect can replace one another, displaying circular relations.
á
An event
may have multiple effects. The
order of importance may shift in time.
á
An effect
may have an independent life of its own.
Removing the cause will not necessarily remove the effect.
Emergent Properties are the property of the whole, not the parts, and thus cannot be
analyzed; they are the product of interactions among the parts. The mere notion of interaction
signifies a dynamic process. In
other words, the emergent phenomenon is a time-dependent state reproduced
continuously online and
real time. Therefore, life, love, happiness,
and success are not a one-time proposition; they have to be reproduced
continuously. If the processes
that generate them come to an end, the phenomena cease to exist as well. They
cannot be stored or saved for future use.
Systems Dimensions[4]. The parameters that coproduce state of a
socio-cultural system are found in the following five dimensions: wealth, power, knowledge, beauty, and
values, in my experience, form a comprehensive set of variables that
collectively describe the organization in its totality.
á
The
generation and distribution of wealth, or the production of necessary goods and services and
their equitable distribution.
á
The
generation and dissemination of truth, or information, knowledge, and understanding.
á
The
creation and dissemination of beauty, the emotional aspect of being, the meaningfulness and
excitement of what is done in and of itself.
á
Formation
and institutionalization of values
for the purpose of regulating and maintaining interpersonal relationships:
cooperation, coalition, competition, and conflict.
á
Development
and duplication of power,
the question of legitimacy, authority, and responsibility or, in general, the
notion of governance.
4: Interactive Design
Creating
feasible whole with infeasible parts
Interactive
design is essentially identified with Russell Ackoff. He explicitly recognizes that choice is
at the heart of human development.
ÒDevelopment is the capacity to choose; design is a vehicle for
enhancement of choice and holistic thinking. Designers seek to choose rather than predict the future.Ó
Interactive
design is about creating feasible whole with infeasible parts. It is both the art of finding
differences among things that seem similar and the science of finding
similarities among things that seem different. Designers try to understand rational, emotional, and
cultural dimensions of choice and produce a design that satisfies a multitude
of functions.
Four distinct
elements of interactive design are: 1) Participation, 2) Formulation of the
Mess, 3) Idealization, and 4) Realization. Each one of these elements adds a unique characteristic to
this captivating process.
Participation
Self-organizing, purposeful,
socio-cultural systems are self-evolving.
They do not simply adapt to their environments but co-evolve with
them. They can change the rules of
interaction as they evolve over time.
However, like all open systems a purposeful socio-cultural system
exhibits a tendency toward a predefined order. Its behavior is guided by an implicit, shared image. In the short term it tends to
approximate and reproduce its pattern of existence very closely. To change this
pattern of behavior the implicit shared image or (the organizing attractor)
needs to be changed. This can only be done by a participative design process. People are more likely to accept an idea when they have had
a hand in shaping it.
Formulation of the Mess
Separation of
defining problems from designing solution is a unique characteristic of
interactive design. According to
Ackoff, ÒWe fail more often not because we fail to solve the problems we face
but because we fail to face the right problem.Ó Problem is defined neither as deviation from a norm
nor in terms of the universal constraints (lack of time, resources, or
knowledge). It is defined as a mess,
an interactive set of problems reflecting the future implicit in the present
operation.
Idealization
The distinctive characteristic of
idealization is the notion of backward planning. It starts with the assumption that the system has been
destroyed overnight and that the designers have been given the opportunity to
recreate the system from a clean slate.
The only constraints are that the outcome be technologically feasible
and operationally viable.
Design is a
process for operationalizing the most exciting vision of the future that the
designers are capable of producing.
It is the design of the next generation of their system to replace the
existing order.
Realization
Successive
approximation is at the core of realizing an ideal design. Realization takes
place in a real-world environment.
Therefore, designers must identify all the constraints that might
interfere with proper implementation of the design. These constraints usually
fall into the following three distinct categories.
Type
I Constraints
Type I
constraints cannot be removed within the existing framework. Such constraints would require
revisions and improvisations of the design in order to create a target design
capable of being implemented.
Target I would be the first approximation of the unconstrained design. If necessary, subsequent approximations
will identify Target II and Target III generations of the desired design. It is critical that Type I constraints
be continuously monitored so that the target design can further approximate the
idealized design as soon as these constraints are removed.
The
realization effort, therefore, will not be a one-time proposition. Successive
approximations of the desired state make up the evolutionary process by which
the transformation effort is conducted.
It may take a number of attempts before the desired design is
implemented.
Type
II Constraints
Type II
constraints are those constraints whose removal will require extensive
preparation. They consist of
activities that consume considerable time and resources, as well as knowledge
and management talent.
These
activities usually involve redesign of the products (if necessary), redesign of
throughput, and redesign of organizational processes. Design of the measurement
and reward system with variable budgeting and target costing seems to be an
integral part of all successful realization efforts. This usually is the most
resource-intensive part of the change effort. For
control purposes, all critical assumptions and expectations about the selected
course of actions must be explicitly recorded and continuously monitored.
Type
III Constraints
Type
III deals essentially with behavioral constraints. These are the constraints that can be removed if designers
so desire. Selling the idea,
removing resistance to change, ensuring acceptance, cultivating support, and
providing training are among the efforts targeted at constraints that are
basically self-imposed. These
constraints, taken together, act as the cultural default of the organization,
and their function is to reinforce the status quo. Without a prior foundation of trust and commitment, the
system would simply refuse to undergo the planned transformation
irreversibly. And in this context,
dissolving the "second-order machine" is the most critical phase of
realizing the design.
Conclusion
The beauty of interactive design and the magic of the iteration of structure, function, and process when combined with the power of operational thinking, and profound understanding of systems principles and dimensions, in my experience, create a competent and exciting methodology that goes a long way in dealing with emerging challenges of seemingly complex and chaotic socio-cultural systems.
References
Ackoff, R. L., Creating the Corporate
Future. New York: John
Wiley, and Son, 1981.
Ackoff, R. L., and Fred E. Emery, On
Purposeful Systems. Chicago:
Aldine-Atherton.1972
Beer, Stafford, Brain of the Firm. Harmondsworth: Penguin Press, 1967.
Boulding, Kenneth E, The image, Ann Arbor: University of Michigan Press,
1956
Churchman, C. West, Design of
Inquiring Systems. New
York: Basic Books, 1971.
Dewey, John, Freedom and Culture. New York: Prometheus Books, Great Books
in Philosophy, 1989.
Forrester, Jay W., ÒCounter Intuitive
Behavior of Social SystemsÓ Technology Review, Vol. 73 No. 3 Jan. 1971, pp. 52-68.
Gleick, James, CHAOS: Making a New Science. Viking Penguin Inc: New York, 1987.
Gharajedaghi, Jamshid, Systems
Thinking, Managing Chaos and Complexity a Platform for Designing Business
Architecture: Boston,
Butterworth Heinemann, 1999.
Richmond, Barry, An Introduction to
Systems Thinking, (ithink software),
High Performance Systems, Inc, 2001
Singer, E. A.,
Jr., Experience and Reflection. C.W.
Churchman ed., Philadelphia:
University of Pennsylvania Press, 1959.
Sterman, D, Business
Dynamics, systems Thinking and Modeling for a Complex World. Irwin, McGraw Hill 2000
Wolfram
Stephen. New Kind of Science Canada: Wolfram Media inc., 2002