“Dynamic Systems are an approximation that should help solve problems to the Management and professionals …The Solution to small problems yield small rewards Very often the most important problems they are a little less difficult to manage than less important … Many people predetermined mediocre results when setting very low initial goals … The attitude must be one of business design … The expectation should be of greater improvement … The attitude that the goal is to explain the behavior, which is very common in circles academic, it’s not enough. The goal should be to find Management Policies and Organizational Structures that lead more success ”
Jay Forrester – Dynamic Systems – MIT
“The Experience is a very expensive school”
Benjamin Franklin
“Experience is something that you get right after you do not need it”
Anonymous
The constant of modern times is change
Many times our best effort to solve a problem actually makes it worse.
Henry Adams, a perceptive observer of the great changes since the industrial revolution, formulated the Law of Acceleration describes the exponential growth of technology, population and production. Adams believed that the radical changes in society induced by these forces would require a new mental change in society.
Many philosophers, scientists and managers have since studied these phenomena, and many of them have turned to the development of Thinking Systems – the ability to see the world as a complex system, where you have to understand that you can not do one thing without affecting to others and that everything is connected with everything.
If people had a way of seeing the world around in a holistic way, they could act in consonance with the long-term interests of the system as a whole, being the way how the development of thinking systems are crucial for the survival of humanity .
From there, Dynamic Systems are born as a method that helps learning in complex systems.
Learning complex dynamic systems requires much more than technical tools to create mathematical models. Dynamic systems are fundamentally interdisciplinary.
The behavior of complex systems, dynamic systems, is based on dynamic nonlinear theory and feedback control developed in mathematics, physics and engineering. And as those tools are applied to human behavior, as well as to physics and technical systems, dynamic systems are also derived from cognitive psychology, sociology, economics, other social sciences.
To build models of dynamic systems to solve real-world problems, we must learn how to work effectively with groups that implement these systems policies, and how to catalyze sustained change in organizations.
Successful learning approaches about complex dynamic systems require: – tools to represent mental models, which hold about the nature and difficulty of the problem we will study; – Formal models and simulation methods to test our mental model of study, design of new policies and practice of new skills; and finally, – Methods that sharpen our reasoning of scientific skills, improve the groups of processes, and that overcomes defensive routines for individuals and work teams.
“As we found in a complication of contagious diseases, that when applying a remedy for an ulcer, it could cause another ulcer; and that in trying to remove a symptom of the disease could produce another ”
Sir Thomas More
When you are confronting a problem of a complex social system, with things that you do not feel satisfaction with or with something that you want to fix, you can not just take a step and fix that something with a lot of hope that you will find the problem. You can not interfere in a part of a complex system from the outside without being certain of the risk or disaster that it may cause in some other part or remote part of the system. If you want to fix something you are first obliged to understand, the whole system … Intervening in parts is a way to cause problems.
Cause- Effect: the inventory is very high because the sales inexplicably change. Sales change because the competitors lowered prices. Competitors lowered prices because … Such level of explanation of these events can be extended indefinitely, in an unbreakable Aristotelian chain of cause and effect, until we arrive at a first cause, or probably lose interest along the way.
Systems react to your solutions: as your sales grow, competitors cut prices and sales fall again. Yesterday’s solutions are today’s problems. This is what is called feedbak- interaction, answers – the results of our actions will define the situation by which we face the future.
However, people tend to resist, this policy of resistance is caused because we often do not fully understand the ranges of feedbacks-interactions, results-operating in the system. As our actions alter the state of the systems, other people react to restore and balance what we have moved.
Our actions can also trigger collateral effects.
When we take actions then there are several effects: primary or intentional effects are those that benefit the system, the collateral effects are those that damage the system.
In general, the effects tend to be distant in time and space given the complexity of the real world, as in business, society, ecosystems.
To avoid a policy of resistance and find levels of leverage, we need to expand the limits of our mental models, in order to realize these, understand the implications of the results-feedbacks-created by the decisions we make.
This is to understand about the structure and dynamics of the immense and complex system in which we are involved.
The art of modeling dynamic systems is to discover and represent the feedback-interactions, result- of the processes, that together with the structure of the stock-accumulation- and the flow -flow- the times of delays- in everything- and the lack of linearity, determine the dynamics of the systems. In fact, the complexity of the behaviors usually comes from the interaction-feedback- between all the components of the system, and not from the complexity of the same components.
All dynamics come from the interactions of two types of feedback-interaction cycles – one positive (self-reinforcing cycle) and the other negative (self-corrective cycle).
Positive cycles tend to reinforce or amplify whatever is happening in the system. Positive cycles are all processes that generate their own growth.
Negative cycles react instead in opposition.
Since the dynamics depend on the interactions-responses-feedback-so all the learning depends on the feedback. We make decisions that alter the real world, we obtain information exchanged from the real world and using the new information we review and understand the world and the decisions that we make bring to our perception the state of the systems close to our goals.
The interaction – feedback – as such in its whole is pervasive and as a fundamental aspect the behavior is invisible as the air we breathe. Literally, speaking of behavior, we do not know much about our own behavior, but we do know about feedback-interaction-about what this entails and the results that are generated.
So what you have to study and learn are the interactions of the cycles, the iterative cycles of invention, observation, reflection and action.
Learning in an explicit feedback of cycles and interactions that has been practiced in management tools such as the Total Quality Management, where Deming called the cycle (PDCA- Plan, Do, Check, Act), is the heart of the improvement processes in Quality Management.
Description of mental models: mental models are widely discussed in psychology and philosophy. Different theorists describe the mental models as collections of routines and standard operating procedures, scripts of possible selective actions, cognitive maps of a domain, typologies of categorization of experiences, logical structures of interpretation of languages and individualized attributions that we find in daily life.
All the decisions we make are based on models, usually in mental models. In dynamic systems, the term mental model includes our beliefs about the connections of causes and effects that describe how a system operates, with the limits of the model and the relevant horizon time to consider our consideration of the problem.
Most of us do not appreciate the omnipresence and invisibility of mental models, on the contrary we naively believe that our senses reveal the world as it is. On the contrary our world is actively built (modeled) by our senses and brain. The illusions are extremely powerful.
Research shows that the neural structure is responsible for the ability to see illusory contours, where the optic nerve and areas of the brain are responsible for the processing of visual information.
Usually we do not realize that these mental models exist.
Interaction information about the real world not only alters our decisions within the context of the existing framework and decision rules but also the feedback-feed backs alter our mental model. If our mental models change we change the structure of our system, creating different decision rules and new strategies.
Barriers to Learning: Interactions with the real world can stimulate changes in mental models. Learning involves new understanding or re-branding of the new situation and leads to new goals and new decision rules, not just new decisions. In the real world, knowledge becomes obsolete so mental models must change.
Even though today in the world we are faster and more astute and we learn faster than before, the rates of organizations and corporations of failures are high. Today the rate of change in our systems are much faster and the complexity much higher. Learning delays due to pressure problems remain long. It lacks the ability to run experiments and the delays between interventions and results are longer. The rate of change is accelerated through society, while learning remains slow, uneven and inadequate.
The main way in which learning cycle and interactions can fail, is because dynamic systems are complex, contains imperfect information about the state of the real world, are puzzling, contain ambiguous variables, there is a poor reasoning of scientific skills, failures in implementation, and the lack of perception of the interactions, which sink our ability to understand the structure and dynamics of complex systems.
Dynamic Complexity: dynamic systems are complex, they emphasize multiple cycles, multiple states, non-linear characters of systems of interactions – feedback – where we live. The decisions of any agent form not only one but several cycles of interactions (feedbacks) that operate in any given system.
These cycles react to the decisions of the decision makers in both ways: one anticipated and one not anticipated; these can be positive cycles as well as negative cycles, and those cycles can contain many stocks (accumulations), states of variables and many non-linear forms.
Most people think that the complexities are due to the terms of the number of components in a system or the number of combinations one should consider when making a decision. The Dynamic Complexity in contrast, can come from a simple system with a low number of combinations.
Dynamic Complexity comes from the interaction (Feedbacks) between agents over time
The timing of delays between making a decision and the effect it has on the state of the system is commonly a problem generator. Obviously, delays reduce the number of times one can cycle around a learning cycle, decreasing the ability to accumulate experience, test hypotheses, and improve.
For example, reducing the error of an operator in a specific task in complex processes with long times of delays, such as product developments, had improved by half in the life cycles of several years or more.
Dynamic complexity not only decreases learning cycles, it also reduces the learning gained in each cycle.
The existence of multiple interactions and feedbacks means that it is difficult to maintain other aspects of the system constant and to isolate them from the effect of the variable of interest. Many variables change simultaneously, confusing the interpretation of system behavior and reducing the effectiveness of each cycle around the learning cycle.
Delays also create instability in dynamic systems. Adding delay times in negative interaction cycles creates the tendency for the system to oscillate. Many systems, such as the construction of buildings up to driving a car, contain delay times, between the initiation of control of the action and its effects on the state of the system. As a result, decision makers often continue to intervene to correct apparent discrepancies between the desired state and the current state of the system, even after taking sufficient corrective actions to restore the balance of the system.
Oscillation and instability reduce our ability to control by confusing variables by distinguishing cause and effect, and then decrease the rate of learning.
Limited Information: In the Real World we experience many filters, No one knows the current sales of a corporation, the current rate of production, or the true value of accumulations at any given time. On the contrary, one receives estimates of that data based on averaged examples and untimely measures. The act of measuring introduces distortions, delays, prejudices, errors and other imperfections, some known and others unknown and unrecognizable.
Among all the act of measuring is an act of selection.
For example, the price of many goods in the economic system does not include the cost of the waste resource or environmental degradation, these externalities receive little weight in decision-making.
Changes in our mental models are constrained by what we previously decided to define the measure and address this. Seeing is believing and believing is seeing. They interact one in another.
The eyes see only what the mind is prepared to understand.
The self-reinforcement of the interactions between expectations and perceptions has been repeatedly demonstrated in a great variety of experiments studied.
However, the mutual feedback between expectations and perceptions limits learning, blinding us of the anomalies that could challenge our mental model.
Confusion of variables and ambiguities: to learn one must use the limited and imperfect information that is available, in order to understand the effect of our own decisions, and thus be able to adjust our decisions and align the state of the systems with our goals, and so we can review our mental models and redesign the system itself. Thus, much information that we have is ambiguous, Ambiguity comes because the changes in the state of the system resulting from our own decisions are confused with simultaneous changes in the hosting of other variables.
The number of variables that could affect the system would vastly overwhelm the available data, to rule out alternative theories and competitive interpretations. The identification of the problem is harassed by both: qualitative and quantitative approximations. In the qualitative field, ambiguity comes from linguistic ability supported by multiple meanings.
In the quantitative field, engineers and economists have long fought with the problem through the identification of the structure and parameters of a system of their observed behavior.
In order to draw inferences from the data described by the econometric texts, it is necessary to make fanciful assumptions.
Limited Rationality and Lack of Perception of Interactions (Feedbacks). Dynamic complexity and limited information reduce the potential for learning and development, limiting our knowledge of the real world. But how wisely do we use the knowledge we have? How do we process the information? We process it in the best way and we make the best decisions we can? Unfortunately the answer is not.
Humans are not only rational beings, but we weigh the possibilities coldly and judge the probabilities. Emotions, reflexes, unconscious motivations and other non-rational or irrational factors all play a large role in our behavior and ways of judging.
But even when we find time to reflect and deliberate we do not behave in a totally rational way.
So brilliant that the human mind is, the complexity of the real world shrinks our cognitive abilities. Herber Simon, has better articulated the limits of the ability of decision-making, in his famous principle of limited rationality, with which he won the Nobel Prize in Economics in 1979.
“The ability of the human mind to formulate and solve complex problems is very small compared to the size of the problem whose solution is required by objective rational behavior in the real world, or even by a reasonable approximation for such objective rationality” Simon 1957, p 198
To face with these corrals complexities of the real world, pressure of time and limited cognitive capacities, we are forced to fall into the procedure of memory, habits, rules of chance and simple mental models to make decisions.
Even though sometimes we do our best we can, limited rationality means that we frequently fail, limiting our ability to learn from experience.
Experimental studies have shown that people still behave poorly in systems with even modest levels of dynamic complexity.
These studies have suggested to me that the dysfunction observed in dynamic complexities derives from the lack of perception in the interactions (feedbacks).
The mental models that people use to guide decisions are dynamically deficient. People generally adopt events based on open cycles of causality, ignoring processes of interactions (feedbacks), failing to appreciate: times of delays between actions and responses; and with the report of the information, they do not understand the accumulations (stocks) and flows (flows); and they are insensitive to the non-linearity that can alter the strengths of different cycles of interactions (feedbacks) that a system contains.
Subsequent experiments show that the greater the complexity of the dynamic system environment, the worse it is for people to do related to their potential.
The robust lack of interaction perception (feedback) and the poor development that they cause are due to two basic and related deficiencies in our mental model. First, our cognitive map of the causal structure of systems are vastly simplified compared to the complexity of the systems themselves.
Second, we are disabled in correctly inferring the dynamics of everything except that of a simple causal map. Both are directly consequences of limited rationality, such as, the limitations of attention, in memory, in remembering, limitations in the capacity to process information and the time constrained by human decision making.
Cognitive Maps Imperfect: Causal Attributes is the central feature of the Mental models. We create and update the cognitive maps of causal connections through entities and actors. Dorner (1976) found that people tend to think of a simple causal series form and had difficulties with systems in the side effects and the multiple causal pathways (much less in the feedback cycles).
Similarly, experiments in causal attributions show that people tend to assume each effect due to a simple cause and frequently stop looking for explanations when the first cause is found.
Heuristic methodologies (trial and error) that we use to judge causal relationships systematically lead to cognitive maps that ignore feedbacks (interactions), multiple interconnections, non-linearities, delay times and other elements of complex dynamics.
Within a causal area, people use various causality clues including temporal and spatial approximations of cause and effect, temporal precedence of causes, covariations and similarities of cause and effect. That heuristic methodology leads the difficulty in complex systems, where cause and effect is often distant in time and space, where the actions have multiple effects and where the delays the distant consequences are different from the approximate effects.
Multiple feedbacks in complex systems cause several variables to be correlated with one another, confusing the tasks of judging the causes. However, people are poor in judging these correlations.
We have great difficulty in the presence of errors of chance, of non-linearities, and of negative correlations of discovering true relationships.
A fundamental principle of dynamic systems states that the structure of systems rises from their behavior. However people have a great tendency to attribute to the behavior of others the disposition preferably than to the situational factors, that is, to the character and especially to imperfect characters preferably that to the system in which those people are acting.
The tendency to blame people preferably is that the system is a form that psychologists call “Attribution of Fundamental Error” Ross (1977), different people in complex systems placed in the same place tend to behave in this same way.
The attribution of the individual behavior and of a special circumstance preferably that to the structure of the system deconcentrates our attention of the high level of leverage, where the redesign of the system or of the policy of government can have significance, substantial and beneficial effects in performance.
When it is attributed to the behavior of people preferably that the structure of the system, the focus of management becomes pure guilt, and guilt prela instead of the design of the organization, where ordinary people can achieve extraordinary results.
Erroneous Inference about Dynamics: Even if our cognitive maps were perfect, learning, especially the double cycle of learning, could be very difficult. Using a mental model to design a new strategy or organization, inference must be made about the consequences of decision rules that have never been addressed and for which we do not have data. To do this requires intuitive solutions of a high-caliber nonlinear differential equation, a task that far exceeds human cognitive abilities in everything except a simple system.
Poor Development in some tasks is due to the inability to make reasonable inferences about the dynamics of the systems as opposed to perfection and complete knowledge of the structure of the system.
People can not mentally simulate even the simplest system of interaction-feedback-the first positive linear order of interaction cycle.
The dysfunctions in dynamic systems comes from the lack of perception of the interaction of the structure in the environment. A Rich Mental Model that captures those resources of complexity, can not be used reliably to understand the dynamics. Dysfunctions in complex systems come from the lack of mental simulation – lack of interaction perception of the dynamics. Those two limits in rationality must be about making effective learning happen. A perfect mental model without a simulation capacity yields very little insight (internal ideas), a calculation to reliably infer about dynamics systematically yield erroneous systems when simplistic models are applied.
Ramon Marcel Arellano Baez 