This is available as part of a book offering that is free for community members of Cyb3rSynLabs. Please check here (https://www.cyb3rsynlabs.com/c/books/) for Second Order Cybernetics Essays for Silicon Valley. The e-book version is available here (https://www.cyb3rsyn.com/products/soc-book)
My last post was a review of Mark Graban’s wonderful book, Measures of Success. After reading Graban’s book, I started rereading Walter Shewhart’s books, Statistical Method from the Viewpoint of Quality Control (edited by Dr. Deming) and Economic Control of Quality of Manufactured Product. Both are excellent books for any Quality professional. One of the themes that stood out for me while reading the two books was the concept of Cybernetics. Today’s post is a result from studying Shewhart’s books and articles on cybernetics by Paul Pangaro.
The term “cybernetics” has its origins from the Greek word, κυβερνήτης, which means “navigation”. Cybernetics is generally translated as “the art of steering”. Norbert Wiener, the great American mathematician, wrote the 1948 book, Cybernetics: Or Control and Communication in the Animal and the Machine. Wiener made the term “cybernetics” famous. Wiener adapted the Greek word to evoke the rich interaction of goals, predictions, actions, feedback, and response in systems of all kinds.
Loosely put, cybernetics is about having a goal and a self-correcting system that adjusts to the perturbations in the environment so that the system can keep moving towards the goal. This is referred to as the “First Order Cybernetics”. An example (remaining true to the Greek origin of the word), we can use is a ship sailing towards a destination. When there are perturbations in the form of wind, the steersman adjusts the path accordingly and maintains the course. Another common example is a thermostat. The thermostat is able to maintain the required temperature inside the house by adjusting according to the external temperature. The thermostat “kicks on” when a specified temperature limit is tripped and cools or heats the house. An important concept that is used for cybernetics is the “law of requisite variety” by Ross Ashby. The law of requisite variety states that only variety can absorb variety. If the wind is extreme, the steersman may not be able to steer the ship properly. In other words, the steersman lacks the requisite variety to handle or absorb the external variety. The main mechanism of cybernetics is the closed feedback loop that helps the steersman adjust accordingly to maintain the course. This is also the art of a regulation loop –compare, act and sense.
Warren McCulloch, the American cybernetician, explained cybernetics as follows:
Narrowly defined it (cybernetics) is but the art of the helmsman, to hold a course by swinging the rudder so as to offset any deviation from that course. For this the helmsman must be so informed of the consequences of his previous acts that he corrects them – communication engineers call this ‘negative feedback’ – for the output of the helmsman decreases the input to the helmsman. The intrinsic governance of nervous activity, our reflexes, and our appetites exemplify this process. In all of them, as in the steering of the ship, what must return is not energy but information. Hence, in an extended sense, cybernetics may be said to include the timeliest applications of the quantitative theory of information.
Walter Shewhart’s ideas of statistical control works well with the cybernetic ideas. Shewhart purposefully used the term “control” for his field. The term control is a key concept in cybernetics, as explained above. Shewhart defined control as:
A phenomenon is said to be controlled when, through the use of past experience, we can predict at least within limits, how the phenomenon may be expected to vary in the future. Here it is understood that prediction within limits means that we can state, at least approximately, the probability that the observed phenomenon will fall within the given limits.
Shewhart expanded further:
The idea of control involves action for the purpose of achieving a desired end. Control in this sense involves both action and a specified end.
..We should keep in mind that the state of statistical control is something presumable to be desired, something to which one may hope to attain; in other words it is an ideal goal.
Shewhart’s view of control aligns very well with the teleological aspects of cybernetics. From here, Shewhart develops his famous Shewhart cycle as a means to maintain statistical control. Shewhart wrote:
Three steps in quality control. Three senses of statistical control. Broadly speaking, there are three steps in a quality control process: the specification of what is wanted, the production of things to satisfy the specification, and the inspection of things produced to see if they satisfy the specification.
The three steps (making a hypothesis, carrying out an experiment, and testing the hypothesis) constitute a dynamic scientific process of acquiring knowledge. From this viewpoint, it is better to show them as a forming a sort of spiral gradually approach a circular path to which would represent the idealized case, where no evidence is found in the testing of hypothesis indicates a need for changing the hypothesis. Mass production viewed in this way constitutes a continuing and self-corrective method for making the most efficient use of raw and fabricated materials.
The Shewhart cycle as he proposed is shown below:
One of the criterions Shewhart developed for his model was that the model should be as simple as possible and adaptable in a continuing and self-corrective operation of control. The idea of self-correction is a key point of cybernetics as part of maintaining the course.
The brilliance of Shewhart was in providing guidance on when we should react and when we should not react to the variations in the data. He stated that a necessary and sufficient condition for statistical control is to have a constant system of chance causes… It is necessary that differences in the qualities of a number of pieces of a product appear to be consistent with the assumption that they arose from a constant system of chance causes… If a cause system is not constant, we shall say that an assignable cause is present.
Shewhart continued:
My own experience has been that in the early stages of any attempt at control of a quality characteristic, assignable causes are always present even though the production operation has been repeated under presumably the same essential conditions. As these assignable causes are found and eliminated, the variation in quality gradually approaches a state of statistical control as indicated by the statistics of successive samples falling within their control limits, except in rare instances.
We are engaging in a continuing, self-corrective operation designed for the purpose of attaining a state of statistical control.
The successful quality control engineer, like the successful research worker, is not a pure reason machine but instead is a biological unit reacting to and acting upon an everchanging environment.
James Wilk defined cybernetics as:
“Cybernetics is the study of justified intervention.”
This is an apt definition when we look at quality control, as viewed by Shewhart. We have three options when it comes to quality control:
Source: Measures of Success (Mark Graban, 2019)
Final Words:
Shewhart wrote “Statistical Method from the Viewpoint of Quality Control” in 1939, nine years before Wiener’s Cybernetics book. The use of statistical control allows us to have a conversation with a process. The process tells us what the limits are, and as long as the data points are plotted randomly within the two limits, we can assume that whatever we are seeing is due to chance or natural variation. The data should be random and without any order. When we see some manner of order in the likes of a trend or an outside data point, then we should look for an assignable cause. The data points are not necessarily due to chance anymore. As we keep plotting, we should improve our process, and recalculate the limits.
I will finish off with Dr. Deming’s enhancement of Shewhart’s cycle. This is taken from a presentation by Clifford L. Norman. This was part of the evolution of the PDSA (Plan-Do-Study-Act) cycle which later became famous as PDCA cycle (Plan-Do-Check-Act). This showed only 3 steps with a decision point after step 3.
The updated cycle has lots of nuggets in it such as experimenting on a small scale, reflecting on what we learned etc.
Always keep on learning…
In case you missed it, my last entry was My Recent Tweets:
Note: The updated Shewhart cycle was added to the post after a discussion with Benjamin Taylor (Syscoi.com).
Today’s post is a follow-up to an earlier post, Solving a Lean Problem versus a Six Sigma Problem:
In today’s post, I am looking at “The Ashby Space.” The post is based on the works of Ross Ashby, Max Boisot, Bill McKelvey and Karl Weick. Ross Ashby was a prominent cybernetician who is famous for his “Law of Requisite Variety.” The Law of Requisite Variety can be stated as “Only variety can destroy/absorb variety.” Ashby defined variety as the number of distinguishable states of a system. Stafford Beer used variety as a measure of complexity. The more variety a system has the more complex it is. An important concept to grasp with this idea is that the number of distinguishable states (and thus variety) depends upon the ability of the observer. In this regard, variety of a system may be viewed as dependent on the observer.
Max Boisot and Bill McKelvey expanded upon the Law of Requisite Variety and stated that only complexity can destroy complexity. In other words, only internal complexity can destroy external complexity. If the regulatory agency of a system does not have the requisite variety to match the variety of its environment, it will not be able to adapt and survive. Ashby explained this using the example of a fencer:
If a fencer faces an opponent who has various modes of attack available, the fencer must be provided with at least an equal number of modes of defense if the outcome is to have the single value: attacked parried.
Boisot and McKelvey restated Ashby’s law as – the range of responses that a living system must be able to marshal in its attempt to adapt to the world must match the range of situations—threats and opportunities—that it confronts. They explained this further using the graphical depiction they termed as “the Ashby Space.” The Ashby Space has two axes, the horizontal axis represents the Variety of Responses, and the vertical axis represents the Variety of Stimuli. Ashby’s law can be represented by the 45˚ diagonal line. The diagonal line represents the requisite variety where the stimuli variety matches the response variety. To adapt and survive we should be in on the diagonal line or below. If we are above the diagonal line, the external variety surpasses the internal variety needed and we perish. Using Ashby’s fencer example, the fencer is able to defend against the opponent only if his defense variety matches or exceeds that of the opponent’s offense variety. This is shown below.
Boisot and McKelvey also depicted the Ordered, Complex and Chaotic regimes in the Ashby space. In the ordered regime, the cause-effect relationships are distinguishable and generally has low variety. The complex regime has a higher variety of stimuli present and requires a higher variety of responses. The cause-effect relationships are non-linear and may make sense only in hindsight. The chaotic regime has the most variety of stimuli. This is depicted in the schematic below. Although the three regimes may appear equally sized in the schematic, this is just for representational purposes.
The next idea that we will explore on the Ashby Space is the idea of the Adaptive Frontier. Ashby proposed a strong need for reducing the amount of variety from the external environment. He viewed this as the role of regulation. Ashby pointed out that the amount of regulation that can be achieved is limited by the amount of information that can be transmitted and processed by the system. This idea is depicted by the Adaptive Frontier curve. Any variety that lies outside this curve is outside the “adaptation budget” of the system. The system does not have the resources nor capacity to process all the variety that is coming in, and does not have the capacity to allocate resources to choose appropriate responses. The adaptive frontier is shown in the schematic below as the red dotted curve.
Combining all the ideas above, the Ashby Space can be depicted as below.
Boisot and McKelvey detail three types of responses that a living system might follow in the presence of external stimuli. Consider the schematic below, where the agent is located at “Q” in the Ashby Space, which refers to the stimuli variety, X.
Final Words:
The Ashby Space is a great representation to keep in mind while coping with complexity. The ability of a system to discern what is meaningful and what is noise depends on the system’s past experiences, world views, biases and what its construes as morals and values. Boisot and McKelvey note that:
Not everything in a living system’s environment is relevant or meaningful for it, however. If it is not to waste its energy responding to every will-o-the wisp, a system must distinguish schema based on meaningful information (signals about real-world regularities judged important) from noise (meaningless signals). Note that what constitutes information or noise for a system is partly a function of the organism’s own expectations, judgments, and sensory abilities about what is important —as well as of its motivations— and hence, of its models of the world. Valid and timely representations (schema) economize on the organism’s scarce energy resources.
This also points to the role of sensemaking. As Karl Weick notes, “an increase in complexity can increase perceived uncertainty… Complexity affects what people notice and ignore… The variety in a firm’s repertory of beliefs should affect the amount of time it spends consciously struggling to make sense. The greater the variety of beliefs in a repertoire, the more fully should any situation be seen, the more solutions identified, and the more likely it should be that someone knows a great deal about what is happening.”
The models or representations we construct to represent a phenomenon do not have to be as complex as the phenomenon itself, just like the usefulness of a map is in its abstraction. If the map was as complex as the city it represented, it would become identical to city, with the roads, buildings etc., an exact replica. The system however should have the requisite variety. The system should be able to filter out unwanted variety and amplify its meaningful variety to achieve this. The agent must wait for “meaningful” patterns to emerge, and keep learning.
The agent must also be aware to not claim victory or “Mission Accomplished” when dealing with complexity. Some portion of the stimuli variety may be met with the existing schema as part of routinizing. However, this does not mean that the requisite variety has been achieved. A broken clock is able to tell time correctly twice a day, but it does not mean that you should assume that the clock is functional.
I will finish off with a great insight from Max Boisot:
Note that we do not necessarily require an exact match between the complexity of the environment and the complexity of the system. Afterall, the complexity of the environment might turn out to be either irrelevant to the survival of the system or amenable to important simplifications. Here, the distinction between complexity as subjectively experienced and complexity as objectively given is useful. For it is only where complexity is in fact refractory to cognitive efforts at interpretation and structuring that it will resist simplification and have to be dealt with on its own terms. In short, only where complexity and variety cannot be meaningfully reduced do they have to be absorbed. So an interesting way of reformulating the issue that we shall be dealing with in this article is to ask whether the increase in complexity that confronts firms today has not, in effect, become irreducible or “algorithmically incompressible”? And if it has, what are the implications for the way that firms strategize?
Always keep on learning…
In case you missed it, my last post was Nietzsche’s Overman at the Gemba:
I welcome the reader to read further upon the ideas of Ross Ashby. Some of the references I used are:
This is available as part of a book offering that is free for community members of Cyb3rSynLabs. Please check here (https://www.cyb3rsynlabs.com/c/books/) for Second Order Cybernetics Essays for Silicon Valley. The e-book version is available here (https://www.cyb3rsyn.com/products/soc-book)
