Featured Product
This Week in Quality Digest Live
Lean Features
Maria Guadalupe
A new study suggests that investing in soft skills will result in higher individual—and national—productivity
Jake Mazulewicz
There isn’t one magic bullet, but rather a ‘consolidation of subtleties’
William A. Levinson
We must train ourselves to recognize the waste around us
Jeetu Patel
Two studies shed light on how work cultures need to evolve
Mark Rosenthal
Nailing down the key concepts

More Features

Lean News
Weighing supply and customer satisfaction
Specifically designed for defense and aerospace CNC machining and manufacturing
From excess inventory and nonvalue work to $2 million in cost savings
Tactics aim to improve job quality and retain a high-performing workforce
Sept. 28–29, 2022, at the MassMutual Center in Springfield, MA
Enables system-level modeling with 2D and 3D visualization, reducing engineering effort, risk, and cost
It is a smart way to eliminate waste and maximize value
Simplified process focuses on the fundamentals every new ERP user needs
DigiLEAN software helps companies digitize their lean journey

More News

Harish Jose


A System Is Designed to Do Exactly What It Does, Or...

A constructivist’s view of POSIWID

Published: Wednesday, May 4, 2022 - 11:03

he dictum, “purpose of a system is what it does” (POSWID) is famous in cybernetics, attributed to the management cybernetician Stafford Beer.

Beer notes: “A good observer will impute the purpose of the system from its actions and thus from the resultant state.”

Hence the key aphorism and acronym. There is, after all, no point in claiming that the purpose of a system is to do what it consistently fails to do. I’ve written about this before here and here. In cybernetics, the emphasis is on what a “system” does, and not especially on what it is or what the designer or management of the “system” claims it’s doing. Thus, we can see that POSIWID has a special place in every cybernetician’s mind.

A “system” is a collection of variables that observers purposefully select to make sense of the world around them. The boundaries and parts of the system vary according to who is doing the observing. The purpose also is assigned by the observer.

Beer explains this clearly:
“The point that I find that I am most anxious to add is that this System has a PURPOSE. The trouble is: WHO SAYS SO?
So where does the idea that Systems in general have a purpose come from? IT COMES FROM YOU! It is you, the observer of the System, who recognizes its purpose. Come to think of it, then, is it not just YOU—the observer—who recognizes that there is a System in the first place?”

Another key point is that an observer may impute several purposes for the system.

Beer continues:
“Consider the System called a tiger. The purpose of a tiger is to:
• Be itself
• Be its own part of the Jungle System
• Be a link in animal evolution
• Eat whatever it eats, for Ecology’s sake
• Provide tiger skins
• Perpetuate the genes of which it is the host

“For the moment, I am prepared to say that the purpose of a tiger is to demonstrate that the recognition of a System and of its purpose is a highly subjective affair.”

Understanding the purpose of a system helps us to understand how we construct the systems themselves. All of this turns out to mean that we simply can’t attribute purposes, or even boundaries, to systems as if these were objective facts of nature. The facts about the system are in the eye of the beholder.

This sounds like an unproductive conclusion, but we can make something of it. It means that both the nature and the purpose of a system are recognized by an observer within his perception of what the system does.

From Beer’s writing, it’s clear that the POSIWID depends on the observer. This is also the basis of constructivism, in which the observer is the king or queen. The system is a selection of variables chosen by the observer to improve their understanding of a phenomenon. The boundaries drawn by the observer are entirely arbitrary and contingent on the oberver’s mood. A “system” is thus a mental construct of the observer.

For example, an educational system may have physical artifacts in the world such as buildings, books, chalkboards, etc. However, depending on the observer, what an educational system entails will change. For a student, it’s a system for learning, or a system to get away from their hometown. For a teacher, it’s a system to provide meaning to their lives or a system to earn money while doing another job on the side. There can be as many systems involving the same collection of parts as there are observers of that system.

Beer continues:
“The definition of the purpose of a System as being what it does lays the onus not on ‘nature’ but on the particular observer concerned. It immediately accounts for UNRESOLVABLE disagreements about systems, too. For two people may well disagree about anything at all and never become reconciled. They say that they will be convinced, and give way, if the FACTS show that they were mistaken. But the facts about the nature and purpose of a System are not objective realities. Once you have declared, as an observer, what the facts are, the nature and purpose of the System observed are entailed.”

As a constructivist, this is an important concept to grasp. If there are two observers, and each is constructing the system, they each will come up with their own systems and varying POSIWIDs. Our first step in systems thinking, then, is to understand how the other participants view the system by their assigned purposes, and how they see the POSIWIDs. Even if they assign a purpose for the “system,” the outcome that they perceive may not match what they expect.

Here are some important points from from this information:
1. There are always multiple participants in the social realm. It’s important to understand what a system means for each stakeholder. This includes the parts, the whole, the assigned purposes, and the POSIWIDs. There’s no POSIWID without an observer.
2. There’s always a gap between what we believe the purpose of a system should be, and what it actually is doing. We’re tempted to assign an objective reality to the “it” here, but we should resist this and understand that the “it” or system is an “as-if” model or abstraction that we employ to make sense of our environment.
3. To understand the gap between the expected system and the actual one, we need good comparators in place to allow us to measure what the gap is. POSIWIDs depend entirely on the type of the observer to distinguish what’s happening. A good fictional example is Sherlock Holmes and Inspector Lestrade. Holmes, the master observer, is able to distinguish more system attributes than Inspector Lestrade, which would in turn correlate to more POSIWIDs.
4. Similarly, what we think the system is doing could be faulty. This means that we need an ongoing error-correction process to improve our ability to manage the system. We must interact with the system as much as possible, and also welcome other participants’ input and perspectives. We can’t manage a system unless we’re a part of it. We should practice epistemic humility.
5. POSIWID(s) should be reinterpreted as often as possible, with input from others. They help us understand the dynamics of the various parts and how they interact with each other.
6. We should focus on only a few POSIWIDs at a time. Because we can't manage all the external variety thrown at us, we should filter out the unwanted POSIWIDs.
7. We can’t predict what the POSIWID(s) will be beforehand. Due to the complex connections between the parts, as well as their nonlinear relations, POSIWIDs are more likely to be unpredictable. This is another reason we should resist the temptation to treat systems as objective realities in the world.

One of the main struggles I had when I started my journey into constructivism was how we could manage a system if it’s entirely “subjective.” I’ve put the term in quotes because there’s no subject/object distinction in constructivism. I’ll write more about this later.

Beer explained this issue well:
“Question: How is it that systems are subjective, while some of them can be singled out and declared to be viable?
Answer: Once you have defined them, you can tell whether they are viable or not.
Question: And those criteria are suddenly supposed to be objective?
Answer: Well, it’s all about necessity and sufficiency within a stated frame of reference.

“If systems are subjective phenomena," Beer continues, "then we are going to have trouble in determining a measure. The whole idea of measures is to be objective.... Yet the problem we face is not unique. In fact, the measures that we are accustomed to call objective work only because we accept a set of conventions about how they are to be employed. For example, if we quote the height of Mount Everest, we do not mean that this is the distance you would travel from the base camp to climb it; nor do we mean that if we look at Mount Everest while holding a ruler at arm’s length, we can read off its height. We might have agreed on either of these conventions: They would both work, given certain other stateable conditions. It seems that objective measures, like objective systems, exist only as conventional crystallizations of one out of a virtually infinite number of subjective possibilities.”

Stay safe and always keep on learning....

First published on Harish's Notebook.


About The Author

Harish Jose’s picture

Harish Jose

Harish Jose has more than seven years experience in the medical device field. He is a graduate of the University of Missouri-Rolla, where he obtained a master’s degree in manufacturing engineering and published two articles. Harish is an ASQ member with multiple ASQ certifications, including Quality Engineer, Six Sigma Black Belt, and Reliability Engineer. He is a subject-matter expert in lean, data science, database programming, and industrial experiments, and publishes frequently on his blog Harish’s Notebook.


Nice to see this!

Back in the '90s, there was a discussion group (a listserv) called the Deming Electronic Network. Illuminaries in the world of Quailty held regular discussions there on various aspects of the System of Profound Knowledge, Deming's 14 points and 7 deadly diseases, and other related subjects. People like Myron Tribus, Heero Hacqueboord, Michael Tviete, David Kerridge and many others were regular discussants. Don Wheeler chimed in occasionally. Some of the topics really plumbed the depths of epistemology (especially pragmaticist epistemology, a la C. I. Lewis and C. S. Peirce), systems thinking, statistical thinking (especially analytic studies). Some excellent explanations of some otherwise very arcane topics (as well as some lively debates) emerged in those discussions, and I've always been disappointed that after the listserv shut down, no one compiled it or indexed it or made it otherwise available. For a while, you could access an archive at the Deming Institute, but I don't believe that's possible any more. 

Anyway, this is a long way to go to say I really enjoyed this article, because it brings some of those philosophical concepts back into the light. I especially thought the discussion of the height of Mt Everest was remeniscent...it brought all the discussions around the vital importance of operational definitions and Deming's quote "There is no true value of anything." 

Reference of quotes should be provided

Hi Harish,

Thank you for sharing your thoughts about systems and their purpose as well as Beer's view of this matter.

However, I think it could be prudent to share references to the sources of the Beer's quotes that you used in your post.


Alexander Solodkin


Good suggestion. The source is The Heart of Enterprise. I will add that to the source post.