I’m super excited to announce the working title of an upcoming book, by Harish Jose: For Us… Cybernetics as a Human Practice.

This is very much his life’s work on Cybernetics - over 400 pages of profound insights. A masterpiece that I can’t wait to read myself. The book is scheduled to be published later this year and will be available for free to all the members of the Cyb3rSyn Community.

Ebook and print editions will also be available for purchase - I’ll send out discount codes to all the subscribers of the Cyb3rSyn Newsletter when the book is ready to order.

Without further ado, here is the sneak preview of one of the chapters, focussing on a topic that always triggers passionate conversations within the systems thinking community.

I have seen The Gods Must Be Crazy several times, and the Coke bottle still stops me.

The movie came out in 1980, directed by Jamie Uys. The setup is simple. A pilot flying over the Kalahari Desert throws an empty Coke bottle out of his cockpit window. For the pilot, the bottle had one purpose: to hold a soft drink. That purpose is finished. The bottle is trash.

It falls into a different world.

Xi and his San tribe have never seen glass. The bottle lands without breaking. They take it for a gift from the gods and begin, with considerable ingenuity, to find out what it is for. They use it to cure animal hides, to carry water, to grind roots, to roll dough, to make music, to trace decorative circles in the sand. Each person who picks it up finds a different use. Each use is the purpose, for that person, at that moment.

Then it becomes something else again. Only one bottle exists. Everyone wants it. The peaceful tribe fractures along the lines of its demand. The purpose is now conflict. Xi eventually concludes the bottle is an evil thing and makes a long journey to throw it off the edge of the world.

From outside the movie, watching, the bottle has a purpose that none of the characters could name. It is a plot device. It sets the story in motion. That purpose belongs to the filmmaker and the viewer. From a further remove still, it provides work: for the actors, the crew, the technicians who made the movie possible. It exists at a level entirely removed from any use Xi’s tribe found for it.

Same object. Pilot, tribe member, viewer: three different positions, and purposes that do not overlap. Not one of them is wrong.

Stafford Beer, the management cybernetician whose Viable System Model occupies the next chapter, had a name for what the bottle is demonstrating:

The purpose of a system is what it does. POSIWID. Beer arrived at this formulation after years of working with organizations that said one thing and did another. It is a diagnostic commitment more than a definition. When you want to understand what an arrangement is actually for, watch what it consistently produces. Watch who benefits and who does not. Watch what gets prioritized when resources are scarce and the declared purpose conflicts with something else. The Coke bottle’s purpose, in that sense, is never in the bottle. It is always in the encounter between the bottle and whoever is holding it.

Dan Lockton, a design researcher who applied Beer’s formulation to the study of designed environments, clarified what the commitment means in practice:

Intentionality is irrelevant. That is a strong claim. It does not mean that intentions have no moral weight. They plainly do. It means that for the purpose of diagnosis, declared intention is not evidence. Behavior is evidence.

The Coca-Cola Company did not design the bottle to generate conflict in a San tribe or to set a movie plot in motion. Its designers had a specific and limited intention. But those designers are not present in the Kalahari. The bottle is. And what the bottle does, in the Kalahari, is the bottle’s purpose there. The gap between what the designer intended and what the object consistently produces is not a failure of communication. It is a structural feature of designed things encountering lives they were not designed for.

Organizations work the same way. A company may declare that its purpose is to serve customers, to develop its people, to act sustainably. These declarations shape some behavior. They do not determine all behavior. Between the declaration and the actual pattern of decisions, resource allocations, and outcomes, there is always a gap. That gap is where diagnosis begins.

Ackoff’s rain dance, which closed Chapter 4, names this exactly. The ceremony is coordinated, the steps are practiced, the procedure runs on schedule. The weather does not change. POSIWID is that different kind of looking: not questioning anyone’s sincerity, but asking what the arrangement consistently produces.

The black box gives this a precise engineering form. The concept originated in electrical engineering: a sealed unit whose inner workings are hidden, but whose inputs can be varied and whose outputs can be observed. By varying the input and watching the output, an engineer can discern what the box does without ever seeing inside it. Beer made the connection to POSIWID explicit: “It is not necessary to enter the black box to understand the nature of the function it performs.” You do not need to know how the arrangement works internally to understand what it consistently produces. The output is the evidence and therefore the output is the purpose.

The Coke bottle is a black box to every observer who picks it up. The pilot does not know its chemical composition. Xi’s tribe does not know what a soft drink is. The filmmaker may not know how or where the glass was manufactured. None of them need to. Each imputes purpose from what the bottle does in their hands. POSIWID and the black box are the same diagnostic commitment from different angles.

Beer did not stop at “watch what it does”. He pressed one level deeper, and the result is more interesting than the slogan.

A purpose is imputed by an observer. The word matters. To impute is to attribute, to read in from outside, not to discover as an intrinsic property. Purpose does not live in the object or the organization. It lives in the encounter between the object and whoever is looking. Beer was explicit about this:

Revisiting the Coke bottle: Pilot sees trash. Tribesman sees grinder, water-carrier, pattern-tracer. China’s Xi sees ideological poison. Filmmaker sees plot device. Each is doing POSIWID correctly. Each arrives at a different answer because each is observing from a different position, with a different repertoire of categories, at a different moment.

Beer made this even more specific with an example of his own:

A single observer finding multiple simultaneous valid purposes for the same entity. None of them wrong.

In organizational life, the implication is direct. When two people observe the same company and arrive at different imputed purposes, they have not made an error. They have observed from different positions. The auditor looking at compliance rates imputes a risk-management orientation. The frontline worker observing what actually gets escalated imputes something different. The external shareholder observing capital allocation imputes something different still. The disagreement between them is not noise to be resolved. It is information. Where their imputed purposes converge, something durable is visible. Where they diverge, something worth investigating is there.

Beer drew the conclusion plainly:

What the practitioner needs, then, is not the correct imputation. There is no view from nowhere that provides one. What the practitioner needs is multiple imputed purposes held together, compared, and treated as a richer description of what the arrangement is actually doing than any single perspective can provide.

The observer’s own variety matters here, and the connection to Chapter 3 is direct. The variety an observer can perceive determines the POSIWIDs available to them. Arthur Conan Doyle’s detective Sherlock Holmes and his Scotland Yard counterpart Inspector Lestrade stand at the same crime scene. Lestrade sees a struggle. Holmes sees a left-handed person of between thirty and forty years who recently returned from a warm climate. Both are doing POSIWID. Only one of them has the perceptual variety to impute a purpose that is diagnostic in practice. Building that variety, through experience, through deliberate contact with different perspectives, through sustained attention to what the arrangement actually produces, is part of the practitioner’s ongoing work.

Hegel’s dialectical method is notoriously dense. He reportedly admitted only one person had ever understood him, and even that person had not quite managed it. One insight from his approach is useful here regardless.

The insight is simple, even if the surrounding philosophy is not. Every apparently stable idea contains an inner contradiction. That contradiction does not invalidate the idea. It drives its development forward. Hegel illustrated this with a plant:

The bud is not wrong. The blossom does not correct an error. Each stage is partial, and each carries the previous ones within it. What looks like contradiction is development.

The Coke bottle moves through at least three such moments.

The first moment is the pilot’s frame. The bottle has a designed purpose, declared by its manufacturer, enacted in the specific transaction of buying and consuming a soft drink. That purpose is clear, bounded, and finished the moment the drink is gone. This is the working assumption of most management science: organizations have a designed purpose, declared by founders or governing bodies, and the task of everyone in the arrangement is to achieve it. The assumption is not wrong. Declared purposes do shape behavior. The first moment is partial, not false.

The second moment arrives when the bottle lands in the Kalahari. The tribe has no access to the designer’s intention. What they have is the object and their own ingenuity. They find the bottle purposeful in ways the pilot never imagined, and their found purposes are not errors. They are the encounter between a designed object and purposeful people who were not part of the design. Organizations face this constantly. The declared purpose does not flow cleanly down the hierarchy and get implemented. It encounters the purposes of the people doing the work, who adapt it, reinterpret it, resist it, or perform compliance while doing something else in practice. The designed purpose is the beginning of a negotiation, not the end of one.

The third moment is POSIWID. Once both the first and second moments are visible, the question becomes: what does the arrangement actually produce? Not what was it designed for, and not what do the people inside it want, but what does it consistently do, as observed across time and across observers? Beer’s insight is that this third moment is not an endpoint. It is a comparator. The arrangement should maintain a function that continuously measures the declared purpose against the purpose imputed from what it actually delivers. That feedback modifies the original purpose. What converges over time is neither exactly what the higher level wanted nor exactly what the operational units preferred, but what the ongoing encounter between declaration and behavior has produced.

The pilot’s bottle, thrown away as trash, becomes the tribe’s most contested object. That is its third-moment purpose. No one designed that. It emerged.

A screwdriver makes the same point more plainly. Its designed purpose is to drive screws. But the purpose of any given screwdriver, at any given moment, is what its user needs it to do: drive a screw, pry open a paint can lid, score a line in timber. The purpose is dynamic. It is determined by the user at the time of use. The instruction card inside the packaging is not evidence about purpose. The garage is evidence.

The bottle is a purposive object. It was designed with a specific use in mind. It does not choose its purpose. Whatever it is used for, the purpose comes from outside it.

Xi is purposeful. He can decide the bottle is evil. He can undertake a long journey to dispose of it. He can choose, based on his own assessment of what the situation requires. The bottle has no such capacity. This distinction, between things that have purposes assigned to them and beings who generate their own, is foundational to understanding why POSIWID matters beyond diagnosis.

Aristotle distinguished between things that move toward their purpose from within and things whose purpose is assigned from without. The seed unfolds into the tree because that unfolding is the seed’s own direction. The stone shaped by a sculptor has its purpose given to it entirely from outside. The stone does not care about being a statue.

Werner Ulrich, a Swiss systems thinker working in the tradition of his teacher C. West Churchman, built this distinction into the foundation of his framework for examining the assumptions hidden in any design or intervention. His formulation of purposiveness is precise:

And he drew the implication directly: all design of tools represents somebody’s solution to somebody’s problems. Both words carry weight. The designer had specific concerns. The users have different ones. The tool embeds the designer’s purposes in its structure, and when you use it, you enact that design whether or not it fits your situation.

Purposefulness is different. Ulrich defined it as:

A purposeful being can step back from a procedure and ask whether the procedure should exist. They can ask whose problems it solves and whose it ignores. A purposive component cannot ask this. It does not know what it is for.

Kant grounded this distinction in a categorical imperative:

Kant was not saying that people cannot be useful to one another. He was saying that treating a person as nothing but a means, managing them as a purposive component rather than engaging them as a purposeful agent, is a moral violation regardless of whether it produces efficient outcomes.

Geert Hofstede, the organizational researcher, gave this an operational edge: as soon as people are part of the process, the effects of interventions are not known. Purposeful beings generate responses that purposive design cannot anticipate, because purposeful beings have their own assessments of whether the procedure makes sense. When those assessments are suppressed rather than engaged, they do not disappear. They manifest as workarounds, quiet noncompliance, the gap between the policy and the rate.

Ulrich pressed to the diagnostic conclusion. Any framework that does not include the intrinsic purposefulness of the people in it falls back on a machine model of social arrangements. Tools have purposes and are purposive. People are purposeful. The purposiveness of tools depends entirely on the purposefulness of the people using them. Purpose only makes sense when you are talking about a being that can reflect on it.

The POSIWID implication is direct. When an arrangement treats the people in it as purposive components and then imputes purpose from what it consistently produces, it will arrive at imputed purposes that nobody chose and for which no one is responsible. What the arrangement produces is the emergent result of purposeful people adapting to being treated as purposive ones. That emergent result is the POSIWID. The gap between it and the declared purpose is, among other things, a measure of how thoroughly the machine model has been applied.

Geoffrey Vickers, the British systems thinker, spent much of his career arguing that goal-seeking was an inadequate description of how human arrangements actually work. What organizations do, in his account, is not pursue goals but maintain relationships: relationships between people, between operational units, and between the organization and its environment. These relationships are continuously generated from within, by the ongoing judgment and perception of the people in them. The course being followed is not given from outside, as it would be for a vessel steered toward a fixed destination. It is continuously produced from inside.

POSIWID is a useful diagnostic for things that are visible. It does its harder work on things that are not.

The Coke bottle generates conflict in Xi’s tribe because only one exists and everyone wants it. But what sustains the tribe’s capacity to manage that conflict, to negotiate, to maintain relationships under pressure, is not the bottle. It is the accumulated social fabric of the tribe: the shared understandings, the informal protocols, the habits of relation that have developed over years of living together. That fabric is invisible precisely because it is working. It becomes visible only when it breaks down.

Susan Leigh Star, an American sociologist, named this phenomenon with precision. Her concept of infrastructure extended well beyond roads and servers. Infrastructure, for Star, is the sociotechnical fabric that sustains any organized activity: the standardized forms, the unwritten protocols, the informal networks, the skills that exist only in the hands and habits of people who have been doing the work for years. She wrote:

Her key insight was simple and unsettling: infrastructure is invisible when it works. It becomes visible only when it fails. We will review the concept of ready-to-hand in detail later in Part 3.

Star identified several attributes that describe how infrastructure operates. It is embedded: thoroughly integrated into surrounding arrangements, not easily separable from the activities it supports. It is transparent to expert users: the experienced nurse does not think about where the relevant information lives; she simply knows. And it becomes visible upon breakdown: when the infrastructure fails, or when the people who embodied it leave, its absence makes apparent what its presence had been silently providing.

Star illustrated this with nursing work in hospitals. Much of what nurses do is coordinative and interpretive: managing the gap between standardized procedures and the specific person in the bed, translating between the clinical record and the actual condition, catching what the monitoring equipment does not register. This work is not captured in formal job descriptions. It is not measured by efficiency metrics. It is simply, as Star’s respondents noted, thrown in with the price of the room.

The POSIWID of nursing work is what nursing work actually does. And what it actually does (the coordination, the gap-bridging, the catch) is not what the job description says it does. The job description describes the visible, documentable, billable portions. POSIWID requires observing the whole pattern of what gets produced, including the portion that no metric tracks.

When organizations pursue efficiency without attending to this invisible layer, the consequences follow a predictable pattern. The staff who managed the informal networks, who held the tacit knowledge, who filled the gaps that policy never saw: when they leave, the gaps appear. The infrastructure becomes visible, in Star’s formulation, upon its breakdown.

From a variety standpoint, the mechanism is exact. The environment generates variety that formal procedures were never designed to absorb. The gap was always being filled by informal adaptation: the judgment call, the workaround, the piece of knowledge that only one person held. When that person leaves, the formal procedures face variety they were not built to handle. The breakdown is not a new problem. It is the old solution becoming visible in its absence.

Star and her colleague Geoffrey Bowker posed the question that follows:

The purpose of the informal infrastructure is what it does. What it does is absorb variety that the formal arrangement cannot absorb. No policy document states this, because policy documents are not designed to acknowledge the gap between what the policy covers and what the situation requires. The gap is the infrastructure’s domain, and the infrastructure is, by definition, invisible.

The gap between declared purpose and imputed purpose is not a failure to be corrected. It is the primary diagnostic signal.

A declared purpose that matches the imputed purpose across multiple observers, under varying conditions, over time, is either describing something very simple or describing something very well. Either way, it is information.

A gap between the two is also information, and usually more generative. It tells the practitioner where the declared description has lost contact with what the arrangement is actually producing. It locates the operative constraint not where the policy placed it but where behavior has placed it. It reveals where a purposive framework has been applied to purposeful people and what they have produced in response. It shows where invisible infrastructure is absorbing variety that the formal description cannot see.

Beer proposed that the organization should maintain a comparator: a function that continuously measures the declared purpose against the purpose imputed from what the arrangement delivers, and feeds the gap back into the ongoing adjustment of purpose itself. The purpose is never fixed. It converges toward a working position between what the policy formulated, what the operational activity produces, and what the environment is currently demanding. That convergence is what viable organization looks like from the inside.

The Coke bottle’s purpose never stabilized. Each time someone new picked it up, the purpose shifted. Xi’s eventual solution was to remove the bottle from the arrangement entirely, because no stable convergence was possible while only one bottle existed and many people needed it. He could not redesign the bottle. He could not create more of them. He could only change the conditions under which the bottle was operating. He threw it off the edge of the world.

Most organizations cannot throw the bottle off the edge of the world. They have to keep working with the gap between what they say they are for and what they consistently produce. The practitioner working with POSIWID has four questions to move through. What does the arrangement actually do, as observed from outside its declared intentions? Who is observing, and what variety do they bring? Are the people in the arrangement being engaged as purposeful beings or managed as purposive components? And what invisible labor is sustaining the visible function that no metric is currently tracking?

These questions do not produce a final answer. They produce a more honest account of what is actually happening.

If purpose is what an arrangement does, and what it does depends on who is watching, and the people inside the arrangement are purposeful beings continuously generating their own purposes, what architecture can hold all of this and remain viable? That question is what the Viable System Model was built to answer.

I will finish with Ralph Stacey:

That’s it for this week. If you are in the northern hemisphere, Happy Spring Break! I’ll write to you again week after next.