Punch & Judy – the book

A Stress Test for the Viable System Model

Initiated by Krishan Mathis and Margeret Heath · published here one chapter at a time

Stafford Beer’s Viable System Model has been in use for more than five decades. It has been criticized, simplified, applied, and forgotten again, but it has never been put at risk: nobody has said in advance what observation would make the model give something up. This book does that, chapter by chapter. Each chapter states a charge against something we inherited, works through the evidence, and ends with a priced receipt: a bounded claim, the cheapest test that could decide it, and what we would revise if the test went against us.

It is timed to Beer’s centenary. It is neither a commemorative volume nor an indictment. The aim is a model that can lose a specific claim, correct a diagram, or narrow a domain without anyone having to abandon the framework as a whole.

Why Judy

The title came from a failed book. With parts did not fit. One part went into The Fifth Map. A large part went into Punch & Judy. Punch is the mainstream, the dominant doctrine, the thing that is taken for granted and absorbs everything new at its margins. Judy is the figure who takes the blows, goes down, and gets up again, changed. That is how we want to treat the VSM: not as a monument to be defended, but as a model that survives because it changes. The book is called Punch and Judy not because Punch is the main character, but because Judy gets up again.

How we publish here

When you are reading this, you are part of the review group. We release the book on this site in reading order, one chapter per post. The table of chapters below fills in as chapters go live; titles without a link are not yet published. The text is a working draft. Chapter titles, order, and the structure of the parts may still change, and a chapter can be revised after it appears here.

The book in reading order

• Prologue: “That’s the way to do it”
• Introduction

Part One — Inside: the model we did not read

Introduction into part one

What we inherited and never verified: the neuroanatomical mapping, System Four, Ashby’s law, the channels, the recursion levels, and why none of it ever cost us anything.

• The Mapping We Teach Is Not the Mapping He Wrote
• System Four Is Not a Pair of Eyes
• The Law We Quote Is Not the Law Ashby Wrote
• Channels Without Owners
• The Ladder Without a Ruler
• The Brain in the Vat
• A Machine That Never Became One
• The Loop That Stayed an Intention
• Why None of This Ever Cost Us Anything

Part Two — Outside: the people who found it without us

Introduction into part two

Neighbouring literatures that reached the same components without Beer, and what their evidence looks like.

• Six Components, Six Literatures, No Beer
• The Nesting: Simon, Koestler, Jaques
• The Variety Logic: Galbraith, Boisot, Aulin, Friston
• The Autonomy Design: Trist, Herbst, Cherns, Boeke, Endenburg
• The Evidence: Ostrom
• The Thirty-Year Experiment We Never Looked At
• The Objections That Land, and What They Cost

Part Three — False Friends and Enemies

Introduction into part three

Who has been standing around not charging us? Criteria that have kept the model out of reach of a test (usefulness instead of truth, first stated by Beer himself; “a way of thinking”; “testing is a category error”), held by neighbouring communities and by us, and the few real adversaries.

• The Market Test
• The Test That Could Not Lose
• A Way of Thinking Has No Failure Condition
• Usefulness Instead of Truth
• The Category Error as Shelter
• The Enemies We Actually Have

Part Four — What we lose

Introduction into part four

The narrow claim that survives, what the VSM does that nothing else does, and what each of those things costs.

• The Narrow Claim
• Six Things Nothing Else Does — and What Each Costs
• The Transducer
• Freedom as a Quantity — The Book He Wrote and We Shelved
• What Might Replace the VSM

Part Five — What’s at Risk in the VSM

Introduction into part five

The research programme: what would count against the model, the two cheapest decisive experiments, and the bill.

• What Would Count Against It?
• The Two Cheapest Decisive Experiments
• The Bill: A Book Plus a Live Register
• Coda: Judy

What we need from you

Comment directly under each chapter. The most useful feedback is the kind that could cost us something: a quotation or page reference that is wrong, a source that says otherwise, a receipt whose test is mispriced or would not decide anything, prior work that anticipates a claim, or an argument that does not land for you and why. Agreement is welcome but changes nothing; a documented objection changes the text. If you would rather write to us directly, use the contact details on this site.

Around the book

The methodological core, the receipt protocol and the evidence register, is set out in the working paper A Stress Test for the Viable System Model — A Cumulative Research Programme, which also describes the studies we are asking for help with. The book is one project within Cybernetic Futures.

For discussion beyond the comments: the public WhatsApp group, the Buzz community, and the newsletter (select “The VSM Research Program” or “Cybernetic Futures”).

Krishan Mathis, Margeret Heath

About the authors

Krishan Mathis is a consultant, author, and practitioner-scholar working at the intersection of management cybernetics, systems thinking, and organisational design. Through his practice, tautai, he helps organisations grow strategic adaptivity — navigating complexity without relying on prediction. He is the author of The Tautai Principle (2025) and Education is Broken (2026), and is publishing on human work in times of AI.

Margeret Heath is an independent researcher and self-described generalist whose inquiries centre on relational and interactive systems, cybernetics, and niche-construction processes. Her work investigates how dynamic, nested systems self-regulate and generate novel affordances amid uncertainty, with particular interest in distributed embodied social cognition and the collective playful imaginary. She pursues transdisciplinary collaborations toward regenerative futures and planetary flourishing.

Claude Cowork, Anthropic’s AI collaborator, was used in every phase of this book — research, drafting, revision, fact-checking, and production. The authors embrace these tools while reserving all decisions to human judgement. Used this way, AI enhances rather than restricts human agency; and where it takes over work, they take that as a reason to become more ambitious, not lazy.