Epoché
Essential Innovation for Those Drowning in Information but Starving for Wisdom
A Free book by Michiel Buisman
Published chapter by chapter.
updated 3-4-2026
Coffee that tastes like work. A presentation going into the last few slides. A small audience, diverse in their energy at the start of this meeting, but growing more homogenous as the peak has passed. The joy of checklists checked, stakeholders informed/involved, quarters reviewed, code committed, steering performed and workshops attended, has passed. It is time for learnings. The speaker attempts to inject suspense, but everybody already knows what went wrong and is eager to move on. On, to the next project under the exact same governance, but with a fresh topic and new set of hopeful stakeholders. We just need to get through this rehearsed language, careful attributions and start nodding in conclusion. Rest assured: the lessons learned will be carefully captured and filed away.
This project in particular had everything. Budget, preconditions met, correct internal and external stakeholders with time and mandate, executive sponsorship, consultant reporting and KPIs. This project was the culmination of years of refining the playbook and appeasing the whole RACI matrix. Everything was done correctly, but the result will never see broad implementation.
This was a digital transformation initiative. Or a knowledge platform. Or an innovation lab. Or an agile rollout. It happened years ago and still happens today. Different decades, different domains, same shape. Every one of these failed. Every one of them had an acceptable explanation ready. And not one of those explanations was true.
The explanation is one of: adoption failure, change management or “people problems”. They are familiar, comfortable and lead to a shallow conclusion. They allocate it to human behaviour which feels tractable and satisfying but ephemeral at the same time. The verdict is familiar: people matter in every project, and next time we will manage them better. We will preserve the methodology, the governance structure, and the career capital of everyone in the room who championed the approach, that had been championed so many times before. With excellent learnings as a result.
These explanations are not untrue because they are dishonest. They are wrong because they are structurally necessary. An organization cannot easily conclude that its governance model is broken. The system is not covering up the truth out of malice. It is generating the most survivable interpretation of the evidence. A feature, not a bug. This is what makes it so hard to fix.
To process the failure, a workshop is organised. It is a long workshop, to stress the importance and underline that we are doing everything we can to capture the learnings. The setting is psychologically safe in the right way. The paper is the correct color of brown for this session. The notes are just sticky enough. Unblaming observations and insights are captured and dot-voted for priority. After much work transforming all this paper into a .doc format, it gets reviewed, annotated, refined or redacted and stored on a network drive. It gets distilled down to two birds eye view slides for the next quarterly board meeting, along with 60 other topics. And that was the end of it. Execution was not the problem. But you cannot say that in the post-mortem meeting. Not without a better explanation. Let’s go inside.
Caution: once seen, it cannot be unseen. Like the emperor’s new clothes, innovation theatre is missing something essential to justify the gravitas and effort put into it. It is like mime: a striped-shirted, gloved artist, miming a box, or a rope, or a room. They go through all the motions in great detail and plausibility. Sometimes even in collaboration with a real or equally imaginary partner, with which they sometimes switch places. It is mesmerizing to watch. Nothing remains, but a memory. Another quality performance of the innovation theatre troupe.
A design thinking workshop is scored for attendance. It truly teaches new insights, attitudes and even some skills. Then the attendees are left to their own devices and return to the order of the day, within a day.
An innovation lab with interior design that is substantially different from the office, where play is encouraged by a prominently placed fussball table. Its main output is slides for the quarterly review, if they are lucky.
A hackathon takes place. Sometimes even with an actual sleepover, but usually not. Always with pizza. And a winner! We built in 36 hours what it would normally take a year. But that was never the problem to begin with.
The stage gates are designed to stringently reduce uncertainty. To eliminate it entirely. So no genuinely novel initiative passes gate two without some lubrication, defeating the system.
We are not failing to innovate. We are innovating at failure. And we have become very, very good at it.
By now you will have guessed: it was never execution of the carefully developed governance that failed. What if the governance, the gates, the KPIs were missing something real and essential, like the pantomime? What if the governance itself were the pantomime and real results required adding a significant aspect; we can and should continue doing hackathons, stage gates, innovation labs and design thinking workshops. But just add this one insight.
The governance model is broken, because it assumes that “knowledge” (in quotes, because you will see new refinements in chapter 3 and 6) existed before the initiative, that it only needed to be shaped and presented, captured and applied. But what if the “knowledge” did not exist before? What model of governance supports the creation of “knowledge” through the act of innovation?
Every knowledge management practice, every innovation framework, every governance model in the modern organization is built on a single unspoken premise: that “knowledge” already exists somewhere, and the job is to find it, capture it, refine it, and distribute it. Like some precious gems.
These assumptions are plainly visible in the practice of questionnaires. Give me your “knowledge”. I specifically and explicitly ask you for it. You are responsible for formulating it in a way that is suitable to my needs. Which will probably involve storing it in a database, which will then be promoted to “knowledge base”. Or a benchmarking exercise. Or a best practice library. These do capture some genuine, explicit “knowledge”. But not as much as hoped. And after campaigning strongly for increased use (after all, learnings are a valuable way to consolidate current governance), access drops sharply. By then the campaign has run its course and nobody tracks it in the quarterly anymore.
And in a stable, controlled environment, this is the correct way to do things. A hospital infection protocol. Aircraft maintenance checklist. Financial audit procedure. Masthead lookout protocol. These have been refined for decades or even hundreds of years. Validated, codified, dissemination optimized and applied. That is not the problem. What happens outside these domains is. The assumption is not wrong. It is situated. And every system that embeds it, however sophisticated, inherits both its power and its blind spot.
Before any practice can be called “best”, it needs some time to ripen. Get the kinks out. Make it future proof. Then codify it. That is exactly what ISO does. And for innovation and knowledge management, 11 standards exist today, that undergo periodic adjustment of even complete rewrites. They provide auditability, scalability, repeatability, liability management. Where these work extraordinarily well: manufacturing tolerances, financial reporting, pharmaceutical quality control. ISO is one of the most successful governance inventions of the twentieth century. To ignore its austerity and wisdom is shooting oneself in both feet. Twice.
ISO addresses innovation and knowledge management systems directly in a range of standards (56000-56010 and 30401). Full certification for innovation management is possible since 2024. 30401 is in revision at time of writing, from the 2018 version. These are standards that serve inside their domain. They are purpose-built exactly for this territory. They are genuinely good, structured, comprehensive and serious. ISO is one of the most successful governance systems ever created. Precisely because of that, it cannot solve this problem: Like all best practices, they optimize for the known so aggressively that the uncertain becomes illegible. And even when, through Herculean effort, it is finally seen, it is rendered ineligible.
Any implementation of innovation worth its salt, starts with defining it for the organisation. I would not dare to do this for you, it represents the unique value it brings and the strategic anchoring that resonates with your organisation. But somehow, somewhere, there probably will be something in there that resembles this: unknown, undefined problems can be solved with unknown, new approaches. Innovation explores both the problems and the solutions and makes the unknown known and actionable. This has everything to do with wrangling uncertainty and nothing with management systems that function under stable conditions. Though it is possible and necessary to apply strong governance to the uncertainty-wrangling. And aspects of it can come from ISO best practices.
At one point, this was the subtitle of the book. It should be on a nice Delft Blauw tile somewhere in plain sight. It was the core of this book, before I went beyond it, on to the “how” and “why”. If only one thing is added to the ISO range for innovation, it is that “knowledge” should be appreciated as a primary valuestream. It should be valorized. And while that is a single word, it is quite alien to many organizations. It is associated with academic circles, where this is done all day, every day. Only in academics, it works somewhat differently. I am referring to enterprise valorization. But before we can start to discuss the process of creating, enhancing, or realizing value from knowledge, waste, or assets, we need to discuss an elephant in the room: what you refer to as “knowledge”, isn’t.
I assume you feel “Knowledge” is in protocols, documents, spreadsheets. It represents proven rules to follow, repeatability and can be tested for compliance. It is one-dimensional and can ideally be captured in a checklist with boolean YES/NO checks. It is about Execution.
But before this level of certainty, things pass a judgement phase. In this phase, people determine what makes one thing better than the other and put that into rubrics and frameworks. Just before finalizing the final judgementframework, is when canonical knowledge peaks.
Before judgement is possible, hypotheses need to be formed. And in order to that, one needs to be able to sit with multi dimensional uncertainty, without frameworks, experiencing and observing, without judging yet. A state of wonder. That is the state of epoché. A domain where tacit knowledge and sensing of first principles develops.
These three modes of knowing have 2.350 year old ancient Greek names, that we will delve into some more in chapter 6. For now just listing them is enough. No more “knowledge” in quotes from now on.
Techne: the knowledge of execution, codified and repeatable. Episteme: the knowledge of causes, built through deliberation and judgment. Phronesis: earliest stage sensing, situational awarenes and cultivates practical wisdom. Aristotle named all three around 350 BC, which is either reassuring (smart people have been thinking about this for a long time) or alarming, because we are still getting it wrong. Chapter 6 will give these their full philosophical depth. For now, the names are enough.
Around 1910, Scientific Management or Taylorism emerged. It was an adequate solution to a contemporary problem. It removed mandate, thinking and judgement from the workers, to a higherup, who did the thinking and judging for them. Then that higherup themselves, along with a bunch of their peers has their thinking done for them by a set of even-higherups, and so on. This creates a desire for predictability and control. It demands certainty quarter by quarter. Career structures became rule-following. Result: everything gets treated as an execution problem.
If you apply rules that solve one problem very well to a different problem, you get unstatisfactory results. Applying a hammer to a screw comes to mind. Hitting it harder, training to hit it just right, it will not make the screw connect more tightly.
Similarly, when phronesis and episteme are managed with the KPI’s of certainty, they underperform and therefore need to change or go. And so it becomes what it needs to be, in order to survive. But that no longer performs the function it once had.
There are many wrong shapes this can take: phronesis for known problems leads to endless exploration and analysis paralysis. Episteme without execution leads to insights that never scale. Many innovation failures are domain and category errors, not bad ideas.
ISO optimizes for execution techne brilliantly. And techne is dominant in modern organizations and the basis of governance and indeed careers.
For innovation in high uncertainty, we require the other two modes too. They are mandatory phases in the creation of new techne. We need governance for all three, not just the one.
We keep demanding business cases before exploration. We keep documenting lessons learned that cannot be put to use. We keep enforcing processes that don’t fit anymore. The first mode, wonder, is the most fragile and the most essential: it is the nursery of new techne.
The three modes are not a new invention. They surface repeatedly across
frameworks developed independently, in different disciplines, over centuries.
Five of the most recognizable:
| Framework | Wonder / Phronesis seed | Judgment / Episteme | Execution / Techne |
|---|---|---|---|
| Aristotelian Virtues | Phronesis: practical wisdom as the disposition that enables genuine inquiry; epoché as its stance; nous (intuitive grasp of first principles) emerging through sustained wonder | Episteme: true, canonical knowledge of causes through deliberation; hypothesis formation and testing; rubric-building for what counts as “good”; peak intellectual virtue before crystallization | Techne: codified skill and craft; rule-bound production; what remains when episteme and phronesis are stripped out in distribution; becomes brittle and potentially harmful without the wisdom that generated it. Phronesis returns as a calm recognition of obsolescence symptoms by experienced practitioners. |
| Cynefin | Complex/Chaotic: probe-sense-respond; emergent patterns, no imposed order | Complicated: sense-analyse-respond; expert judgment, good practices | Clear: sense-categorise-respond; best practice checklists, standard procedures |
| Kolb’s Learning Cycle | Concrete Experience: immersive, sensory, first-hand encounter | Reflective Observation + Abstract Conceptualisation: patterns, theories, causal models | Active Experimentation: applying concepts until they become routine |
| Design Thinking | Empathise + Ideate: immersive observation, divergent thinking, no predetermined solutions | Define + Prototype + Test: hypothesis formation, iterative testing, refinement | Implement + Scale: validated solutions deployed as standardised practice |
| Diagnostic indicators | Reality is genuinely surprising you. No coherent hypothesis yet. Questions proliferating faster than answers. Senior practitioners say “I don’t know, let’s look.” | Hypotheses failing in informative ways. Criteria for “good” are explicit and debated. Boundary conditions becoming clear. | Processes reproducible by non-experts. Outcomes predictable. Compliance audits pass. Unless: edge cases proliferating, practitioners gaming the system, outcomes varying despite compliance. |
The full table is in Appendix A. it maps all three modes across seventeen frameworks including
Heidegger, Foucault, Bourdieu, Polanyi and Nonaka.
It is not required reading. It is there for the reader who wants to see how
deep the structure goes.
I recommend using the Aristotelian names in discussions that should involve knowledge creation. It prevents confusion and permits other people to continue using “knowledge” as they always have.
Many organizations that say they want innovation are structurally preventing the phase that makes innovation possible. This is not hypocrisy. It is a logical outcome of applying techne governance to phronesis.
The pattern is consistent enough to feel like a law. An organization commissions an innovation effort. Resources are allocated, a team is formed, a kickoff is held. Within weeks, sometimes days, the first pressure arrives: what is your hypothesis? What are your success metrics? When will we see results? In the Burcht and the Stad (see interlude 3: Burcht, Stad and Platteland), these are reasonable questions. In the Platteland they are a death sentence. The phronetic phase requires the temporary absence of these questions. The questions assume that knowledge to answer them already exists. It does not. That is the entire point.
The language barrier is real and structural. When someone doing phronetic work says “I don’t know yet,” a techne-trained manager hears incompetence. When they say “I need more time to observe,” the system hears avoidance. When they say “I can’t give you a business case yet,” the governance process hears an unacceptable answer and produces a rejection. Neither party is wrong in their own frame. The frames are incommensurable. One is asking for what the other cannot yet produce.
This creates a translation problem that no amount of goodwill resolves. The phenomenological practitioner cannot defend their work in techne language without betraying it. The moment they produce a premature hypothesis to satisfy the governance requirement, they have stopped doing the work. They have started performing it. And the organization, satisfied that the box has been checked, has no way of knowing the difference.
| Techne hears | Phronesis means |
|---|---|
| “Play” | Essential exploration |
| “I don’t know yet” | Intellectual honesty |
| “Uncertainty” | The honest truth state |
| “Intuition” | Pre-reflective pattern recognition |
| “I can’t give you a business case” | The work hasn’t produced one yet |
Power makes it structural. Phronetic work requires the absence of coercive pressure to conclude. The moment someone with budget authority asks “when will we see results,” the calculus changes. Premature (epistemic) closure becomes the rational career move. The explorer who keeps saying “I don’t know yet” is not the person who gets promoted. That honour goes to the explorer who produces a confident-sounding hypothesis, runs a workshop, generates a slide deck and moves on to implementation. The system selects for performance, not for genuine inquiry.
This is why good managers rationally kill the phenomenological phase. They are not anti-innovation. They are responding to the incentive structures the previous chapters described. The quarterly cycle punishes open-endedness. Annual budgets demand predictable milestones. Procurement requires defined deliverables. Legal requires documented purpose. Audit requires traceable decisions. None of these requirements are wrong in themselves. Together they constitute an environment in which genuine not-knowing cannot survive.
The result is the innovation theater described in chapter 1. Organizations want the outputs of phronetic work: breakthrough insights, reframed problems, options that did not exist before. They cannot tolerate the process that produces them. So they fund the performance of that process instead. Design thinking workshops. Innovation sprints. Hackathons. These are adaptive responses to an impossible requirement: produce the outputs of genuine exploration without the conditions that make genuine exploration possible.
There is one more mechanism worth naming before Part II. Even when protected space for phenomenological work is created, the phase tends to collapse from within. Practitioners trained in techne environments become uncomfortable with their own not-knowing. The absence of a hypothesis feels like failure. The absence of a framework feels like incompetence. The internal pressure to conclude can be as strong as the external one. Epoché is a discipline, not a condition. It requires active maintenance of a stance that the entire organizational environment is working to dissolve.
Part II is about legitimizing this phase without romanticizing it. It is disciplined work that produces specific outputs, though not the outputs governance systems know how to receive. Before those outputs can be described, the phase itself needs to be recognized as real work, with its own logic, its own demands, and its own integrity.
(Pasteur, Einstein, Bohr and Darwin walk into a bar…)
Defense-in-depth is not a modern invention. The logic of concentric rings: outer layers that absorb contact, inner layers that protect what cannot be lost. It runs through ISO 27001, NIST frameworks, and eight centuries of European fortification design. Each ring has its own rules and tolerance for exposure. That graduated difference is what makes the system effective.
In the European landscape this pattern is almost architectural memory. At the center stands the Burcht: thick stone walls, narrow windows, everything audited, locked down, and minimized. Around it lies the Stad: walled but with gates, markets, and town-hall deliberations under controlled oversight. Beyond the walls stretches the Platteland: open countryside, trade routes, villages, outsiders, ideas carried on the wind. Serendipity lives here. So does exposure.
Map this onto the three modes of knowing.
The Burcht is pure techne — execution, repeatability, doelbinding, data minimalisatie, audit trails that satisfy any regulator. In regulated environments this inner keep is genuinely non-negotiable. GDPR and ISO 27001 were built for exactly this zone and they excel inside it.
The Stad aligns with episteme — judgment, hypothesis testing, strategic deliberation. Gates open for trusted partners. Debate is possible, but boundaries are negotiated.
The Platteland belongs to phronesis and the stance of epoché. This is where tacit sensing, first-principles observation, weak-signal detection, and conversations with outsiders happen. It is deliberately more porous because genuine novelty rarely originates inside the keep. It drifts in from the outside.
The structural failure is that most organizations have extended Burcht logic all the way to the outer perimeter. What should be graduated rings has collapsed into one uniform standard optimized for execution. Principles that work brilliantly for known assets (purpose limitation, data minimization, zero-trust) are applied indiscriminately to the earliest, most fragile phase of knowledge creation.
Phronetic work first becomes illegible inside the dominant frame. It produces no tidy audit trail, no compliant risk score, no business case the system can digest. And when, through sheer persistence, a raw insight is pushed through the gates anyway, the system evaluates it against techne criteria and declares it ineligible.
This is not ignorance or malice. It is survivable organizational logic. The Burcht must be defended; boards, regulators, and insurers demand it. Yet the unintended consequence is epistemic amputation: we systematically starve the Platteland where new techne is born.
Aristotle would recognize this as a category error. Phronesis is the practical wisdom to calibrate security to the actual mode of knowing. Epoché requires openness to the unknown. You cannot suspend judgment while simultaneously demanding full documentation and purpose limitation before any sensing has begun. The requirement itself closes the door.
Security best practices themselves warn against this over-extension. Defense-in-depth works precisely because the outer rings are not governed by the same rules as the inner Burcht. Over-fortifying the Platteland does not make the organization more secure. It makes it more brittle.
Three diagnostic questions are worth sitting with:
Protecting the Burcht remains necessary. Treating the entire landscape as if it were the Burcht is how organizations end up performing innovation theater while the real breakthroughs happen elsewhere — usually with competitors who still dare to leave the gates open from time to time.
The remedy is not to tear down the walls. It is to restore graduated security rings that match the three modes: tight and auditable for techne, deliberative for episteme, and deliberately more porous — though never naive — for phronesis. Only then can epoché do its quiet, essential work without being suffocated at birth.