Knowledge

In twenty years of interim work I have written and received dozens of handovers. They all have the same shape: an overview of files, a list of live cases, the state of the finances, the risks, the names that matter. A good handover is complete, current and readable within a day. And in no handover I ever received was there any account of why things are the way they are. Which alternatives were weighed when the regulation was drafted. Which battle was fought over which clause. What the previous occupants knew about the question that will come back in three years. The handover records what holds. The why does not travel with it.

That is not carelessness on the part of those who write them; it is the shape of the instrument. Decisions record outcomes, not the judgements behind them. Whoever designed a scheme had a definition of the problem in mind that never made it into the decision, and at every changeover the file moves while the definition of the problem stays behind. The rotation is, moreover, institutional: the tenures of the Senior Public Service (the Netherlands’ senior civil-service corps, rotated between posts) are shorter than the lifespan of the schemes that fall under the roles. After two or three changeovers the scheme is still there and the reason is gone. The previous paper in this series called that an orphaned design.¹ This paper is about the first link in the chain that prevents it: knowledge.

Two kinds of knowing

The distinction is older than the problem. Gilbert Ryle separated knowing that from knowing how; Michael Polanyi showed that we know more than we can tell.² Applied to the state: the first kind of knowing is thoroughly institutionalised. The text sits in the statute database, its application in the work instruction, its explanation in the manual, and training the people who apply it is an industry. The second kind of knowing — command of the room for judgement around the rule, knowing which amendment the edifice will survive and which will bring it down — was never institutionalised. It stayed behind in the heads of the people who made the design. Organisations survive rotation; heads leave. That asymmetry is the engine behind every orphaned design.

The objection is ready to hand: the judgements are on paper, surely. The explanatory memorandum, the parliamentary papers, the archive. True — and it does no more than relocate the problem. A statement of reasons accessible only through archival research functions as source material, as raw stock for a reconstruction that takes weeks and requires specialists. Knowledge in the sense of this series is operative: someone can exercise it today on a live question. Whoever wants to know why a parameter stands as it does must be able to consult. Whoever has to reconstruct is thrown back on an excavation, and a system that needs an excavation for its own reasoning knows nothing in the sense meant here. The pattern language Nederland 2040 (The Netherlands 2040) makes exactly this distinction — consulting versus reconstructing — and builds a public facility for it.³

Where the question has an address

One domain saw this problem coming two centuries ago and solved it: water. Rijkswaterstaat (the Directorate-General for Public Works and Water Management) has existed since 1798. Since 1965 there has been a standing technical advisory body for the flood defences, which under changing names runs on to this day.⁴ Around it lies a chain: a knowledge institute, chairs at Delft, the water boards as a daily practice ground, and since 2010 a government commissioner with an annual programme and a fund of his own.

The test came in 2017. The Netherlands then replaced the foundation under its flood safety: the standards moved from probability of exceedance to probability of flooding, with a basic level of protection for every citizen behind a defence.⁵ That is an intervention at the heart of the edifice, precisely the amendment that is impossible in an orphaned design. It succeeded because the question of why the standard stands as it does had an address: a community that commands the room for judgement and replenishes itself through doctoral candidates, chairs and the standing network. Nº 00 described the government commissioner under the General Administrative Law Act (Awb) as a personalisation of memory, effective until the person falls away. The Delta Commissioner is the next stage up: a role with a chain around it, so that the address goes on existing when the occupant changes. This is how the first link ought to be lodged: as a reserve with its own supply.

Where no one sees the whole

Set beside that the distribution model of the Municipalities Fund, the arrangement through which tens of billions are shared out among municipalities each year. Since the Financial Relations Act of 1997 the model has been adjusted layer over layer: dozens of indicators, each added or recalibrated to solve an acute distribution problem.⁶ Extension, never renovation — the amendment pattern from Nº 00. When central government wanted to recalibrate the model, the task turned out to be archaeological. The fund managers themselves gave as their reason that the model had become too complex and could no longer be explained; the Council for Public Administration judged critically the traceability of what took its place.⁷ Command of the model sat with a handful of specialists at the ministry and with a few research consultancies. Municipalities that came off worse could not point to where in the model that disadvantage arose, because no one saw the whole, and the recalibration that took effect in 2023 remains contested to this day — not least because no party has the authority to answer the why conclusively.

The contrast with water is total. There the foundation could be replaced; here it had to be excavated. Both systems are equally Dutch, equally technical, equally politically sensitive. The difference lies in one thing: whether the knowledge lives somewhere.

The countable substitute

Nº 00 established that every link has acquired a substitute that imitates its shape and is countable. For knowledge that substitute is the document. The knowledge base, the intranet, the lessons learned after every incident, the evaluation after every programme: all countable, in reports delivered, sessions held and databases filled, and all reserve without supply. A document is the residue of knowing. Command does not grow inside it; it grows only in people who exercise the knowledge.

There is a second substitute, sharper and closer to home: the outside consultancy as memory. When the in-house occupancy rotates and the archive falls silent, the organisation hires the knowing in. Advisory firms that have known the organisation longer than the sitting board. Research consultancies that command the distribution model better than the fund manager. Interim managers brought in because they still know the institution from a previous assignment. I am not writing this from the outside; my trade owes its existence to this gap. The interim manager is paid, in part, to know what the organisation itself once knew. Anyone who earns from that has a duty to name the gap and to leave it smaller on departure than it was on arrival. That is why my own assignments carry a return clause: coming back two years after departure to test what still stands. It makes the safeguarding question part of the assignment rather than an intention.

Why does the substitution catch no one’s eye? Because the knowledge base is tickable and the consultancy hireable, and both are visible in the annual accounts. Command is invisible, until the day it is absent.

Consultable rather than reconstructable

The action perspective at this link consists of two movements and one test.

The first movement: for every scheme, record the reasoning per parameter as part of the decision itself, publicly and by version. The question of why this figure and not another is then consultable in twenty years’ time rather than reconstructable. This calls for no change to the system; it calls for the half-page of considerations, alternatives and the chosen line to be written before the decision falls, and published with it.³ Any organisation can start tomorrow, with its own by-laws and subsidy schemes.

The second movement: lodge command in a place with its own supply. A chair, a standing chain of advice, a discipline that replenishes itself. The water domain shows what that looks like; the annual risk-equalisation cycle under the Health Insurance Act (Zvw), described in Nº 00, shows that it can also be done with a young law. An archive preserves. Only a community that exercises the knowledge keeps it operative. How that exercise can be organised is the question of the practice regime, further on in this series.

And the test, to be run for your own organisation in a single movement: point to the scheme your organisation runs on, and answer the question of who can explain, within a week, why the most important parameter stands as it does, without archival research. If the answer is a role, this link is lodged. If the answer is a name, work out when that person retires. If there is no answer, you have an orphaned design under management.

With water the Netherlands knows why each standard stands as it does; it has invested two hundred years in that and renews the investment every year. With the distribution of its money it does not. The difference is not a matter of aptitude. It is a choice about where knowledge lives, and that choice can be made afresh for every file. An archive remembers. A discipline knows.

Footnotes

1. Jacob Huibers, Orphaned Design (Nº 00), Statecraft, 2026, statecraft.nl. ↩

2. G. Ryle, The Concept of Mind, 1949; M. Polanyi, The Tacit Dimension, 1966. ↩

3. Nederland 2040, een patroontaal (The Netherlands 2040, a pattern language), pattern 29 (Information with certainty), on consulting versus reconstructing, and pattern 169 (The reasoning before the decision). ↩ ↩²

4. Technische Adviescommissie voor de Waterkeringen (Technical Advisory Committee on Flood Defences), established 1965; continued from 2005 as the Expertise Netwerk Waterkeren, and since 2014 as the Expertise Netwerk Waterveiligheid (flood-safety expertise networks). Delta Commissioner and annual Delta Programme since 2010; Delta Act in force 2012, Delta Fund 2013. ↩

5. New standards for primary flood defences from 1 January 2017 (Water Act): the shift from probability of exceedance to probability of flooding, with a basic level of protection of at most 1 in 100,000 per year for every citizen behind a primary defence, to be achieved by 2050 at the latest. ↩

6. Financial Relations Act 1997; recalibrated distribution of the Municipalities Fund from 1 January 2023. ↩

7. Council for Public Administration, opinions on the recalibration of the Municipalities Fund, 2021 and 2022, including on the traceability of the distribution model. ↩