The Fable Of Rational Systems

Convention, once it stabilises, has a habit of presenting itself as reasoning.

This is what happens when convention graduates from language into process.

Institutions are where that habit gets a badge.

We still talk as if organisations think. They review, assess, evaluate evidence, make trade-offs, and arrive at reasoned conclusions. This is flattering nonsense.

Most institutional decisions are not acts of judgement. They are acts of self-protection carried out in a serious tone of voice.

The question is rarely what is right. It is what can be defended, documented, circulated, and survive contact with governance, procurement, legal, leadership, audit, and the long afterlife of organisational blame.

We do not decide because it is right. We decide because it is recognisably defensible.

That sounds close enough to rationality that people rarely inspect the difference. But the difference is doing quite a lot of work.

Institutions do not usually fail to think. They fail more elegantly than that. They are often full of intelligent people operating inside systems that reward legibility over judgement, precedent over discernment, and procedural cleanliness over the slight discomfort of actually deciding something. That is why so much organisational behaviour has that peculiar dead-eyed familiarity to it.

A procurement process selects the safe vendor. Not the best one, or even the most suitable one in any deep sense. The safe one. The known name. The proven category. The one with references that calm the room down and that nobody has to explain twice.

If the work is mediocre, nobody looks reckless. If it fails, the paperwork survives. The choice does not need to be especially good. It needs to be difficult to criticise without sounding unreasonable.

That is not judgement. That is institutional fear with a spreadsheet.

Roadmaps work much the same way. Officially, priorities were weighed, evidence was gathered, trade-offs were considered, and a coherent direction emerged. Sometimes that is true. Quite often it is fiction written after the event.

The roadmap is assembled around decisions that had already hardened through budget gravity, executive preference, delivery inertia, inherited commitments, or simple organisational exhaustion. The direction comes first. The rationale arrives later, cleaned up and ready for company. What gets presented as reasoning is often just whatever the system was already prepared to tolerate, rewritten as intent.

Rory Sutherland has a useful line here: it is easier to be fired for being illogical than for being unimaginative. That is one of the more useful observations in Alchemy, his book about the distance between tidy logic and actual human behaviour. Institutions know this instinctively. The unusual idea has to survive scrutiny. The familiar mistake only has to survive procedure.

Metrics help complete the illusion. Organisations love to say they measure what matters. Often they measure what behaves.

The stable metric wins. The dashboard-friendly one. The one that sits still long enough to be reported upward without an awkward scene. Not because it is the most meaningful signal, but because it is the least socially dangerous one.

A metric that changes shape with reality is a nuisance. A metric that keeps its composure looks mature. So teams optimise towards what holds together in review decks, then speak as if stability itself were proof of significance. The proxy hardens into truth. The dashboard becomes less a view of reality than a way of disciplining what reality is allowed to count.

Governance, naturally, makes all this worse while sounding like the cure. Another review, another sign-off, another checkpoint, another group of serious people performing concern at approximately the right altitude.

The stated purpose is better decision making. The actual effect is usually better insulation.

Behaviour does not materially change. The same incentives remain. The same compromises survive. The same timid choices pass through untouched. What improves is the explanation. There is a paper trail now, a challenge session, a template, a record that the issue was surfaced, reviewed, considered from multiple angles, and moved through the proper channels before being allowed to become someone else’s problem.

This is not nothing. It is just not judgement.

It should sound familiar by now. The first article made the point that you are not designing decisions. You are designing what stops being a decision. The second sharpened it: you are designing what people will learn to say without thinking.

At institutional scale, the loop closes.

You are designing what organisations can say without risk.

That is what process really is. Not neutral structure. Not the machinery around decision making. Not some innocent scaffold that good judgement politely climbs. Process is habit that has been formalised and approved.

That is why institutions trust it so much. An individual habit still looks contingent. An institutional habit arrives wrapped in policy language, approval routes, precedent, compliance, and the faint moral prestige of having been reviewed by several people with calendars full of meetings. It no longer feels like repetition. It feels like order.

That is the trick.

Process does not turn repetition into judgement. It turns repetition into something much harder to challenge.

Which brings us to the fable itself. The Fable Of Rational Systems is the belief that enough process produces reason. That if a decision passes through sufficient layers of seriousness, documentation, challenge, and review, then what comes out the other side has somehow earned the status of judgement.

Usually what it has earned is protection.

Institutions are not especially good at deciding. They are often exceptionally good at manufacturing decisions that look as though nobody really made them.

That is the real achievement. The safer vendor wins and it looks prudent. The roadmap solidifies and it looks strategic. The stable metric survives and it looks disciplined. The extra governance layer appears and it looks responsible. Everything acquires the texture of thought, while very little of it requires anyone to think especially hard at the point of consequence.

Behaviour stabilises faster than understanding. Language stabilises faster than intent. Process stabilises faster than judgement.

And once process stabilises, repetition starts to dress itself up as proof.

This is why institutional rationality is so persuasive. It has all the right props. The documents are in order. The criteria are visible. The challenge has been minuted. The review happened. The sign-off exists. The recommendation appears to have emerged from somewhere respectable.

Decision theatre belongs here quite comfortably.

Not because everyone is lying. That would almost be cleaner. It is worse than that. Most people are participating sincerely in a system that has confused procedural completion with intellectual honesty. The performance is not fake in the theatrical sense. It is real enough to be convincing, and hollow enough to leave the actual question untouched.

It is the meeting where everyone performs challenge without introducing danger. The workshop where options are discussed but only one is survivable. The governance forum where risk is surfaced mainly so that responsibility can be diluted before impact arrives.

Everybody is present. Very little is truly at stake for the process itself.

And so the institution gets what it really wants. Not the best answer. Not even the clearest one. The answer that leaves the smallest bruise.

That is why explanation expands while behaviour stays put. A system can become incredibly articulate about what it is doing without becoming any more honest about why it is doing it. It can produce elegant rationales for choices that were never really alive as choices in the first place.

Clarity, in that environment, is not the opposite of dysfunction. It is often its preferred visual style.

This is also why institutional language becomes so uncanny after a while. Confidence rises as ownership falls. The rationale becomes cleaner as conviction becomes thinner. Documents sound decisive in exact proportion to the degree that nobody wants to be the person who actually decided.

The process decided.
The governance decided.
The criteria decided.
The precedent decided.
The framework decided.

Everyone was involved.

Nobody, somehow, was responsible.

That is the part worth sitting with.

Institutions do not usually collapse because they are irrational. They collapse because they are highly rational in ways that protect themselves first. Highly rational about avoiding blame. Highly rational about distributing consequence so widely that it never quite lands. Highly rational about preferring the answer that is easiest to defend over the one that is hardest to ignore.

That is still a kind of intelligence.

Just not one we should keep flattering with the word judgement.

Institutions do not fail because they cannot explain themselves.

They fail because explanation has become the substitute for responsibility.