Why Euclid Still Matters.

A reasonable objection. Geometry is two and a half thousand years old. The world has produced calculus, set theory, computers, physics. Why begin a project about contemporary cognition with a book whose latest revision predates almost everything that matters?

The objection is real, and worth answering carefully. The answer is not that the Elements is the only good place to begin. It is that, considered as a piece of cognitive infrastructure — as a designed environment for training a particular set of mental operations — the Elements still has properties no modern alternative quite reproduces.

Five of those properties are relevant.

§ 01Closure: every step is licensed

In the Elements, every move in every proof is licensed by something explicit: a postulate, a definition, a common notion, or a previously established proposition. There are no informal steps. There are no places where the argument relies on a thing the reader is expected to bring from outside. The system is, in this sense, closed.

Closure is rare. Most domains in which one might want to train explicit reasoning involve at least some appeal to background knowledge that is not being tracked. The result is that learners in those domains acquire reasoning habits which depend, silently, on the unstated. They become unable to separate the inference they are licensed to make from the inference they are smuggling.

Closure makes that smuggling visible. If a step in a Euclidean proof is unjustified, the reader can locate the missing license. The exercise of reading and rebuilding such proofs is the exercise of doing this location. It is unusually direct training for a habit one rarely otherwise gets to practise.

§ 02Economy: a small toolset

Five postulates. Twenty-three definitions. Five common notions. The whole of Book I is built on these. The economy is itself a pedagogical fact. A reader who has internalised the toolset can hold all of it in mind simultaneously. Every move in a proof can be checked against a list of around thirty items, all known to the reader, none of them deferred to a textbook somewhere else.

This is the opposite of the situation in nearly every modern technical training, in which the toolset is enormous, partially tacit, and impossible to hold in mind. Economy permits a kind of cognition that abundance forecloses. You cannot trace dependencies in a tool you cannot keep in your head.

Closure makes smuggled assumptions visible. Economy makes them rare. The two together produce an unusually transparent environment for the practice of inference.

§ 03Visibility: the figure is honest

Euclidean propositions come with a diagram. The diagram is not decoration. It is the simultaneous public artefact against which the proof is checked. A diagram that does not satisfy the construction cannot be drawn; a proof that does not match the construction cannot be completed.

Modern interactive media give us something Euclid did not have: a diagram that updates as the construction is being made. Drag a point, and the figure adjusts. The result is an honesty machine. Claims that hold only for the special case the original draughtsman happened to draw become visible the instant the figure is perturbed. Claims that are genuinely invariant remain invariant under every drag.

This is one of the few places where modern tooling actually improves on the original medium of the Elements. An interactive Euclidean proof is more honest than a static one. Not less.

§ 04Necessity: the mode of conviction

The Elements trains a particular epistemic posture: the distinction between appears true and must be true. A diagram may suggest something is the case. The proposition establishes that it is the case, given the postulates, by demonstrating that no consistent assignment of the construction's parts can avoid the conclusion.

This distinction — between contingent appearance and logical necessity — is one of the most important moves a thinking person learns to make. It is also under-trained. Most contemporary intellectual environments operate at the level of plausibility: arguments that seem reasonable, conclusions that feel right, sources that look credible. Plausibility is not necessity. The person who has practised the move from one to the other in geometry is in a measurably better position to recognise its absence elsewhere.

§ 05Transfer: the operations are general

The objection that geometry is no longer relevant frequently rests on a confusion between the content of geometry — triangles, circles, parallels — and the operations a geometric proof exercises. The content is, indeed, mostly irrelevant to most people's lives. The operations are not.

The operations the Elements trains are these: identify what is given; identify what must be shown; produce a sequence of moves licensed by explicit primitives; track the dependency of each move on what came before; test the conclusion against the construction; recognise when the argument has and has not closed.

Each of these operations transfers. Out of geometry, into law. Out of geometry, into program design. Out of geometry, into the reading of any argument in any field. They are general cognitive moves, not local geometric ones, and the rare property of the Elements is that it offers an environment unusually well-tuned for practising them in isolation, without the contaminating noise of a domain in which other things are also going on.

§ 06What this argument is not

It is not a claim that the Elements is the only or even the best place to learn rigorous reasoning. There are many such places. It is not a claim that a person who has worked through Book I will automatically reason well in domains far from geometry; transfer is real but partial, and the rest of the project exists partly to make the transfer more reliable.

Nor is it nostalgia. The Elements is not honoured here because it is old. It is used because, of the available training environments, it has the rare combination of properties — closure, economy, visibility, necessity, transfer — that make it unusually well-suited to the cognitive work this project is organised around.

Geometry is not the destination. It is the dojo. One trains in the dojo because the surface is flat, the rules are clear, and the consequences of every move are visible. One leaves the dojo in order to do something else, with habits that the dojo made possible.

One trains in the dojo because the surface is flat, the rules are clear, and the consequences of every move are visible.

Two and a half millennia later, the surface is still unusually flat, and the rules are still unusually clear. That is enough.