01Executive summary
Six milestones, and the single most consequential idea in this entire set: that there are general laws, and that they can be known.
Before the sixth century BC, technical knowledge was a collection of rules that worked. Thales and his contemporaries began asking about lines, angles and solids in the abstract rather than about this field and that stone. Euclid organised the result deductively around 300 BC. Eupalinus drove a tunnel through a mountain from both ends. Hippodamus laid out cities as coherent wholes. Archimedes established the mechanics of levers, pulleys and floating bodies. Ctesibius and Hero built force pumps, water clocks and a steam reaction turbine that did no useful work at all.
02From the particular to the general
Mesopotamian and Egyptian surveyors could compute the area of an irregular plot and the volume of a cylindrical stone. What they did not do was think about triangles or cylinders. Every calculation was about a specific object, and the knowledge was a set of procedures for cases that had come up before.
The Greek move was to treat the shape as the object of study. Once a triangle is a general thing rather than a particular field, statements can be proved about all triangles at once, and those statements apply to cases nobody has encountered. That is the entire basis of engineering analysis, and it happened once.
Greek natural science had remarkably little effect on Greek engineering. The geometry was used — principally in architectural proportion — but the physics was not. Engineering contributed considerably more to science than science did to engineering, and that remained true until the second half of the nineteenth century. Roughly twenty-five centuries separate the birth of Greek science from the point at which scientific knowledge became a routine input to engineering practice. Any account that presents ancient science as the foundation of ancient engineering has the relationship backwards.
03Eupalinus at Samos, and how the Greeks worked
In the sixth century BC, Eupalinus of Megara drove an aqueduct tunnel roughly a kilometre long and about 1.8 metres square through a mountain on Samos, working from both sides toward the middle. It is the same problem as Siloam two centuries earlier, at greater length, and it is the clearest single demonstration that Greek surveying was a real technical discipline rather than an approximation.
The organisation of the work
Greek public works ran on a contract system that is recognisable today. A city-state hired master craftsmen who brought their own helpers and apprentices, working under an architecton who was responsible to the state. The contract was often inscribed on stone and placed on the site, so that the workmen and the public could both read it. The specification for the naval arsenal at Piraeus — a building roughly 120 metres long — ran to about four printed pages in modern type, fixing wall thicknesses, stone sizes and window dimensions but leaving everything a modern drawing would show to be carried in the heads of the master workmen.
The specification was public
Inscribing the contract on stone at the site is an accountability mechanism, not a decoration. Everyone could see what had been agreed and compare it with what was being built.
Proportion replaced dimensions
Working to a modulus — column diameter two or three units, height ten or twelve — let masons with marked rods translate a design into stone without detailed drawings. It is dimensional coordination, and it substitutes for documentation.
04Archimedes and the mechanics of machines
Archimedes of Syracuse, roughly 287 to 212 BC, is the one practising engineer among the great Hellenistic mathematicians, and the distinction shows in what he worked on: mensuration, specific gravity, the pressure of liquids, and the action of levers. The compound pulley and the screw are both attributed to him, though each may well have been in use before he made it famous.
The military engineering at Syracuse in 212 BC — levers, cranes and catapults that frustrated a Roman siege — is the part usually told as a story. The part that matters is that the machines were designed from stated principles rather than copied from precedent.
05Ctesibius, Hero, and the machine that did nothing
Ctesibius of Alexandria and his successor Hero describe a hydraulic clock, a hydraulic organ, a fire engine, a force pump, an air gun, a steam turbine and an automatic theatre whose figures moved by steam or hot air. The devices are correct in principle and were, as far as anyone can tell, essentially toys.
The aeolipile deserves precise description because it is so often misdescribed. It is a hollow ball free to rotate on a horizontal axis above a cauldron, with two tubes bent in opposite directions projecting from opposite sides. Steam enters through the pivots and escapes from the bent tubes, and the ball spins from the reaction of the unbalanced escaping jets. It is therefore a reaction turbine, not a device driven by steam pushing on a movable surface, and it did no useful work.
It is tempting to treat the aeolipile as a missed industrial revolution. It was not. A reaction turbine of that form has negligible efficiency, there was no material capable of holding useful pressure, no fuel economy to make it worth running, and — most decisively — an economy resting on slave labour in which mechanical power solved no problem anyone was trying to solve. The principle was available for eighteen centuries and stayed a curiosity because none of the surrounding conditions existed. Working principles do not become technologies until something wants them, and reading history backwards from a device to a supposedly inevitable outcome is the most common error in popular engineering history.
Pressure piping at Pergamon
The Hellenistic period did produce one hydraulic work of genuinely surprising ambition. Around 200 BC, water was brought some 55 kilometres to Pergamon in three tile pipes laid side by side, discharging into a reservoir on a hill above the city. From there a single main ran down across several valleys — the deepest more than 180 metres below the reservoir — and up over the intervening ridges as an inverted siphon. At the low point the pressure approached 20 bar, several times that in a modern town main. Tile and lead would both fail at that pressure, and the material of the main is genuinely uncertain; on tensile grounds wood is the most plausible candidate.
06Takeaways for current practice
- Generalise from the case to the class. It is the single move that turns accumulated procedure into analysis.
- Machines trade force against distance; they do not create work. Everything about efficiency follows from taking that seriously.
- A working principle is not a technology. The aeolipile lacked materials, fuel economics and any problem worth solving.
- Publish the specification where the work happens. The Greeks inscribed contracts on site, which is an accountability mechanism worth more than it costs.
- Dimensional coordination substitutes for documentation. Working to a modulus let complex buildings be built without detailed drawings.