Copper cannot cut granite. That single fact is the engine behind every lost-technology theory about Egyptian stonework, and it is entirely true. Copper sits at about three on the Mohs scale and the quartz crystals in granite sit at seven. Drag a copper blade across granite and the granite wins. What the theory misses is that the Egyptians were not trying to cut granite with copper. They were cutting it with sand, and using copper only to push the sand around.
The short version
- Granite was extracted by pounding, not cutting, using hammerstones of dolerite, a rock harder and tougher than granite itself.
- Shaping and drilling used loose quartz sand as the abrasive, with copper saws and tube drills acting as carriers for the sand.
- Experimental archaeologists have reproduced the sawing, drilling and dressing marks found on Egyptian granite using nothing but these materials.
- The quarry at Aswan still holds the unfinished evidence: an abandoned obelisk surrounded by pounding trenches, cracked and left in place.
Ask the right question and the mystery evaporates
The framing of the puzzle is usually wrong. People ask how a Bronze Age civilisation cut granite, imagining a blade slicing stone the way a saw slices timber. Hard stone has almost never been worked that way, not in Egypt and not anywhere. It is worked by abrasion: rubbing away material with particles harder than the stone, suspended in water and dragged across the surface under pressure. That is how granite benchtops are finished today, with industrial diamond instead of desert sand. The principle has not changed in five thousand years, only the abrasive.
Once you see the process as abrasion rather than cutting, the toolkit makes sense. A flat copper blade with no teeth, worked back and forth in a trench of wet sand, will sink into granite. The copper wears away faster than the stone does, which is why the archaeological record contains so little of it, and why the blades had to be replaced constantly. The sand does the work.
What the unfinished obelisk shows
The most useful object in this whole debate is a failure. In the granite quarries at Aswan lies an obelisk still attached to the bedrock along one side, abandoned when a crack opened through the shaft. Had it been raised it would have been the largest ever erected, well over a thousand tonnes. It is conventionally associated with the reign of Hatshepsut in the Eighteenth Dynasty.
Because work stopped mid-process, the surfaces were never dressed smooth. What survives is the raw method: long shallow trenches running down both flanks of the shaft, their floors covered in overlapping concave scoop marks roughly the width of a fist. Those are pounding marks. Workers stood in the trenches and hammered downwards with balls and picks of dolerite, a dark igneous rock harder and far tougher than granite, crushing the rock a few millimetres at a time until the trench was deep enough to undercut the block. Discarded dolerite pounders still litter the site.
| Detail | Figure |
|---|---|
| Mohs hardness, copper | About 3 |
| Mohs hardness, quartz in granite | 7 |
| Dolerite pounders | Harder and far tougher than granite |
| Aswan quarries to Giza | Roughly 900 kilometres by river |
| Unfinished Aswan obelisk | Over 1,000 tonnes, never raised |
| Granite roof beams, King's Chamber | Individual blocks of tens of tonnes |
| Abrasive used for sawing and drilling | Loose quartz sand and water |
| Metals available in the Old Kingdom | Copper and arsenical copper; no iron tools |
The experiments that settled it
This is not armchair reasoning. Experimental archaeologists have spent decades actually doing it. The most systematic programme, carried out by Denys Stocks over many years, reconstructed Egyptian copper saws, copper tube drills, bow drills and stone pounders from period-accurate materials and used them on real granite while measuring wear rates, sand and water consumption, and the marks left behind.
The results matched the ancient artefacts. Copper tube drills rotated with a bow or a weighted lever, fed with wet sand, cut annular grooves in granite and produced cylindrical cores identical in profile to the cores Flinders Petrie collected at Giza in the nineteenth century. Petrie, working before anyone had tried the experiment, concluded the Egyptians must have had fixed jewel points set into their drills, because the spiral striations on his cores looked to him like a single hard point tracking through the stone. Replication showed that loose abrasive produces exactly those striations, as individual sand grains momentarily jam and score the surface. Petrie's observation was excellent; his inference was simply premature.
Flat copper blades with sand produced the long straight kerfs visible on granite sarcophagi, including the overcuts and false starts where a sawyer went slightly too far and had to correct. Those mistakes are among the best evidence available. A machine does not leave hesitation marks.
What replication has demonstrated
- Granite can be quarried, sawn, drilled and polished with copper, stone, sand and water.
- The tool marks produced match those on genuine Egyptian granite objects.
- Rates are slow but entirely workable given the labour and time available.
- Petrie's jewelled-point hypothesis is unnecessary.
What no evidence supports
- Powered rotary saws or lathes in Old Kingdom Egypt.
- Lost hardened alloys or chemical stone-softening agents.
- Any tool type absent from tomb depictions, quarry debris and workshop finds.
- The claim that the work is physically impossible with known materials.
Slow is not the same as impossible
The honest sticking point is rate. Abrading granite with sand advances at millimetres per hour, and people reasonably object that the sheer volume of finished granite in Egypt looks incompatible with such a pace. The objection dissolves once you put real organisation against it.
Egyptian state projects ran for decades with rotating workforces numbering in the thousands, fed and housed by an administration that existed largely to do this. Work was seasonal, timed around the agricultural year and the Nile inundation, which also provided the high water needed to float multi-hundred-tonne loads downstream from Aswan. Hundreds of sawyers working in parallel, each advancing a few millimetres an hour, for months, is a very large amount of granite. The bottleneck in the ancient world was never mechanical ingenuity. It was calories, logistics and administrative continuity, and pharaonic Egypt was unusually good at all three.
There is also a selection effect in how we perceive the problem. Egypt produced granite objects for three thousand years, and the survivors are disproportionately the largest and hardest things ever made, because those are what survive. We compare a civilisation's greatest hits against our intuition about a single workshop's daily output, and conclude that something impossible happened.
What the AI angle actually adds
Asking a language model how the Egyptians cut granite produces a summary of the abrasion consensus, because that is what the literature says. It is a reasonable way to get oriented and a poor substitute for the experiments themselves. Models reproduce the distribution of published claims, which means they will also faithfully reproduce a confident-sounding version of any error that is widely repeated online.
Where computation genuinely contributes here, it does so unglamorously: photogrammetry and structured-light scanning of tool marks at sub-millimetre resolution, statistical comparison of striation geometry between ancient and replicated surfaces, wear modelling of copper under abrasive load, and provenance work matching stone to specific quarry faces. That work is answering questions Petrie could only guess at. None of it required a new theory of the Egyptians. It required someone to pick up a copper blade and a bucket of sand.
The verdict
The mystery of Egyptian granite is a mystery only in translation. Substitute abrasion for cutting and the toolkit becomes ordinary, the quarry marks become legible, and the abandoned obelisk at Aswan becomes exactly what it looks like: an enormous, brilliantly organised, extremely patient job that went wrong. The impressive part was never the tools. It was the willingness to spend a generation of human effort on one stone.
About the author
Motech researches and produces long-form video on technology and the unexplained for an audience of more than 270,000 subscribers. Corrections and tips: info@richardelhaj.media