The pillars in the temples of southern India weren't carved. They were turned.
Stand close and the evidence is on the surface. Concentric rings running around the shaft, perfectly parallel. A profile that repeats down the whole length without drifting. Flat end plates where the piece was held. And at one end, the small ridge where a cutting tool entered the stone and the matching mark where it left.
Start and stop marks. On granite.
Those are the signatures of a lathe, and a lathe for a monolithic granite pillar is not a small machine. The piece has to be mounted, spun true and held against a cutter harder than itself, at a scale that has no known equivalent in the ancient world.
Some of the same temples hold pillars that ring. Strike the columns at Vittala in Hampi or at Nellaiappar in Tirunelveli and each gives a different pitch, cut from a single block of stone.
One researcher who went to see them in person called them unexplainable and put them before the flood, more than twelve thousand years old.
Some argue it's generations of temple masons with templates, abrasive powder and time nobody today would spend. Others point to the machine the stop marks imply.
So what was turning granite in southern India?
