Puma Punku sits at 3,840 metres altitude on the Bolivian altiplano, adjacent to the ancient city of Tiwanaku, and it should not exist. Not because the ruins are not real — they are very real, and extensively documented — but because the technology required to produce what lies scattered across its surface has no plausible conventional explanation within the accepted timeline of Andean civilisation.
The site consists primarily of massive stone blocks, many of them andesite (one of the hardest stones on Earth, with a Mohs hardness of 7) and red sandstone, cut with a precision that staggers modern engineers. The H-shaped blocks that have made Puma Punku famous feature interior angles machined to tolerances of less than a millimetre. Drill holes are perfectly cylindrical. Channels are routed with the consistent depth of a machine cut. The blocks interlock in three dimensions, fitting together without mortar, in a jigsaw pattern that requires the cutting of each stone to be precisely planned in relation to every adjacent stone.
What makes this particularly difficult to explain is the dating. The Tiwanaku culture — to whom Puma Punku is conventionally attributed — reached its peak around 600–800 CE and had no writing system, no iron tools, and no wheel. The site sits near no forest, meaning the timber required for conventional construction techniques would have had to be transported considerable distances. And the andesite blocks weigh, in some cases, over 130 tonnes — requiring, on conventional assumptions, approximately 1,700 people working in coordination to move a single stone across the 90 kilometres from the nearest quarry.
The mainstream archaeological explanation invokes human ingenuity, ramps, ropes, and community organisation on a scale comparable to the Egyptian pyramid builders. This explanation is possible. But it requires a pre-Columbian Andean civilisation with organisational and technical capabilities that left no other trace in the record — no comparable structures elsewhere, no documentation, no material culture of equivalent sophistication. Alternative researchers, including Brien Foerster and David Hatcher Childress, have argued that the evidence better supports the hypothesis of an earlier, more technologically advanced civilisation whose achievements were absorbed or memorialised by later peoples.
The question of who built Puma Punku, and with what technology, remains genuinely open. The conventional answer is not impossible — but it is deeply unsatisfying, and the standard of proof applied to it is nowhere near the standard that the evidence demands.