The Wrong 20th Century

How to read this archive

On 30 June 1908 something came down over the Tunguska river. In our century it exploded in the air and flattened two thousand square kilometres of forest. In this archive it does not explode. It lands, intact, and people spend the next ninety years cutting it up.

Every document below was generated by Claude through ComfyUI on Comfy Cloud, and every one was generated from an earlier document. The only image it was shown was the one it cites. When the 1949 newspaper reprints the 1944 war map, that is because the war map was the input. The catalogue numbers are Numonic asset hashes, and the web of citations is drawn from the lineage Numonic extracted from each file, not from captions.

The canon was kept in a ledger as the work went on. Each artifact had to reuse at least one fact an earlier artifact had established and add at least one consequence of its own. The counters beside the scroll are that ledger.

Divergence, as the ledger records it: 1908-06-30: the Tunguska object does not explode; it lands intact in the Siberian taiga (the Fall).

I · The citation web

Which document was made from which

Each line is a lineage edge Numonic extracted from a file: the later document’s ComfyUI graph loaded the earlier document’s exact bytes (matched by SHA-256). Nothing here is drawn from captions. Hover or focus a line to see which facts it carried; select a document to open its provenance.

III · The consequence ledger

Every fact, who established it, who relied on it

Kept as the archive was made. A document could only be accepted if it reused at least one earlier fact and added one of its own.

Fact	Established by	Relied on by

The Wrong 20th Century

How to read this archive

Which document was made from which

The century, one document at a time

Every fact, who established it, who relied on it

Every MCP call made to build this page