Semantic Attractors and the Recursive Mirror: Corpus Topology as a Method for Semantic Archeology
Zenodo (CERN European Organization for Nuclear Research) July 8, 2026 Peer reviewed DOI: 10.5281/zenodo.21266296 via OpenAlex
Summary
A reproducible pipeline that combines corpus analysis with network topology is introduced and applied to over 224,000 text segments from more than 2,700 documents. The texts cover religious canon, government FOIA records, UAP testimony, psychedelic ethnography, and comparative mythology. The pipeline enables systematic comparison of semantic structures across these diverse domains, revealing how language and narrative patterns emerge and connect across different cultural and institutional contexts.
Study at a glance
| Design | corpus analysis with network topology pipeline |
|---|---|
| Population | text chunks from documents spanning religious canon, government FOIA records, UAP testimony, psychedelic ethnography, and comparative mythology |
| Key finding | A reproducible corpus-topology pipeline was developed and applied to over 224,000 text chunks from more than 2,700 documents across multiple domains, enabling systematic semantic network analysis. |
Abstract
We present a reproducible corpus-topology pipeline and apply it to 224,215 text chunks from 2,700+ documents spanning religious canon, government FOIA records, UAP testimony, psychedelic ethnography, and comparative mythology.