01 / SELECTED INVESTIGATIONSOPEN METHODS, FALSIFIABLE IDEAS
Two studies.
No shared department.
Our initial archive features independently conducted work associated with the founder. It demonstrates the kind of research Meridian intends to support across fields—not the output of an already staffed institute.
PLATE II · COMPUTATIONAL LINGUISTICSOPEN SOURCE · WORKING PAPER · ACTIVE FOLLOW-UP
Project RBT.
A non-neural system for target-conditioned corpus transformation. In its original French–Latin investigation, a 614,000-proposal search reached a plateau; held-out comparisons revealed a metric that primarily detects typological proximity, not linguistic ancestry.
Later experimental runs include documented corrections, ablations, and unresolved tests. Findings and limitations are recorded in the public repository.
ORIGINAL RUN REPORTED · FOLLOW-UP UNDERWAY
Read the paper, code and audit trail ↗
PLATE III · ECONOMIC HISTORYMONOGRAPH · DOCUMENTARY & FIELD RESEARCH
Who killed Cuervo?
A long-form investigation of a New Mexico community through railroads, roads, commercial changes, local accounts, and archival evidence.
The 160-page research monograph is in the process of being prepared for public presentation. Its purpose isn't to vindicate a tidy ghost-town legend; it's to ask what the record can actually support.
RESEARCH COMPLETED · PUBLIC EDITION IN PREPARATION
Be notified when the public edition is ready ↗ These projects began as independent work before Meridian's organizational launch. Neither is represented here as having received Meridian funding, an editorial acceptance decision, or completed external peer review. The RBT artwork is an illustrative schematic, not experimental output.
