Measuring what humanity knows, and what we're losing
The Deep Time Research Institute conducts cross-domain research on cultural knowledge systems: measuring the accuracy of oral traditions, mapping knowledge extinction in endangered languages, and building open-access tools for preservation triage.
DTRI has a peer-reviewed publication in archaeoastronomy and a second accepted for publication in ethnopharmacology, manuscripts in review, and several submissions held pending appropriate cultural review. See the papers page for the current list.
Our central question is whether the accuracy of a cultural tradition is governed by how directly its claims can be checked against the world by the people transmitting them — the External Referent Constraint. That framework is under active development. Its cross-domain measurement was retired in August 2026; the published work testing it within single domains stands.
All preprints, data, and code are open access except where a deposit is restricted during review.
Every statistical claim is paired with the code and data that produced it. Numbers are computed from raw data via code rather than carried forward from earlier drafts. Papers are paired with a public Zenodo deposit, and manuscripts run through an adversarial multi-model review before submission.
That apparatus is load-bearing, not decorative. In August 2026 an internal source-verification audit of our own flagship dataset found that a substantial share of its values could not be traced back to their cited sources. We retired the analysis rather than repairing it, withdrew the figures that depended on it, corrected an accepted paper before typesetting, and are publishing the audit itself as a methods contribution.
The standard is that a number should not appear unless a script in the deposit produced it. That standard is what we work to, not a property we claim to have achieved.
AI is core research infrastructure here. Claude Code is the primary computation engine for statistical analysis, Monte Carlo simulation, and data pipelines, under direct author supervision and disclosed in every manuscript's methods section. Methodological choices, including AI use, are documented on the methods page.
Whether the work holds up is for peer review to decide. That's the whole point of submitting.
Elliot Allan — Founder & Director. Quantitative archaeology, cultural evolution, and empirical analysis of oral traditions.
Contact: elliot@deeptime-research.org