Files, fields, identifiers, and metadata

Dataset Documentation

Everything needed to interpret, archive, validate, cite, and reuse the release responsibly within its stated terms.

Release v1.0-AI-ECR Published 2026-07-31 Matthew 5–7 AI-reviewed · not human peer review

File guide

FilePurpose
central-propositions.csv179-row finalized central analytical register
historical-occurrences.csv213 preserved historical OCC-MT records
occurrence-central-links.csv220 explicit links between historical and central layers
textual-issues.csv68 textual or semantic-control records
mapping-adjudication.csv45 provisionally confirmed mapping controls
pericope-results.csvThirty literary-unit summary table
command-family-profile.csvCommand-family frequency and membership profile
topic-profile.csvPrimary and inclusive topic profile
soteriological-relation-profile.csvMulti-label soteriological relation profile
data-dictionary.csvPublic field definitions

Stable identifiers

CEN-MT

Stable identifiers for the 179 final central propositions.

OCC-MT

Stable identifiers for the preserved 213 historical occurrence records.

TV-MT

Stable identifiers for textual variants and semantic-control records.

MT-SM

Stable identifiers for the thirty Matthew 5–7 literary units.

CF-MT

Stable identifiers for the 56 command families represented.

T identifiers

Controlled topic identifiers used for primary and secondary topic assignment.

Central register fields

The central register includes release identity, proposition wording, source occurrences, mapping cardinality, evidence type, instructional form, polarity, scope, temporal force, normativity, command families, topics, soteriological relations, canonical status, textual links, mapping status, confidence, final AI-ECR decision, theological-framework status, and review classification.

Download the complete field-by-field data dictionary.

Machine-readable metadata

Repository release

The deployable package includes a repository-ready release directory containing a README, release page, methods, limitations, citation metadata, data dictionary, checksums, release manifest, and source-to-public filename map. It can be uploaded to a Git repository without changing the website paths.

Versioning rule: preserve this release unchanged as v1.0-AI-ECR. Corrections should be published as a new version with a change log rather than modifying archived files silently.

Download bundles