Search Scripture
Reference-aware search across passages, records and catalogues. Type a reference, a name, or a question.
102 passages seeded with text
Loading
CodexTrace
A source-linked research platform for Scripture, textual variants, manuscripts, original languages, contradictions, commentary, and ancient context. Every claim is traceable backward toward the record it came from.
30
Contradiction claims
across all collections
9
Variant units
witness data curated separately
8
Manuscripts
catalogue metadata only
15
Source records
every claim resolves to one
Counts are reported per collection. There is no single objective count of contradictions in the Bible — the number depends on the collection, canon, translation, manuscript tradition, grouping rules and definition. See Methodology.
Reference-aware search across passages, records and catalogues. Type a reference, a name, or a question.
102 passages seeded with text
Recorded claims that passages conflict, each with propositions, formal analysis, harmonisations and their premises.
30 claims in codextrace_research
Differences between manuscript witnesses at a defined variant unit — a transmission question, never a logical one.
9 variant units
Catalogue metadata and authoritative repository links. No imagery is mirrored; every record works in link-out mode.
8 manuscripts
Hebrew, Aramaic and Greek tokens, lemmas and glosses — ingested from licensed datasets, never generated.
Lexical index
Mesopotamian, Ugaritic, Egyptian and Second Temple literature, related by motif and form. Similarity is not dependence.
9 documents
Entry points into the corpus by subject, person, place, object and event.
25 topics · 29 entities
Side-by-side comparison of accounts of the same episode, on the axes each account actually addresses.
8 groups
Anyone can browse and search. Signed-in contributors can discuss records, cite evidence inline, and submit sources for editorial review. Community content is a separate, clearly-labelled layer — it never becomes canonical research automatically, and it never enters the AI trusted corpus by default.