Our Methodology

AI-Assisted Historical Research

Caribbean Name Archive uses OpenAI language models to synthesize evidence from a curated documentary base. AI does not invent sources or facts — it organizes, cross-references, and presents evidence drawn from real historical records. Every significant claim cites a specific source.

Confidence Ratings

Every claim in our reports carries one of four confidence ratings:

• HIGH — Supported by multiple corroborating primary sources.
• MEDIUM — Supported by a single primary source or multiple secondary sources.
• LOW — Contextual inference from related evidence.
• SPECULATIVE — A plausible hypothesis without direct evidence.

What We Will Never Claim

• That a surname proves ethnic origin, DNA ancestry, or genetic makeup.
• That a surname proves descent from a specific African people or kingdom.
• That a surname proves descent from royalty or entitlement to a coat of arms.
• That a surname identifies a specific enslaved ancestor.
• Fabricated or unverifiable sources.

Our Source Base

Our documentary base includes Barbados Slave Registers (1817–1834), Jamaican parish registers, civil registration indexes, emancipation records, plantation and probate records, Windrush migration lists, historical newspapers, onomastic dictionaries, and peer-reviewed academic sources. Sources are checked for licensing before inclusion.

Limitations

Caribbean Name Archive is educational research, not certified genealogy. It cannot build individual family trees, confirm biological ancestry, or replace professional genealogical services. Historical records for the Caribbean are fragmentary — many enslaved people were deliberately denied surnames, and record survival is uneven across parishes and periods.