About this project
What this is
SurnameDB covers 10,000+ of the most common US surnames (2010 Census) plus a handful of reader-requested additions, each traced back to 49 etymological dictionaries published before 1931 — including works by Charles Bardsley, Mark Antony Lower, Henry Harrison, Patrick Woulfe, Robert Matheson, and Alexander Macbain — and rewritten in plain modern English.
Why pre-1931 sources
Books published before 1931 are in the public domain in the United States, so their text can be used and quoted freely. We pulled the original scans from the Internet Archive, found the passages that actually discuss each surname, and synthesized a modern explanation from them — flagging anywhere the historical sources disagree or the old print itself was unclear.
What "confidence" means
Sourced — the entry is grounded in a genuine dictionary headword entry for that exact surname. Mixed — grounded in historical excerpts, but they were incidental mentions rather than a full dictionary entry, or sources partially disagreed. Pattern-inferred — no historical source in our corpus mentioned this surname (common for names of non-European origin, which 19th-century British/European dictionaries naturally don't cover); the entry instead relies on general, well-established onomastic (name-origin) knowledge, clearly labeled as such rather than presented as a specific historical citation.
What we deliberately don't do
We don't sell coats of arms or "family crests." A surname is not a family — millions of unrelated people share most surnames, and the historical "crest for your name" industry is largely a marketing myth. We'd rather tell you the honest etymology and point you to real historical sources than sell you a certificate.
Use the data elsewhere
The dataset is available as a small JSON API for other apps, sites, and AI agents to query — see openapi.json and llms.txt.