Where the content comes from.

How we choose sources

We work from a fixed bank of acceptable source types, in order of preference: public-domain primary sources first, then Wikipedia (with care taken to use narrow subject-specific articles), then reputable secondary sources, and the landmarks' own official websites for their own history and logistics.

• National Register of Historic Places nomination forms (federal records, public domain).
• The Library of Congress — HABS and HAER architectural records, the Prints & Photographs Division, the Chronicling America newspaper archive.
• The Smithsonian's Inventory of American Sculpture and the Save Outdoor Sculpture! survey records.
• The Historical Marker Database (HMDB) — verbatim civic-marker transcripts.
• The Georgia Historical Society, the New Georgia Encyclopedia, and other reputable secondary references.
• The landmarks' own official websites, when we have permission.

What we don't use

Other Savannah tour-operator websites. Even with proper attribution, we don't republish competitor narration. If a claim can only be sourced from a tour-operator blog, it doesn't go into the tour.

Why this matters

Generative AI products can sound authoritative while being wrong. We treat sourcing as the product, not as a footnote — every landmark's knowledge base traces back to a primary or reputable secondary record, and gaps are honestly documented rather than glossed with invented detail.