The morgue · Chronicling America, read live
Every paper keeps
a morgue. This is ours.
The Library of Congress and the National Endowment for the Humanities have been scanning America’s historic newspapers since 2005 — more than twenty million pages by the Library’s own count, from 1789 to 1963, every state in the union. All of it sits behind one open JSON API. This page reads it directly: type a word and the archive answers with the actual pages it ran on, OCR snippet first, the Library’s own page scan one click away.
Then the instrument: hand the same word to the ink map and it asks the archive fifty-one times — once per state, the District included — and inks a drawn US map with where the word ran, on a scrubber that walks it decade by decade. No key, no account, no server in the middle.
3,205,306 issues on the wire · 1789–1963 · the Library of Congress’s own API · read in your browser
The search
A word, through 174 years of newsprint.
Quotes make a phrase. Narrow it to a state or a span of years if you like. Each hit is one page; the card carries the paper’s name, where it printed, the date, and the OCR reading around your word — and links to the Library’s scan of the page itself.
The morgue desk
waiting for the wire
Not read yet.
The ink map
Where a word ran.
Give the map a word and it asks the archive once per state — a slow question asked fifty-one times, paced to the Library’s rate limit, progress drawn as it goes; expect a minute or two. Then the states ink by how many pages the word appears on. Drag the scrubber and the same ink re-pools decade by decade, 1789 to 1963, from counts already fetched. Darker ink, more pages.
The ink map
idle
Hover a state for its count. Click one to aim the search above at it.
All decades · 1789–1963
Counts are page hits — a page counts once however many times the word appears on it, and a hit is the OCR’s reading, not an article. States digitize at different rates, so the map inks the archive’s coverage as much as history; the territories’ papers are in the archive but not on this map.
About
A reading room, not a library.
Chronicling America is the National Digital Newspaper Program — the Library of Congress and the National Endowment for the Humanities paying state institutions, since 2005, to scan their historic papers to one standard. The result is the largest open archive of American newsprint there is: over twenty million pages by the Library’s count, and 3,205,306 issues on the API this page reads. Nothing here is hosted by LABS — every search is answered by loc.gov directly, and every page image link goes to the Library.
The archive runs from 1789 to 1963 and stops there on purpose: 1963 is where copyright gets complicated, so the morgue ends where the public domain does. Coverage is not even — each state chose which of its papers to scan and how many, so an empty state on the ink map can mean the word never ran there, or that the papers it ran in have not been scanned yet. The map cannot tell those apart, and does not pretend to.
The text you search is OCR — a machine’s reading of smudged column inches, and it misreads plenty. A word broken across a line hyphen is two words to the machine; an f can be an s in old type. What the feed does not say, this page does not invent: snippets are shown exactly as the API returns them, misreadings included, and the cure for doubt is one click away on the Library’s own scan. Your last search and each inked word’s counts are remembered in this browser and nowhere else.