OK: The 1839 Typo That Ate the World

On March 23, 1839, Charles Gordon Greene of the Boston Morning Post was mocking an editor at the Providence Journal, and somewhere in a throwaway sentence about contribution boxes and flying corks he stuck in the letters "o.k." — glossed, helpfully, as "all correct" [S1]. It was a partisan column nobody was meant to remember. It is now, by most popular accounts, the most widely recognized utterance produced by any human language [S5].

Greene was not being original. Educated young men around late-1830s Boston had a running joke: take a phrase, misspell it on purpose, then abbreviate the misspelling [S6]. Documented siblings of "o.k." (for "oll korrect") include "k.g." for "know go," "k.y." for "know yuse," "o.w." for "oll wright," and "s.p." for "small potatoes" [S6]. Nearly all were dead within months [S6]. OK should have died with them.

It survived because of an accident eighteen months later. In the 1840 presidential campaign, the incumbent Martin Van Buren — born in Kinderhook, New York — was nicknamed "Old Kinderhook," and his supporters formed "OK Clubs" and printed "Vote for OK" on rally signs across the country [S5]. A dying typo gag collided with a presidential nickname by coincidence. The abbreviation got a national second life it had no claim to. Van Buren lost to William Henry Harrison [S5]. The word kept going.

The infrastructure didn't exist yet

What kept it going past the campaign was infrastructure that hadn't been invented when Greene printed his column. The commercial telegraph arrived in 1844, and OK's two-character, vowel-then-consonant shape made it ideal as the acknowledgment token a clerk needed — short, hard to garble, unambiguous in transcription [S2]. Railroads adopted it for train-order acknowledgments; the 1860s transatlantic cables carried it across oceans, where it was exported globally [S2]. A pun written by a Boston editor in 1839 turned out to be perfectly engineered for technologies that would arrive five, twenty, and thirty years later.

A century of refusing to read what was printed

Here the story gets strange. For roughly 120 years, professional linguistics refused to read what Greene had actually printed [S4]. A globally dominant word, the field seemed to feel, required a dignified ancestor — and so credentialed scholars in respected journals proposed, in turn: Choctaw okeh for "it is so" [S2], Wolof waw-kay [S4], Scots och aye [S4], Greek olla kalla [S4], German ohne Korrektur [S4], Latin omnis correctus [S4], and a French sailors' port-of-call, aux Cayes in Haiti [S4]. Each had a champion. Each had its journal article. None had a documented usage chain back to the 1839 column [S2].

The Choctaw theory was the most stubborn, because it had real paper behind it. In 1825, missionaries Cyrus Byington and Alfred Wright translated portions of the Bible into Choctaw and ended sentences with the particle okeh, meaning "it is so" [S2]. The Century Dictionary picked up Byington's entry; Webster's 1913 listed okeh as an acceptable alternative spelling [S2]. President Woodrow Wilson read that footnote, believed it, and beginning around 1918 stamped informal White House memoranda with "Okeh W. W." [S2][S7]. When confused secretaries asked what he was doing, he told them the spelling came from the Century Dictionary's Choctaw derivation [S2][S7]. Franklin Roosevelt inherited the habit; State Department paperwork in the 1930s reportedly carried his okeh approval [S7].

Two American presidents, in short, endorsed federal documents on the strength of a missionary's Bible footnote because a dictionary said so. The actual origin was a partisan editor's joke at a rival paper.

One man, one archive

The case was closed by Allen Walker Read of Columbia, who spent years reading 1830s and 1840s American newspapers and published a six-article series in American Speech across 1963 and 1964 [S3]. He pinned the Boston Morning Post citation, traced the 1840 campaign rescue, and showed that no documented chain linked any exotic ancestor to the English appearance [S2]. Merriam-Webster and Wikipedia now treat his series as the definitive resolution [S6][S2]. The older dictionaries that had bet on Choctaw quietly stopped betting.

Pause on what Read's work actually destroyed. It wasn't an etymology. It was a wish — the wish that a word at OK's scale should have an origin proportional to its reach. Greene's joke was small. The fad it belonged to was small and short-lived [S6]. The presidential rescue was a coincidence inside a losing campaign [S5]. The infrastructure that carried it across the planet wasn't built for it and didn't know it existed.

What a successful token actually looks like

That is the shape of a successful communications token. Not invented brilliantly, but invented carelessly. Not chosen by users, but kept alive by an unrelated political accident. Not designed for any medium, but happening — through dumb luck of phonetic shape — to be the right size and weight for the technologies that arrived afterward [S2]. The word didn't conquer the planet because it was clever. It conquered because it was riding the right infrastructure when the infrastructure showed up.

The strong claim that OK is the single most-used word on Earth is harder to verify than to repeat; it circulates in popular books like Allan Metcalf's OK: The Improbable Story of America's Greatest Word and in journalist coverage, without a published cross-linguistic frequency survey behind it [S5]. The weaker version — that OK is intelligible across a startling number of mutually unintelligible languages — has not been seriously challenged.

An 1860s telegraph clerk and a 2026 text-message thread use the same two characters for the same purpose: received, and nothing is wrong [S2]. Neither of them knows they are completing a joke a Boston editor made about a Providence rival on a Saturday in March 1839 [S1].