The Man Who Made the First One
The crossword was invented by a man from Liverpool, for an American newspaper, as filler for a Christmas edition. Its inventor found it a chore, and the paper's typographers resented setting it.
The Liverpool connection
Arthur Wynne was born in Liverpool in June 1871, the son of a man who edited the Liverpool Mercury. He sailed for America in 1891, aged nineteen, and spent the next twenty years drifting through newspaper work in Pittsburgh, playing violin in the city's symphony orchestra on the side, before arriving at the New York World.
By 1913 he was running FUN, the World's Sunday jokes-and-puzzles supplement, and by December he needed something to fill the Christmas number. He reached back to what he had grown up with - the word squares that Victorian English magazines printed by the yard, which he had solved as a boy under his grandfather's instruction. Then he did the one thing nobody had done: he broke the square.
The first grid
FUN's Word-Cross Puzzle ran on 21 December 1913, on page six of the supplement. It looks strange now.
It's a diamond, not a rectangle, with a hollow centre - a ring of white squares around a hole. There are no black squares anywhere in it, whatever a great many sources tell you. Wynne came to the black square later; it's not in the puzzle that started everything.
The numbering was different too. Modern grids number where a word begins and tell you which way to read. Wynne numbered both ends: clues ran "2-3", "10-18", "6-22", so you were told where each answer started and where it stopped. And the top row was already filled in, because the word sitting there was FUN - the name of the section, doing double duty as the first entry.
One of the original clues, for a four-letter answer, was "What this puzzle is." The answer was HARD.
A diamond with a hole in the middle, and not one black square in it.
Word-cross to crossword
The name flipped early. The received story is that a typesetter reversed it by accident and Wynne, who doesn't seem to have been precious about any of it, let it stand. It's repeated everywhere and traceable nowhere.
What is documented is that he found the whole business a chore, and that the World's typographers resented setting the awkward diamond shape. It survived because readers complained the one Sunday it was missing. Wynne noted that his editor had calculated the supply of reader submissions would last "until the second week in December, 2100."
He made almost nothing from it. In December 1913 he had one grid to fill; by 1915 readers were sending him twenty-five a day. He died in Florida in January 1945.
What came before
Wynne invented less than the word "inventor" suggests, and more than the debunkers allow.
Word squares are the real ancestor - the same words reading across and down, a Victorian parlour staple, and the form Wynne grew up solving. Acrostics were already ordinary newspaper fare.
Giuseppe Airoldi has a genuine claim to something. In September 1890 he published a four-by-four grid in a Milan weekly, under the heading "Per passare il tempo" - to pass the time - with a separate definition for each row and each column. That's a clued interlocking grid, twenty-three years early. Italian sources are honest about the limits of it: no black squares, and therefore no way for the form to grow beyond a neat little box. The priority claim was made in 1959, sixty-nine years after the fact, by an Italian enthusiasts' convention with an obvious stake in the answer. Airoldi died on 13 December 1913 - eight days before Wynne's puzzle appeared.
The Sator square, the Latin palindrome that turns up in Roman ruins, is not an ancestor of anything. It has no clues and no solver. It's the oldest surviving word square, which is a real distinction, and it's regularly mis-described as having been dug up at Pompeii in 1868 - the 1868 find was at Cirencester.