Astronomers Say Van Gogh's Starry Night Shows a July 1889 Dawn

The Starry Night is Vincent van Gogh's most famous painting, and for a long time it has been dated to June 1889. A team of astronomers now says the sky in the picture matches a real sky before dawn on July 22, 1889. They reached that date by comparing the painting with old weather records, the artist's own letters, and even the timing of cicadas.

The crescent moon that did not fit

For decades, the painting was read as imagination or symbolism rather than a real scene. Then a letter from van Gogh mentioning Venus changed that. At least one of the swirly glowing orbs, the one just above the horizon to the right of the cypress tree's branch, was accepted as a real object in the sky. The rest was still treated as memory and invention.

In the 1980s, the art historian Albert Boime used the position of that planet as a guide and matched it to reconstructions of the night sky. He concluded that three stars in the painting corresponded to the constellation Aries as it was visible from the location on June 19, 1889. That date came from letters van Gogh had sent to his brother, along with sketches of the work.

There was one glaring problem: the moon. Looking out from the window of his asylum in late June, van Gogh would have seen a waning gibbous moon, a moon more than half lit, not the thin sliver he painted. Boime put the difference down to artistic license. He noted the soft glow around the crescent and what looked like paint scraped from its curved surface, which he said could mean van Gogh painted the fuller moon and then changed his mind.

A V shape above the horizon

The new study, published in Scientific Reports, says the real clue was the group of stars Boime identified. "To an observational astronomer, the cluster of stars in The Starry Night is the Hyades in Taurus," study author Elijah Jeong-hyun Kim of Kyung Hee University in South Korea said in a statement. "But the moment I read the earlier studies explaining it as Aries, I heard the painting calling to me."

The Hyades form a V shape. On July 22, 1889, they would have been visible alongside Venus near the horizon with a crescent moon, all before dawn. A shape analysis matched five central stars in the painting to the Hyades, and the match was even better with a sketch van Gogh made after finishing the painting.

Then the team hit another snag. Most scholars agree that van Gogh suffered a seizure around mid-July that left him unable to paint for some time. If that timeline is right, the proposed date would be impossible.

Wind, heat and cicadas

So the researchers went back to the letters. In one of them, van Gogh said the seizure struck while he was painting on a windy day. Weather records from that time show only intermittent winds from July 15 to 26, with strong mistral gusts starting on July 27. A windy day, in other words, fits late July better than the middle of the month.

Van Gogh also wrote about The Starry Night together with other summer paintings, and he mentioned the sweltering climate and the cicadas. In June there were only four days above 86 degrees Fahrenheit, which the researchers call cicada weather. July had eighteen, including more than two weeks of consecutive scorchers.

None of this proves the date. The researchers say it suggests that day was the focal point of van Gogh's observations, and they call for the June dating to be reconsidered. Their method was first tried out on night paintings by Elsheimer and Millet, and checked against two night paintings from 1888 whose locations still exist. Whether the stars in the painting came from a single real night or from the artist's memory may never be settled, but the picture itself has not stopped attracting people who want to work it out.