Van Gogh’s “Starry Night” May Depict the Real Sky of 22 July 1889

October 23 — A study published on 22 October in Scientific Reports proposes that the swirling night sky in Van Gogh’s “Starry Night” was not pure imagination: it was most likely inspired by the arrangement of celestial bodies genuinely visible before dawn on 22 July 1889 outside the asylum in Saint-Rémy, France.

If that judgment holds, the birth of this post-impressionist masterpiece moves from the long-accepted June 1889 to late July.

The traditional dating rests on a letter Van Gogh wrote to his brother Theo around 18 June 1889, in which he mentioned “a new study of a starry sky.”

The paper’s first author, Elijah Kim Jeong-hyun — CEO of the Korean space science company SLLAB and a doctoral student at Kyung Hee University — points out a long-avoided contradiction in that date: in mid-June 1889 the Moon over Saint-Rémy was a waning gibbous, an oblique oval in shape, whereas the Moon in “Starry Night” is a thin crescent. June’s lunar phase does not match the painting.

Kim’s team fed the relative positions of the five central stars in the painting into shape-analysis software called Procrustes and, after excluding factors an artist might subjectively embellish such as size and colour, compared them with several candidate constellations. The star group in the painting matched best with the V-shaped Hyades in Taurus, fitting statistically better than the Aries interpretation proposed earlier.

A co-author and Kyung Hee University professor explained that to an observational astronomer, the cluster in the painting “is the Hyades in Taurus,” and that when earlier work read it as Aries, Kim felt “the painting was calling to me.”

Combining the lunar phase with the cluster’s position narrowed the window to around 22 July 1889. In the dawn hours of those days, a waning crescent Moon, Venus and the Hyades appeared together within the painter’s narrow field of view, and the configuration broadly matches the arrangement on the canvas. Historical weather records show clear skies in those days, with visibility good enough for Van Gogh to see the event. The study also cross-checked Van Gogh’s frequent mentions of cicadas in his letters — cicada song means hot weather, and July 1889 temperatures match that clue better.

The new date is far from settled, however. José Luis Aragon, a physicist at the National Autonomous University of Mexico who was not involved in the study, called the work “meticulous and genuinely interdisciplinary,” but still regards 22 July as “an interesting hypothesis rather than a settled answer.”

Aragon also notes that Van Gogh painted in the asylum studio from sketches and memory, so the sky in the picture “is very likely a composite of several nights” — a point the authors themselves concede. Don Olson, professor emeritus at Texas State University, told Scientific American that Venus’s position relative to the Hyades’ V “does not match the painting,” and that Van Gogh’s 18 June letter explicitly mentions “a new study of a starry sky”; overturning that would need “strong evidence.”

Even with the date in dispute, Aragon considers the study’s most convincing element to be a broader conclusion: “This painting is not pure fantasy. Van Gogh looked at the real sky very carefully.”

It is worth noting that “Starry Night” has been in the collection of the Museum of Modern Art in New York since 1941, and the museum still labels it “Saint-Rémy, June 1889.”