The yellow crescent in The Starry Night creates an awkward problem: it does not look like the Moon that should have been over Saint-Rémy during the painting’s usual mid-June date. A new astronomical reconstruction points instead to a July morning.
The proposed answer is unusually specific: before dawn around July 22, 1889, the Moon, Venus and a V-shaped group of stars could have occupied roughly the arrangement van Gogh painted. But this is not a solved mystery. The reconstruction challenges evidence in the artist’s own letters, and an independent astronomy historian disputes its view of the landscape. The intriguing part is the collision between two different kinds of clock: the moving sky and the written record.
The archival MoMA film below introduces the painting itself. Published in 2010, it is not a visualization of the new astronomical reconstruction.
A crescent that will not cooperate
The researchers did not start by treating every swirl as a literal astronomical object. Their question was whether identifiable lights could anchor the scene to a plausible sky. In the publisher’s account of the findings, the conventional mid-June timing produces a Moon substantially fuller than the painted crescent. Their proposed July arrangement includes Venus and the Hyades, a star cluster in Taurus.
The reconstruction combined astronomical simulations, earlier landscape sketches, historical weather and correspondence. Weather records suggested that the proposed July sky could have been visible; references to cicadas offered another, much softer clue about summer conditions. Neither a clear night nor insects in a letter can establish when a canvas was finished. The authors themselves allow for observations from several nights being combined.
Why July 22, rather than any convenient night?
In the October 22 Scientific Reports study, Elijah Jeong-hyun Kim and colleagues reconstruct a narrow view from a landscape drawing. Within that assumed frame, the spacing of the Moon and Venus becomes a constraint, rather than two objects that can be moved wherever the theory needs them.
At 3 a.m. local mean time on July 22, their calculated separation was about 18.6 degrees. On adjacent mornings, the spacing changed sharply. The researchers also compared the arrangement of five painted stars with the Hyades using a shape-analysis method.
The paper treats that resemblance as a consistency check, not a unique identification by itself. Its argument depends on the landscape framing, lunar phase, Venus and star pattern working together.
The letter that refuses to disappear
There is a substantial obstacle to moving the story into July. As Science News reports, van Gogh told his brother Theo about a new study of a starry sky in a letter dated around June 18. That is a reason for the June dating, not an arbitrary museum tradition.
The study’s authors argue that the June letter could indicate a work begun, rather than a completed canvas. That proposed reinterpretation is part of their case, not independent confirmation.
Donald Olson, an astronomy historian at Texas State University, also challenges the reconstruction itself. He argues that the painting’s hills and cliffs require a broader view than the team assumes. He questions the proposed identification of Aldebaran, the conspicuous star associated with the Hyades pattern.
Those objections matter because they attack the route to the answer, not merely the answer’s unfamiliar date. If the frame is wrong, a sky that fits that frame is not necessarily the sky van Gogh used. And if the stars are being identified incorrectly, a persuasive-looking alignment can lose its force.
It was never a photograph
The Museum of Modern Art’s collection record still dates the painting to June 1889. Its account already describes a mixture of observation and invention: a nighttime inspiration painted during daytime sessions, a village that was not visible from the artist’s window, and a cypress brought closer in the composition.
That matters more than an argument over whether the sky is “real.” A painter can remember a bright planet accurately while moving a tree, borrowing a village or changing the light. MoMA also identifies Venus in the painting, so the presence of a genuine celestial object is not the new discovery.
What the new study offers is a proposed relationship between several objects and a particular July sky. What it has not established is that van Gogh painted this canvas on that date, or that the June evidence should be discarded.
Look again at the crescent, then the bright point left of center, then the scattered stars between them. They are now the components of a testable, disputed historical argument. The painting has not surrendered a secret date. It has given researchers something more interesting than a tidy answer: a sky that may fit, and a letter that still needs explaining.
Featured image: Vincent van Gogh, The Starry Night (1889), via Google Arts & Culture and Wikimedia Commons; public domain. Full composition reproduced without cropping.
