About Vincent van Gogh Paintings

Van Gogh painted thick and fast, and a print flattens exactly that. A copy painted in oil keeps the ridges of the brush. The catalogue below dates every work from the record, place by place.

The problem here is chronology, not subject

A catalogue built around a subject spends its effort deciding what a picture shows. This one does not have that problem. Every work is by the same hand, and what it depicts is rarely in doubt. The difficulty is when and where, and on a body of work made in ten years across five places, that is the difference between a coherent catalogue and a pile.

The whole painted output falls between 1881 and July 1890: the Brabant years, two years in Paris, fifteen months in Arles, a year inside the asylum at Saint-Rémy, and seventy days at Auvers-sur-Oise. Those five divisions are the spine of this site, because they correspond to abrupt and visible changes — in palette, in speed, in what was in front of him — rather than to a decade boundary.

Dated from the record, not from the title

Period assignment here comes from the recorded date of execution, and from nothing else. This sounds obvious and is not: an earlier version of this catalogue placed works by matching words in the title, and the results were wrong in ways a reader could see. A canvas painted at Montmartre in 1887 was filed under Arles, 1888, because its title mentioned sunflowers. Two Arles pictures were filed under Saint-Rémy for equivalent reasons.

The date of execution is recorded for essentially the entire painted corpus, so there is no reason to guess from a word. Where a museum publishes a range rather than a year, the range is what appears, and a work that straddles a move between two towns is shown in both periods rather than forced into one.

The letters are part of the record

Van Gogh is unusual among painters in leaving several hundred letters that describe canvases as they were being made, frequently with a sketch of the composition and a line about what was still wrong with it. That correspondence, published and openly accessible, is the reason the chronology can be as tight as it is.

Where a date given here rests on a letter rather than on a signature or an exhibition record, the entry says so. A dating that depends on a described sketch is a good dating, but it is a different kind of evidence from a dated invoice, and a reader is entitled to know which they are looking at.

The corpus is not closed

Works have moved into and out of the accepted body of paintings repeatedly over a century, and they still do. Some pictures long treated as autograph have been reassigned to friends, students or forgers; at least one dismissed for decades was restored to the corpus after technical examination.

This catalogue records the status a work currently has in the published literature, names the institution or catalogue that gives it that status, and where a work is disputed says that it is disputed. It does not arbitrate attribution, and an entry appearing here is not an opinion that a painting is genuine.