Van Meegeren Forgery Case
In 1937, Dutch painter Han van Meegeren produced a canvas titled The Supper at Emmaus and sold it to the Boijmans Museum in Rotterdam as a newly discovered work by Johannes Vermeer. The painting was celebrated by critics as a masterpiece, exhibited prominently, and became one of the most valued works in the museum's collection. Van Meegeren repeated the success with additional 'Vermeers' sold to private collectors and museums, accumulating wealth and critical acclaim. His confession came in 1945 under peculiar circumstances: arrested for collaboration with the Nazis (one fake Vermeer had been sold to Hermann Göring), van Meegeren avoided the collaboration charge by confessing to forgery—a lesser crime. The art world convulsed. The paintings had not changed—the pigment, composition, and visual properties were identical before and after the confession. What changed was the knowledge of origin, and that knowledge transformed the works from masterpieces to frauds. Goodman used this case to demonstrate that in autographic arts, the history of production is not merely of biographical interest—it is constitutive of the work's identity and its aesthetic achievement. The paintings could not be the same works after the confession, because what they achieved in the context of seventeenth-century Dutch painting (a breakthrough in religious representation) versus what they achieved in the context of twentieth-century pastiche (a skilled but derivative imitation) depended entirely on when and by whom they were made.
In the AI Story
The van Meegeren case became the canonical illustration of Goodman's autographic/allographic distinction. In an autographic art, a work cannot be correctly instanced by anyone other than the originating artist, because there is no notation specifying the work's identity independently of its productive history. The history is part of what the work is—and 'is' here is not merely metaphysical but aesthetic. What the painting achieves, what it contributes to the development of the art form, what it exemplifies about the possibilities of the medium, depends on its place in a historical sequence of innovations. A Vermeer advances the sequence; a van Meegeren imitates it. The formal properties may be indistinguishable, but the art-historical achievement is categorically different.
The case has acquired new relevance in the age of AI. When an AI system generates an image in the style of Vermeer—deploying the same compositional structures, lighting techniques, and paint-handling conventions—the image is not a Vermeer, and not even a forgery of a Vermeer (forgery requires deceptive intent). It is a rendering of Vermeer-style visual properties through statistical pattern-matching. The rendering may be perceptually impressive. But it achieves nothing in the art-historical context that Vermeer's work inhabits, because it does not extend, challenge, or deepen the tradition—it samples from it. The sampling is visible in the output's smoothness: where Vermeer's brushwork carries the evidence of decisions made under the resistance of the medium, the AI's rendering is optimized for statistical plausibility, and the optimization eliminates precisely the roughness, the idiosyncrasy, the evidence of struggle that autographic value depends on.
Van Meegeren's technical skill was extraordinary—he developed aging techniques, pigment formulas, and compositional strategies that fooled the most sophisticated experts of his day. The skill was not sufficient for the paintings to be Vermeers, because what makes a painting a Vermeer is not technical mastery but historical position—the specific contribution the work makes to the development of pictorial representation at a specific moment in that development. An AI's technical skill in generating Vermeer-like images is, in relevant respects, even greater than van Meegeren's. The historical contribution is not merely absent but impossible, because the AI is not positioned in the historical sequence—it does not advance the tradition or react against it or deepen it. It replicates surface properties extracted from works that were positioned. The replication is rendering without worldmaking, form without historical substance.
Origin
Van Meegeren's career as a forger began in the mid-1930s, motivated partly by resentment of critics who had dismissed his own original work as derivative. He studied Vermeer obsessively, developed techniques for aging paint to pass chemical analysis, and produced six fake Vermeers between 1937 and 1943. His exposure came through the Göring sale: when Allied forces recovered a 'Vermeer' from the Nazi leader's collection after the war, Dutch authorities charged van Meegeren with collaboration for selling national treasures to the enemy. To avoid the collaboration charge (which carried a death sentence), van Meegeren confessed to forgery and proved his claim by painting another fake Vermeer under supervision in 1945. He was convicted of forgery, sentenced to one year in prison, and died of a heart attack before serving the sentence. Goodman's use of the case in Languages of Art made it the most widely discussed art forgery in philosophical literature, establishing it as the paradigm case for autographic identity.
Key Ideas
Perception cannot distinguish. The paintings looked like Vermeers before the confession and looked identical after—visual properties did not change, yet the works' identities and achievements changed completely.
History is constitutive. What a painting achieves in its art-historical context depends on when and by whom it was made—autographic value is positional, not merely formal.
AI rendering is not forgery. AI-generated Vermeer-style images do not forge (no deceptive intent) but sample—replicating surface properties without the historical positioning that makes autographic contributions genuine.
Technical mastery is not sufficient. Van Meegeren's skill and AI's skill both produce perceptually convincing outputs; neither produces the art-historical contribution that autographic identity requires.
Appears in the Orange Pill Cycle
Further reading
- Edward Dolnick, The Forger's Spell (Harper, 2008)—narrative account of van Meegeren
- Jonathan Lopez, The Man Who Made Vermeers (Harcourt, 2008)
- Denis Dutton, ed., The Forger's Art: Forgery and the Philosophy of Art (California, 1983)
- Alfred Lessing, 'What Is Wrong with a Forgery?' Journal of Aesthetics and Art Criticism 23 (1965)