A pioneering example of two major international museums involved in a disputed attribution of a painting asks us to rethink — and not reduce or distort — the value of different opinions about the attribution of a work of art.
In all fields, there is and should be room for different opinions. Such opinions can be independent and in disagreement, although they must be supported by sound reasoning and interpretations of the evidence. These opinions can give rise to productive discussions in which a definitive word cannot yet be reached. Art historical attributions should be no different.
This is evident in the case of the attribution of Girl with a Flute (1665-1670). In October, the National Gallery of Art in Washington, D.C., which owns four known works by Vermeer, announced that one of them, Girl with a Flute, is not by Vermeer. Rather, the museum believes that the painting was probably made by a student, apprentice or collaborator in his studio, someone who was familiar with the artist’s technique and subjects and who was close to his technique, style and subjects. It should be noted that by 1942, shortly after the painting was donated, its attribution was already being questioned. Indeed, over the years the work’s label has frequently changed from “Vermeer” to “attributed to” to “circle of” Vermeer.
To arrive at its most recent opinion, the National Gallery followed the so-called “three-legged stool” of due diligence: connoisseurship, scientific and forensic analysis, and provenance.
An initial set of observations was based on reasoned comparative connoisseurship, that is, the experience of the specialists’ eye and the evaluation of visual data. For example, the painting was found to be unsigned, but this aspect was considered negligible because 10 other works by Vermeer with secure attributions are also unsigned. In addition, the size of the painting was found to be smaller than most of Vermeer’s works. What is more, the painting was painted on a wooden panel instead of a canvas, as Vermeer typically did. Finally, Girl with a Flute was compared to another work in the collection, the Girl with a Red Hat, whose attribution is secure. Girl with a Red Hat is also very small in size and was painted on panel rather than canvas, which would suggest a correct attribution of both works to Vermeer. But the “quality” of Girl with a Flute was found to be inferior: the team concluded that the style and brushwork were “clumsy” and too different from the artist’s known precision in other known paintings.
The second set of observations was based on scientific and forensic analysis and conservation science technologies. Microscopic pigment analysis and advanced imaging performed by the National Gallery established that the pigments were coarsely ground. The surface of the painting thus appears granular rather than smooth. This would be considered atypical for Vermeer’s painting process, as scientific data collected from other known paintings suggest that Vermeer’s typical process was to coarsely grind the pigments only for the underlayer. He would then grind the pigments finely to paint the final layers of his paintings. This additional incompatibility seemed important.
A third set of observations about the work’s provenance was not made. Its existence has been documented since the artist’s lifetime, although there is still no direct connection to Vermeer
himself. It is thought that Girl with a Flute was in the possession of the family of Pieter van Ruijven, known to have been Vermeer’s patron, and was sold at a 1696 auction in Amsterdam along with other Vermeer paintings. After passing to other documented owners, the painting was purchased in 1923 by American collector Joseph E. Widener, who donated it to the National Gallery of Art in 1939.
Based on its new opinion that the painting was not by Vermeer but rather by someone close to him, the National Gallery hypothesized that, despite our fantasy of a genius painting alone, Vermeer actually had a studio, perhaps with assistants or collaborators or younger artists to whom he gave instructions. Based on this new hypothesis, the museum changed the attribution to “Studio of Johannes Vermeer.”
In all fields, there is and should be room for different opinions. Such opinions can be independent and in disagreement, although they must be supported by sound reasoning and interpretations of the evidence. These opinions can give rise to productive discussions in which a definitive word cannot ye…
In all fields, there is and should be room for different opinions. Such opinions can be independent and in disagreement, although they must be supported by sound reasoning and interpretations of the evidence. These opinions can give rise to productive discussions in which a definitive word cannot yet be reached. Art historical attributions should be no different.
This is evident in the case of the attribution of Girl with a Flute (1665-1670). In October, the National Gallery of Art in Washington, D.C., which owns four known works by Vermeer, announced that one of them, Girl with a Flute, is not by Vermeer. Rather, the museum believes that the painting was probably made by a student, apprentice or collaborator in his studio, someone who was familiar with the artist’s technique and subjects and who was close to his technique, style and subjects. It should be noted that by 1942, shortly after the painting was donated, its attribution was already being questioned. Indeed, over the years the work’s label has frequently changed from “Vermeer” to “attributed to” to “circle of” Vermeer.
To arrive at its most recent opinion, the National Gallery followed the so-called “three-legged stool” of due diligence: connoisseurship, scientific and forensic analysis, and provenance.
An initial set of observations was based on reasoned comparative connoisseurship, that is, the experience of the specialists’ eye and the evaluation of visual data. For example, the painting was found to be unsigned, but this aspect was considered negligible because 10 other works by Vermeer with secure attributions are also unsigned. In addition, the size of the painting was found to be smaller than most of Vermeer’s works. What is more, the painting was painted on a wooden panel instead of a canvas, as Vermeer typically did. Finally, Girl with a Flute was compared to another work in the collection, the Girl with a Red Hat, whose attribution is secure. Girl with a Red Hat is also very small in size and was painted on panel rather than canvas, which would suggest a correct attribution of both works to Vermeer. But the “quality” of Girl with a Flute was found to be inferior: the team concluded that the style and brushwork were “clumsy” and too different from the artist’s known precision in other known paintings.
The second set of observations was based on scientific and forensic analysis and conservation science technologies. Microscopic pigment analysis and advanced imaging performed by the National Gallery established that the pigments were coarsely ground. The surface of the painting thus appears granular rather than smooth. This would be considered atypical for Vermeer’s painting process, as scientific data collected from other known paintings suggest that Vermeer’s typical process was to coarsely grind the pigments only for the underlayer. He would then grind the pigments finely to paint the final layers of his paintings. This additional incompatibility seemed important.
A third set of observations about the work’s provenance was not made. Its existence has been documented since the artist’s lifetime, although there is still no direct connection to Vermeer
himself. It is thought that Girl with a Flute was in the possession of the family of Pieter van Ruijven, known to have been Vermeer’s patron, and was sold at a 1696 auction in Amsterdam along with other Vermeer paintings. After passing to other documented owners, the painting was purchased in 1923 by American collector Joseph E. Widener, who donated it to the National Gallery of Art in 1939.
Based on its new opinion that the painting was not by Vermeer but rather by someone close to him, the National Gallery hypothesized that, despite our fantasy of a genius painting alone, Vermeer actually had a studio, perhaps with assistants or collaborators or younger artists to whom he gave instructions. Based on this new hypothesis, the museum changed the attribution to “Studio of Johannes Vermeer.”
