Attributed to Rembrandt Harmenszoon van Rijn( 1606-1669) or his Studio, Young woman in fantasy costume, first half 1650s, Oil on canvas, 74.0 x 61.0 cm, The Louvre, Paris
For most of its documented life, nobody called this painting Hendrickje. It first surfaces in the collection of Jean de Jullienne (1686–1766), the Parisian silk manufacturer and print publisher best known as the greatest collector and champion of Antoine Watteau’s (1684–1721) work. Jullienne is cited as its owner in 1754 by Descamps, and the painting appeared in his posthumous sale in Paris on 30 March 1767 as lot 133. From there it passed through a sequence of distinguished French collections: to Harenc de Presle, then to the Duc de La Vallière, at whose sale in 1781 it was catalogued as Portrait d’une belle femme, portrait of a beautiful woman. Three years later it appeared in the sale of the Comte de Vaudreuil (1740–1817), a celebrated collector and intimate of Marie Antoinette (1755–1793), under the same title, where it was acquired by the dealer Alexandre-Joseph Paillet on behalf of Louis XVI (1754–1793). It entered the French royal collection as an anonymous beauty by Rembrandt. It stayed that way for almost a century.
The name arrived in 1883, when the German art historian Wilhelm von Bode (1845–1929) proposed that the model was Hendrickje Stoffels (c.1626–1663), Rembrandt’s companion and the mother of his daughter Cornelia, born in 1654. The identification stuck. It has been repeated in catalogues, monographs, and wall labels ever since, and the Louvre’s own title for the painting long reflected it: Portrait de Hendrickje Stoffels au béret de velours. The painting is unsigned. No documentary evidence connects it to a specific commission, and no technical study has conclusively established which passages, if any, are by Rembrandt’s own hand.
There is something worth pausing over in that gap between what the eighteenth century saw and what the nineteenth century wanted. When this canvas was sold in 1781 and 1784 as Portrait d’une belle femme, its French owners felt no need to identify the sitter. The painting was valued for what it was: a richly painted half-length figure in costume, attributed to a celebrated Dutch master, attractive enough to hang in aristocratic interiors alongside Watteaus and Bouchers. It was the nineteenth century, with its appetite for romantic biography and its conviction that great paintings must be confessions wrung from lived experience, that required the woman to have a name and a story. Bode supplied both.
The broader question this painting raises, and which the Louvre’s quiet reclassification throws into sharper relief, is what we lose by insisting on the binary of Rembrandt or not-Rembrandt. The modern obsession with concrete attribution, with sorting every canvas into the master’s own hand or the workshop’s lesser one, is itself a historical artefact, a product of the market’s need for certainty and the museum’s need for hierarchy. Svetlana Alpers, in Rembrandt’s Enterprise: The Studio and the Market (1988), argued that Rembrandt’s studio functioned less as a traditional master-pupil workshop and more as what she called an ‘enterprise,’ a controlled environment in which the master’s manner of painting, his touch, his way of handling models and light, became the product being sold. What the studio offered to the market was not a series of individually authored canvases but a recognisable mode of picture-making: the warm tonality, the fantasy dress, the half-length figure emerging from shadow, the face painted with greater sensitivity than the surrounding costume. Whether the master himself laid every brushstroke was, in this reading, beside the point. The enterprise produced Rembrandts.
This is not an argument for indifference. It matters, for the history of painting, to understand what Rembrandt did with his own hands and what he delegated. But the half-length female figures of the 1650s, of which this Louvre canvas is one, were produced within a studio culture where delegation was the norm, not the exception. Rembrandt’s pupils during this period, among them Ferdinand Bol (1616–1680), Govert Flinck (1615–1660), and later Arent de Gelder (1645–1727), produced their own versions of the type, sometimes close enough to the master’s manner to cause lasting attribution problems. The Rembrandt Research Project, across its six-volume A Corpus of Rembrandt Paintings (1982–2014), downgraded, queried, and sometimes restored the authorship of numerous such works, and the revisions introduced by Ernst van de Wetering (1942–2021) in the later volumes openly acknowledged that the rigid categories of the early Corpus (A for autograph, B for doubtful, C for rejected) had failed to capture the reality of collaborative production. What actually happened in the studio on the Breestraat, and later on the Rozengracht, was messier and more interesting than a sorting exercise. A pupil might lay in the background and costume; the master might paint the face; someone else might glaze the whole. Or none of that. Or all of it in a different order. The finished painting entered the market carrying the studio’s authority, not necessarily the master’s signature, and it was bought and sold on that basis.
So where were these paintings, and who wanted them? Amsterdam’s kunsthandel in the 1650s had a strong appetite for half-length female figures, particularly those carrying an air of poetic refinement or historical allusion. They were not portraits in any commemorative sense. They were tronies, character studies dressed in fantasy costume, and they circulated as independent pictures, collected for their painterly quality and their evocation of a generalised feminine beauty that owed something to the Venetian belle donne tradition of a century earlier (a debt I have discussed in relation to the National Gallery’s Half Figure of a Woman with a White Wrap and will not rehearse again here). Dagmar Hirschfelder has demonstrated that tronies occupied a distinct commercial niche in the Dutch art market, functioning simultaneously as training exercises for pupils, demonstrations of painterly skill, and saleable goods for collectors who cared more about quality of execution than the identity of the sitter. The 1656 insolvency inventory of Rembrandt’s possessions lists various heads and tronies without identifying the models, a pattern consistent with the way the market understood them. They were stock, not biography.
And yet the desire to make them biographical persists, and this painting has been particularly susceptible to it. The face that turns towards the viewer in warm half-light, the velvet beret, the suggestion of intimacy, the feeling that the painter and the model knew each other well: all of this invites the biographical reading, invites the name Hendrickje, invites the romantic narrative of the artist and his common-law wife working together in defiance of social censure. It is a seductive story, and it may even be partly true. But it is worth remembering that Bode, writing in 1883, was building his identification on resemblance to other paintings that were themselves identified on the basis of resemblance to each other. The circle is not grounded in documentary evidence at any point. The Louvre’s own catalogue description, noting that the model has been ‘convincingly’ identified as Hendrickje since Bode, registers the peculiar status of an attribution that rests on consensus rather than proof.
Perhaps the more useful question is not ‘Is this Hendrickje?’but ‘Does it matter?’ If the painting is a workshop product, made within Rembrandt’s enterprise for the Amsterdam market, dressed in costume that self-consciously recalls an older tradition of idealised female portraiture, and sold to French collectors as nothing more specific than the portrait of a beautiful woman, then the most interesting thing about it is not who sat for it but how it was made, how it was sold, and what it tells us about the economy of image-making in a mid-seventeenth-century Dutch studio. The painting does not need a name. It needs to be understood on the terms under which it was produced: collaboratively, commercially, and with a studied ambiguity about identity that was not a failure of documentation but a feature of the genre.
References
Alpers, S. (1988) Rembrandt’s Enterprise: The Studio and the Market. Chicago: University of Chicago Press
Bruyn, J., Haak, B., Levie, S.H., van Thiel, P.J.J. and van de Wetering, E. (1982–2015) A Corpus of Rembrandt Paintings. 6 vols. Dordrecht: Springer
De Winkel, M. (2004) Fashion and Fancy: Dress and Meaning in Rembrandt’s Paintings. Amsterdam: Amsterdam University Press
Hirschfelder, D. (2008) Tronie und Porträt in der niederländischen Malerei des 17. Jahrhunderts [Tronie and Portrait in Seventeenth-Century Netherlandish Painting]. Berlin: Gebr. Mann Verlag.
Musée du Louvre (n.d.) Portrait de Hendrickje Stoffels (1625–1662) au béret de velours, INV 1751. Département des Peintures, collections database. Available at: https://collections.louvre.fr/ark:/53355/cl010062019 (Accessed: 3 May 2024)
Liedtke, W. (2007) Dutch Paintings in the Metropolitan Museum of Art. 2 vols. New York: Metropolitan Museum of Art
Van de Wetering, E. (2017) A Corpus of Rembrandt Paintings VI: Rembrandt’s Paintings Revisited: A Complete Survey. Dordrecht: Springer
RKD – Netherlands Institute for Art History (n.d.) Attributed to Rembrandt, Young woman in fantasy costume, first half 1650s RKDimages, image no. 41222. Available at https://rkd.nl/images/41222 (Accessed 3 May 2024)