This magnificent portrait, first published in 1942 by Georges Wildenstein, generally has been thought to represent Marguerite Josèphe Richard, the first wife of Pierre Jacques Onésyme Bergeret (later called Bergeret de Grancourt) (1715 – 1785), one of François Boucher’s most enthusiastic patrons and the first recorded owner of the painting. It depicts a young woman, resplendent in a full-length gray satin gown, standing in an intimate corner of a garden. A bright bow is tied to her bodice, a spray of flowers is pinned to her sleeve, and a simple straw hat hangs at her side, held by a ribbon tied to her elbow. Her powdered hair, neatly pulled back, is adorned with more flowers; a delicate lace ruff highlights her neck, and her sole piece of jewelry is a pearl bracelet with a cameo portrait. Despite its splendor, her costume is meant to suggest the informal wear of a country girl or shepherdess. The setting reinforces the countrified air of the sitter: she stands in a corner of an overgrown bower or garden, the rampant foliage providing shelter and shade. A wooden bench, visible at the left edge of the composition, is balanced at the right by a stone urn, on which the woman gently rests a hand, her fingers grasping a rose.
There seems little reason to doubt the traditional identification of the sitter as Madame Bergeret. As Colin Eisler astutely observed, the sitter’s rustic, if elegant, costume should be understood as a witty pun on the woman’s name, which is close to the French word for shepherdess, bergère. There is also the provenance of the painting, which can be traced through Bergeret de Grancourt’s heirs until it was sold to Wildenstein around 1920. An inventory of Bergeret de Grancourt’s possessions, drawn up at his death in 1785, included several portraits of Madame Bergeret by or after Boucher and other artists. Wildenstein identified this painting as that listed under no. 22, recorded hanging in a small gallery just off the salon. As was customary for family portraits, the presence of the work was noted but not appraised. Although the inventory did not specify which Madame Bergeret was represented (Bergeret had remarried twice after the death of his first wife), the date visible at the lower left of the painting, 1746, logically suggested to Wildenstein that the sitter had to be the first Madame Bergeret, Marguerite Josèphe Richard, whom Bergeret de Grancourt had married in 1741. The portrait was inherited by their first son, Pierre Jacques Bergeret de Grancourt (1742 – 1807).
The identification and dating of the National Gallery of Art’s portrait was accepted until 1986, when Alastair Laing raised doubts. He pointed out that the portrait was not listed in an inventory dated 1751, compiled on the death of the purported sitter, nor was she identified as feu (the late) in her husband’s inventory of 1785, which suggests that the sitter was still alive. This statement must have been an oversight, however, for the portrait could not have been of the third, still-living Madame Bergeret, Jeanne Vignier, since Bergeret had married her in 1777, seven years after Boucher’s death (Bergeret’s second wife, Louise Mélanie de Lévy, died in 1773). Yet, as Laing noted, the date of 1746 inscribed on the portrait seems improbably early given Boucher’s relationship with Bergeret, who did not become a major patron until the 1750s. Moreover, a date of 1746 would make Madame Bergeret Boucher’s first full-length portrait and thus a remarkably accomplished effort for an artist who was not wholly at ease in the genre. Such an early date is all the more problematic given the close relationship of this portrait to several of Boucher’s representations of his favorite sitter, the marquise de Pompadour.
Boucher painted some nine portraits of Jeanne Antoinette Poisson (1721 – 1764), marquise de Pompadour and mistress of Louis XV, in a variety of settings and attitudes. The portrait of Madame Bergeret relates to several of these in different ways. Its pose and glance — though not its outdoor setting — are similar to those in an oil sketch of Pompadour [fig. 1], which has been connected to a large finished portrait (location unknown) commissioned in 1750. In both paintings the distinctive doll-like presentations of the sitter are remarkably alike: her head perched on a delicate lace ruff, the corsage pinned to her shoulder, one hand grasping the folds of a voluminous dress and the other unconsciously feeling for an object (in the case of Pompadour, the keys of a harpsichord). The garden setting, in contrast, closely resembles those appearing in two other portraits of Pompadour. The first is a small, finished portrait dated 1758, in the Victoria and Albert Museum [fig. 2], where the marquise is represented seated in an overgrown bower, looking up from her reading toward the left of the composition, much as Madame Bergeret glances out of her garden retreat; and the second is a slightly larger 1759 portrait in the Wallace Collection, London [fig. 3], in which Pompadour appears in a similar though more elaborate costume standing in an equally sylvan setting, with abundant foliage surrounding her and roses strewn about. The green bench at the left edge of the second portrait is comparable to, if not an exact duplicate of, the one that appears in the National Gallery’s painting.
These portraits of the marquise de Pompadour, all painted in the 1750s and bearing similarities of pose and setting to the portrait of Madame Bergeret, present a problem to the traditional reading of the National Gallery’s portrait. As Laing was the first to point out, although Boucher frequently reused figures and motifs in his compositions, it seems unlikely that he would have reemployed the pose, format, and setting of a portrait of Madame Bergeret, first created in 1746, for several representations of the king’s mistress, Boucher’s most important patron. Laing notes that it would have been doubly insulting to the marquise, since Madame Bergeret was the wife of a prominent financier, a class from which Pompadour had come but sought to distance herself. One is left with the unlikely scenario that in 1746, years before his association with Boucher, Bergeret de Grancourt commissioned a portrait of his wife, which the artist, despite his complete inexperience in full-length portraiture at that time, accomplished with singular success; and furthermore, that he then adapted this design (still feeling it necessary to work out the composition in oil sketches) to several portraits of the marquise de Pompadour.
The answer to this riddle likely lies in the date inscribed on the portrait. Again, it was Laing who first speculated that the date at the center left edge — 1746 — was spurious, an observation confirmed by technical analysis. To judge from craquelure patterns and overpainting, it is clear that both the date and the signature are later additions, probably strengthening an original inscription that had been damaged by a tear in the canvas. This leaves open the possibility that the painting was produced later than has generally been recognized, perhaps after the first portrait of Madame de Pompadour of 1750, now known only through oil sketches (see fig. 1). It might have been produced around 1751, shortly before the first Madame Bergeret died at the age of thirty, an age not at all incompatible with the young woman in the portrait.
Yet despite its clear relationship to the 1750 portrait of Pompadour, the National Gallery’s portrait is closer still — in costume, garden setting, and handling of brushwork, particularly in the facture of the draperies and foliage — to the portraits of the marquise produced in the late 1750s (figs. 2, 3). Nor is its style inconsistent with even later works, leaving open the possibility that it could have been painted as late as the mid-1760s. Assigning a date in the 1760s would solve several problems: during that decade, unlike the 1740s, Bergeret de Grancourt was actively patronizing Boucher, and this close relationship might have helped in convincing the painter to work in a genre that was not his custom. One might even specify a particular year, 1766, the date of Bergeret’s marriage to his second wife, Louise Mélanie de Lévy (d. 1773). It seems more than plausible that the current date visible on the portrait, 1746, was based on a misreading of a damaged authentic date of 1766 than one of 1750 or 1751. As Laing first proposed, the prominent and exquisitely painted roses — symbols of love and popular in marriage portraits — suggest that the painting may have been commissioned on the occasion of the Bergerets’ wedding in 1766. With this hypothesis in mind, it is less surprising that Boucher would have adapted a pose used in his 1750 portrait of Madame de Pompadour — a composition for which, it should be noted, he needed to make two oil studies (see fig. 1) — for his portrayal of Madame Bergeret rather than the other way around. Pompadour died in 1764, so there would have been nothing to keep Boucher from reusing the pose and setting of his earlier portraits of her.
The identification of the sitter in the National Gallery’s portrait as the second Madame Bergeret must remain speculative. No representations of her have been found with which to compare Boucher’s painting (which very likely was somewhat idealized). A portrait of such scale, beauty, and intimacy suggests a woman who was well loved and provided for. Little is known about Bergeret’s relationship with his second wife, although in 1770 (the year of Boucher’s death) he fathered a daughter with the family governess, Jeanne Vignier, whom he would later marry. If this portrait was indeed commissioned in celebration of Bergeret de Grancourt’s second wedding, it was clearly a time when he still had eyes only for his wife.
This text was previously published in Philip Conisbee et al., French Paintings of the Fifteenth through the Eighteenth Century, The Collections of the National Gallery of Art Systematic Catalogue (Washington, DC, 2009), 33–38.
Collection data may have been updated since the publication of the print volume. Additional light adaptations have been made for the presentation of this text online.
January 1, 2009