Jean Honoré Fragonard’s fantasy portraits or ‘fantasy figures’, as the exhibition at the National Gallery of Art, Washington (to 3rd December), calls them, have long presented a series of problems. He made at least twenty, depending on how one reads the evidence, and the best are among the artist’s greatest works. Their uniform size, free, spirited brushwork and similarity of subject – men and women dressed in fanciful costumes (usually described as Spanish, but more likely seventeenth-century Flemish), shown at half length, their bodies and heads turned in dynamic contrapposto – suggest a series, or at least a concentrated artistic project. Yet we know little about them: the provenance of only three can be traced back to Fragonard’s lifetime and only one is dated, to 1769. Despite variations in handling, more evident when seeing a large number together, it is likely that the pictures were all painted over a short period of time.

Were they meant to be displayed as a group, perhaps to decorate a gallery? Are they truly portraits in any meaningful sense, that is, likenesses of actual people known to the artist, or are they better understood as expressive heads in the tradition of Dutch tronies?1 Several of the figures have long been assumed to represent Fragonard’s friends, or celebrities such as the dancer La Guimard or the philosopher Denis Diderot (cat. no.3; Musée du Louvre, Paris). Two (nos.12 and 13; private collections) can reasonably be identified through family tradition as depicting the dukes of Harcourt and Beuvron, members of a distinguished Norman family. But most of them can also be appreciated as allegories of the arts or other liberal pursuits. Since 1960, when George Wildenstein first identified these pictures as ‘portraits de fantaisie’ and described them, rather unconvincingly, as ‘works in which the expression of the face is not stressed’, scholars have sought to unravel their mysteries.2

For audiences today, as in the past, it is their brilliant execution that is most striking and appealing. Charles Sterling memorably wrote that ‘one could almost say the wind itself had painted them’.3 Two of the finest, both in the Musée du Louvre, Paris, representing Fragonard’s patron the Abbé de Saint-Non (not exhibited) and his brother Louis Richard de La Bretèche (no.7; Fig.87) have old inscriptions asserting that the pictures were ‘painted in an hour’. Recent technical analysis of several has confirmed that they were painted wet-on-wet on commercially prepared canvases and that Fragonard first completed the busts, after which he quickly brushed in the backgrounds. This astonishing performative aspect, understood in Fragonard’s time as a mark of creative genius, allows us to think of the paintings as virtual self-portraits; as Satish Padiyar writes in his outstanding essay in the catalogue, ‘the fantasy figures are in this sense all Fragonard’.4 Painted during the years he was distancing himself from the Académie royale and establishing a freelance career, they project his self-assertion and independent spirit. It is no surprise to learn that when several were exhibited in the 1860s they were immediately appreciated and emulated by Edouard Manet and the young Impressionists.5

The exhibition offered the first occasion since the great Fragonard show in Paris and New York in 1987–88 to study a large number of the fantasy figures in a single room.6 The curator, Yuriko Jackall, is to be commended for assembling no fewer than fourteen from public and private collections in the United States and Europe. The impetus for the show was a new technical study of the National Gallery of Art’s Young girl reading (no.16; Fig.88), which rightly holds pride of place in the centre of the gallery. It is one of the most charming of the fantasy portraits, but also the least typical: uniquely, the sitter is depicted in profile, turned from the viewer and absorbed in her book. It has long been known from x-radiography – confirmed by newer techniques – that Fragonard first painted her head turned toward the viewer, in keeping with the others in the series, but for unknown reasons repainted it, thereby changing an ostensible portrait into a more conventional genre scene.7

A number of questions appeared to be resolved with the discovery at an auction in Paris in 2012 of a sheet of sketches by Fragonard representing many of the known fantasy figures, with several related compositions (no.19; Fig.86). The sketches are unusually slight – mere thumbnails – the majority drawn quickly in pen-and-ink but two are faintly sketched in graphite. Inscriptions on the mount indicate that the drawing remained with descendants of the artist until at least the late nineteenth century. Of the eighteen compositions recorded on the sheet, fourteen can be matched with known paintings, including an oval canvas of a Vestal (no.5; private collection) that had never been associated with the fantasy figures. As specialists were quick to realise, the sheet suggests that Fragonard thought of the paintings as a group or series. More significantly, he inscribed names below all but one, establishing once and for all that the fantasy figures were based on actual people. Although the artist’s habit of spelling names phonetically presents some challenges, Carole Blumenfeld has plausibly linked the inscriptions with known personalities of Fragonard’s era.8 The traditional identifications of Saint-Non and La Bretèche are now confirmed, but the canvas long assumed to depict Diderot is labelled ‘Meunier’, whom Blumenfeld identifies as Ange-Gabriel Meusnier de Querlon, a lawyer, part-time pamphleteer and occasional writer of erotic literature. Other sitters were also drawn from Fragonard’s circle of friends in the world of arts and letters as well as from the financial circles who provided many of his patrons.

Intriguingly, the full-length Cavalier seated by a fountain (no.11; Museu Nacional d’Art de Catalunya, Barcelona), different from and larger than the fantasy figures but often associated with them, also appears on the sheet as a faint sketch in graphite, as do three other large canvases, still to be rediscovered, also showing full-length figures; Blumenfeld has already found an old copy of one (private collection; ill. cat. p.105). The Harcourt fantasy figures do not appear, nor does the Man in costume (no.17; Art Institute of Chicago); they may not yet have been painted when Fragonard made the drawing. The sheet gives the impression of a work in progress, some kind of ongoing catalogue of a particularly important group of pictures. The sketch corresponding to the National Gallery of Art’s Young girl reading – the only one without an identifying inscription – shows the first version of the composition, with the face turned toward the viewer, strong evidence that the sketches are contemporaneous with the paintings.

It is not clear why Fragonard recorded his fantasy figures in this way. He certainly did not consider the drawing a work of art in its own right: he employed a common type of writing paper and, judging from the stains and folds, made no effort to keep it safe. He took care, nevertheless, in arranging his sketches, lining up the two rows of the halflength figures across the width of the page and placing the oval Vestal in the centre of the top row. Whether this proves he intended it as a design for the decoration of a room, perhaps for the enjoyment of one of the many literary salons that flourished in Paris – a tantalising idea proposed by Marie-Anne Dupuy-Vachey – remains an open question.9 The notion of a gallery of such lively and convivial painted faces has appeal, but, as Jean-Pierre Cuzin points out in his catalogue essay, the juxtapositions of the sketches have little of the aesthetic logic one would expect from Fragonard – for good reason Jackall chose not to follow the layout of the drawing in her installation in Washington. Furthermore, had such an extraordinary gallery of pictures actually existed, there would surely be contemporary references to it. Cuzin is probably right to insist that the pictures were not a decorative project but a thematic series, similar to Claude Monet’s paintings of Rouen Cathedral.

In the end, the sheet of sketches raises new questions even as it answers others. We might never know why it was made, but it may not be unique. An old inscription on the mount reads ‘No. 12. Esquisses des portraits’, raising the possibility that Fragonard made at least eleven other such copy-sketches of groups of paintings. One can only speculate as to what these lost sheets might have recorded: his small ‘Dutch’-style landscapes? His large garden paintings, now divided between Washington and the Banque de France, Paris? Or even his original layout for Madame du Barry’s The progress of love, painted for Louveciennes and now in the Frick Collection, New York?

1 See M. Percival: Fragonard and the Fantasy Figure: Painting the Imagination, Farnham and Burlington VT 2012.

2 G. Wildenstein: The Paintings of Fragonard: Complete Edition, London 1960, pp.13–15.

3 C. Sterling: Portrait of a Man (The Warrior): Jean Honoré Fragonard, Williamstown MA 1964, n.p.

4 S. Padiyar: ‘1769: The leap into freedom’ in the catalogue: Fragonard: The Fantasy Figures. Edited by Yuriko Jackall, with technical studies by John K. Delaney and Michael Swicklik and essays by Carole Blumenfeld, Kimberly Chrisman-Campbell, Jean- Pierre Cuzin, Elodie Kong and Satish Padiyar. 160 pp. incl. 190 col. ills. (National Gallery of Art, Washington, in association with Lund Humphries, London, 2017), $49.99. ISBN 978–1–84822–238–0. See also M. Sheriff: ‘Invention, resemblance, and Fragonard’s Portraits de fantaisie’, Art Bulletin 69, 1 (March 1987), pp.77–87.

5 R. Rand: ‘Painted in an hour: Impressionism and eighteenth-century French art’, in A. Dumas, ed.: exh. cat. Inspiring Impressionism: The Impressionists and the Art of the Past, Denver (Art Museum) 2007, pp.135–55.

6 Reviewed by Philip Conisbee in this Magazine 130 (1988), pp.319–21.

7 Y. Jackall, J.K. Delaney and M. Swicklik: ‘“Portrait of a woman with a book”: a newly discovered fantasy figure by Fragonard at the National Gallery of Art, Washington’, The Burlington Magazine157 (2015), pp.248–53.

8 C. Blumenfeld: Une facétie de Fragonard: Les révélations d’un dessin retrouvé, Montreuil 2013.

9 M.-A. Dupuy-Vachey: ‘Fragonard’s “fantasy figures”: prelude to a new understanding’, The Burlington Magazine 157 (2015), pp.241–47.