When it was thoroughly cataloged and discussed by Richard Rand in 2009, Young Girl Reading by Jean Honoré Fragonard was linked, albeit with important qualifications, with the artist’s so-called figures de fantaisie (fantasy figures) or portraits de fantaisie (fantasy portraits)—half-length representations of men and women shown striking dramatic poses in colorful masquerade dress and rendered in loose, gestural brushstrokes. ^[1] Among the most beloved works in the artist’s oeuvre, these pictures are also among the most mysterious and have therefore prompted the most debate: produced for unknown reasons, perhaps representing real individuals, perhaps not. Subsequent research and recent developments—explored in depth in three recent publications and websites ^[2]—have shed new light upon Young Girl Reading and Fragonard’s fantasy figure series in general.

In 2012, a previously unknown drawing appeared on the art market. Covered with eighteen thumbnail-sized sketches that, but for one, were annotated—apparently in the artist’s hand—the drawing suggested that the fantasy figures were indeed portraits of identifiable individuals, members of the artist’s professional network of clients and models.

The emergence of this drawing, now known as Sketches of Portraits, had an additional resonance at the National Gallery of Art, for one of the sketches evoked the Gallery’s own Young Girl Reading. Technical studies (2013–2015) confirmed the presence of an underlying painting in which the model poses as she appears in the Sketches of Portraits—face turned outward—and not as she appears in the finished painting that has long hung on the Gallery’s walls. This discovery, which firmly linked Young Girl Reading to the fantasy figure series, was the impetus for the Gallery’s exhibition Fragonard: The Fantasy Figures (2017). Uniting several fantasy figures with the drawing, the exhibition focused solely on this aspect of the artist’s production, while the catalog that accompanied it explores Fragonard’s technique in detail.

The support is a slightly coarse plain-weave fabric. The painting has been lined, and the tacking margins have been removed. Prominent cusping on all four edges suggests that the painting has not been cut down. The support was prepared with two ground layers: a pale-gray layer covered by a fawn-colored layer. The paint was applied vigorously, with impasto in highlights and thin washes that leave the ground partially visible in the shadows. The gray shadowed lines in the girl’s collar and fichu were created by incising into the wet white paint with the butt end of the brush to reveal the gray layer beneath. The X-radiograph was originally thought to reveal an earlier painting underneath the current head showing the head of a man wearing a feathered hat. Further analysis (false-color infrared imaging and XRF elemental mapping of both lead white and vermillion) conducted between 2013 and 2015 produced images indicating that the underlying head is actually that of a woman wearing a beaded and feathered headdress, her face turned out to gaze directly at the viewer. Due to an increase in the translucency of the surface paint, the outline, eyes, and feathered headdress of the woman are now slightly visible as pentimenti. Cross-sectional analysis shows that there is no intermediate paint layer between the two heads, nor is there varnish or dirt between them. Additionally, it shows that the original paint layer was not disrupted in any way when painted over, suggesting it was dry before the application of the now-visible paint layer. This suggests that Fragonard did not embark on the repainting immediately, as the lower layer would require at least six months to dry to this degree. ^[3]

Overall, the painting is in good condition. To the right of the girl’s neck is a small complex tear. A larger J-shaped tear extends through the pillow and arm of the chair at the bottom right. The paint is slightly abraded in the thinly painted folds and shadows of the dress and in the darks along the bottom of the painting. In 1985 a discolored varnish was removed, and the two distorted tears were realigned and inpainted. The sitter’s head from the previous painting, which had become distractingly visible, was also inpainted at that time. In 1986 a slightly toned varnish was applied. The inpainting of the larger tear has discolored slightly, but the remainder of the inpainting and the varnish have not.