With a tender expression accentuated by her tilted head and slightly parted lips, the girl in this bust-length image gazes downward to the left. Strong light falling from the upper left illuminates her face and the broad, cream-colored kerchief draped loosely around her neck, while a subtle, reflected light models the shaded areas of her cheek and chin. Executed in both thick impastos and thin, translucent paints, this rapidly executed oil study has a compelling presence that belies its small scale.
Wilhelm von Bode was the first scholar in the first half of the twentieth century to publish this engaging head study as Rembrandt’s.1 In 1948, Jacob Rosenberg argued that this sketch was Rembrandt’s preliminary study for Mary in his Holy Family, 1645, in the State Hermitage Museum in St. Petersburg (fig 1). Indeed, the position of Mary’s head, her expression, and the character of the direct and indirect light illuminating her face are almost identical to that of the young girl in the oil study (fig 2). Since the 1960s, however, most art historians have argued that this oil study is a freely executed copy of Mary in Holy Family by a member of Rembrandt’s workshop and not Rembrandt’s preliminary study for that figure.2 The painting has been associated with a number of Rembrandt’s students, including Carel Fabritius (1622–54), Samuel van Hoogstraten (1627–78), and Nicolaes Maes (1634–93), but none of these proposed attributions has proved to be compelling.3 As a consequence, the identification of the artist responsible for this engaging oil study remains an enigma.4
The question of whether or not this bust is Rembrandt’s preliminary study or a copy by a workshop assistant relates to a larger problem within Rembrandt scholarship. As Ernst van de Wetering has demonstrated, Rembrandt often made preparatory studies of figures for his paintings to work out patterns of light and shade.5 In these studies he was particularly interested in rendering light that had been reflected into shadowed areas of the face. Michiel Franken has argued, however, that Rembrandt also encouraged his students to make partial copies after his compositions, and to alter these copies so that they could pass as tronies in their own right.6 It is not always easy to tell the difference between these two types of oil sketches, and debates about their attributions are some of the most complicated issues in Rembrandt studies.
Careful comparisons of the Leiden Collection’s Head of a Girl and the figure of Mary in Holy Family support Rosenberg’s conclusion that this sketch is a preliminary study for the Hermitage painting and not a copy after it. For example, the strokes of white paint evident above the girl’s head, which are also visible in the X-radiograph (fig 3), indicate that the artist initially provided her with a small bonnet rather than the larger headdress worn by Mary in the final composition. The X-radiograph also reveals changes in the position of the white cloth to the left of the girl’s face, an unlikely occurrence should the Leiden Collection painting be a copy. Other differences exist between the oil sketch and the final composition, including the girl’s hairstyle, which is not parted in the middle as is the case with Mary. Mary leans more forward than does the girl in the oil study, a difference in pose that is evident in the position of Mary’s right shoulder, which is higher than that of the girl in the oil sketch. Also revealing is the shape of the young girl’s neckline. In the Hermitage painting Mary wears a straight-edged bodice, but a visible pentimento in this area reveals that her neckline was V-shaped, exactly as in the oil sketch from the Leiden Collection.7 Finally, the handling of paint, which includes both impastos and thin areas in the shadows that reveal the underlying ocher ground layer, is consistent with Rembrandt’s manner and not with a copy. The subtle nuances of reflected light in the shaded portion of the girl’s face are also comparable to those found on Rembrandt’s other oil studies.8
Nowhere in Rembrandt’s oeuvre does one find so many correlations between an oil sketch and the initial appearance of one of Rembrandt’s compositions as in this instance. Indeed, all of the other the oil sketches that Franken describes as workshop copies are based on one of the master’s completed compositions, not on an initial stage in the creative process. Franken argues that the Leiden Collection oil study was made by a copyist who adapted his image so that it could function as “a ‘tronie’ in its own right,” but this conclusion seems improbable since the oil study so closely reflects Holy Family’s underlying design. Van den Boogert and Van de Wetering, recognizing this fact, maintain that Head of a Girl was executed by a member of Rembrandt’s workshop before Rembrandt had worked out his final composition for the Hermitage painting.9 As evidence to support their theory, they contend that an infrared reflectogram of the girl’s head reveals a fairly detailed underdrawing in black chalk or charcoal, a technique that Rembrandt is not known to have used.10 Their reading of the reflectogram, however, mistakenly interpreted the black lines seen in the image as underdrawings rather than strokes of carbon-containing paint used to model the girl’s features. After one discounts this faulty interpretation of the reflectogram, there are, in fact, no compositional or technical reasons for questioning the attribution of this oil sketch to Rembrandt.
A fundamental question when considering the attribution is whether the work is stylistically consistent with Rembrandt’s oil studies from the mid-1640s. This issue is particularly important because no Rembrandt scholar has attributed the painting to the master since 1966, when Bauch deemed the painting to be a partial copy after Rembrandt’s Holy Family. In this respect, the historical situation related to Head of a Girl parallels that of a number of other oil sketches that have been considered to be partial copies by pupils, but that actually may have been preliminary studies Rembrandt made for history paintings. Van de Wetering has identified a number of these, mostly expressive head studies that demonstrate a quick and fluid touch, often boldly executed with thick impastos.12 Van de Wetering, however, has not reattributed Head of a Girl to Rembrandt, and he, like other recent Rembrandt scholars, considers it to be a copy by a member of Rembrandt’s workshop.13
The reason that scholars have resisted attributing this oil sketch to Rembrandt may be that the young woman is more carefully rendered than the preliminary oil sketches generally associated with Rembrandt’s hand, all of which are boldly and expressively brushed. Indeed, Head of a Girl has the character of a figure study rather than a quick compositional sketch. In this respect it resembles the various head studies of Christ that Rembrandt and his workshop painted from a young Jewish model in the late 1640s, none of which were actually preliminary studies for paintings. A particularly close comparison to Head of a Girl is the Head of Christ from Berlin (fig 4). Both young models are posed looking down and to the right, their averted gazes evoking a sense of mystery and inward reflection.14 Their features are similarly rendered, as is the application of paint. In each instance the face has been worked up from a brown monochrome sketch, which has been left exposed in the temple and in the shaded portions of the nose and proper right eye socket. The foreheads and cheekbones of both figures have been comparably enlivened with freely brushed impastos.
Thus, in addition to its high quality, the stylistic, compositional, and technical character of this compelling head study warrant defending its attribution to Rembrandt. It is probable that the master executed this painting around 1645 as a preparatory study for the figure of Mary in Holy Family.15