Figure 1 - Rembrandt, Self-Portrait in a Gorget, ca. 1628, 37.9 x 28.9 cm (10 cm=1 in), The Hague	Figure 2 - Isaac Jouderville (?), Copy after Rembrandt, Self-Portrait in a Gorget, ca. 1630, 38.2 x 31 cm (10 cm=1 in), Nuremberg

One of the most vivid illustrations of the current predicament of Rembrandt scholarship is the exhibition "Rembrandt by Himself," which was recently on view in London and The Hague. The show was directly connected with the Rembrandt Research Project, whose members provided essays for the catalogues and whose (revised) conclusions are adopted in the catalogue entries. In my opinion, the majority of the exhibited "self-portraits" are not by Rembrandt, but rather variations on self-portraits by Rembrandt or portraits of Rembrandt by his students.

There are already indications within the show that things are more complicated than they seem. The most significant example involves one of Rembrandt's most famous paintings, his self-portrait as a young man wearing a military gorget, from the Mauritshuis museum in The Hague (Fig. 1). The curators claim to have proved definitively that this is a copy by a student after Rembrandt's original in the Nationalmuseum of Nuremberg (Fig. 2), previously assumed to be a copy after the original in The Hague.1 This is a radical claim, and either bravely honest or rashly self-deceiving, since the Mauritshuis wa one of the co-sponsors of the exhibition. More specifically, either Rembrandt scholarship has been mistaken all this time and has applied the wrong criteria for determining Rembrandt's autograph works or the current curators are mistaken and apply the wrong criteria. The stakes are particularly high, since to my knowledge all commentators have adopted the "rehabilitation" of the copy as Rembrandt's original. In my opinion, the curators are mistaken, yet their error was only possible because the existing criteria are not clear enough, so that all works attributed to Rembrandt should be subject to question.

But let us first examine the arguments regarding the two "self-portraits." Technical examination of the Hague painting in 1998 revealed an under-drawing of charcoal or black chalk (Fig. 1a). The curators note that such a strategy is unprecedented in Rembrandt's work. They also claim that the positions of the eyes and hairline in the under-drawing correspond exactly to the Nuremberg painting and therefore conclude that the author of the Hague painting copied the Nuremberg painting in his under-drawing and later changed the composition in his painting. The curators observe further that "the meticulous painting technique of the Hague panel contrasts with the ease and spontaneity of the Nuremberg piece." They associate the latter approach with Rembrandt and the former with the student-copyist, specifically Rembrandt's first student Gerrit Dou, who is therefore put forward as the author of the Hague painting. The boundaries in the Hague composition furthermore "seem unnatural in places," "the reflection on the left of the lower lip is not quite right," and "the somewhat elongated face and close-set eyes diminish the resemblance to Rembrandt's features."2

Figure 1 - Rembrandt, Self-Portrait in a Gorget, ca. 1628, 37.9 x 28.9 cm (10 cm=1 in), The Hague	Figure 1a - Computer montage of infra-red reflectographs of Rembrandt, Self-Portrait in a Gorget in The Hague (10 cm=1 in).	Figure 2 - Isaac Jouderville (?), Copy after Rembrandt, Self-Portrait in a Gorget, ca. 1630, 38.2 x 31 cm (10 cm=1 in), Nuremberg

The relation described by the curators is exactly the opposite of what can be expected of a copyist following his master's example. It would be particularly odd for a copyist to follow his master's composition in an under-drawing, and then to elaborate freely on this in his painted composition. In fact, there are already changes in the under-drawing, as noted by the curators: "the contours of the gorget and torso have been drawn several times over, revealing an effort to find the correct position for these elements."3 These changes correspond to pentimenti in the Hague painting revealed by X-rays, which are not discussed in the catalogue of the self-portrait exhibition. The pentimenti were noted in the corpus of the Rembrandt Research Project where, partly for this reason, the painting was identified as the original and the Nuremberg version, which has no under-drawing or pentimenti, as the copy. But what about the correspondence between the under-drawing and the Nuremberg version? As with the ostensible weaknesses of the Hague painting or its lack of resemblance to Rembrandt, this is in my opinion simply a projection (or delusion) of the curators. The under-drawing, illustrated through a computer montage of infa-red reflectographs, is included in their catalogue on a separate page. When they are placed next to one another, the underdrawing makes perfect sense as the first step in the genesis of the Hague painting, whereas there is no evident correspondence to the Nuremberg painting (Figs. 1, 1a, 2).

The under-drawing furthermore corresponds to Rembrandt's self-portrait drawing in the British museum, executed in pen and brown ink and brush and gray wash (Fig. 3). In each case we find the individual strands of thinner hair at the upper left side of his face and across his forehead, as opposed to the thicker bunches of curls behind. In both the drawing and the painting, the shadow cast across the right side of the face marks a strong contour, tracing along the pointed chin, the ridges below the nose, and along the right eye socket, whereas some light falls on the right cheek and right eye, staring out at us from the shadow. The British museum drawing was included in the exhibition, where the curators note a resemblance to the Nuremberg painting and posit a connection with other ostensible Rembrandt self-portraits. They admit in a note that "it is conceivable that the maker of the Hague painting used this drawing as a second example" together with Nuremberg painting.4 It is far more logical to recognize the drawing as preparatory study, developed further in the under-drawing and the painting in The Hague. The fact that under-drawings are unprecedented in Rembrandt's work is less of a problem given that the example in the Hague painting was only discovered last year. There could be other examples. Conversely, this could be a singular example related to Rembrandt's evolving practice, just as the British Museum drawing is in my opinion his only preparatory study for a painted self-portrait.

Figure 3 - Rembrandt, Self-Portrait Study (pen and brown ink), ca. 1628, 127x95cm (not to scale), London	Figure 1a - Computer montage of infra-red reflectographs of Rembrandt, Self-Portrait in a Gorget in The Hague (10 cm=1 in).	Figure 1 - Rembrandt, Self-Portrait in a Gorget, ca. 1628, 37.9 x 28.9 cm (10 cm=1 in), The Hague	Figure 3 (10 cm=1 in.)

There are furthermore direct parallels between the Hague painting and Rembrandt's other autograph self-portraits. His late painted self-portraits in the Frick Museum and Metropolitan Museum in New York are particularly close to the proud, almost haughty bearing in the Hague painting (Figs. 4 and 5). In the earlier work, the three-quarter pose of the body is anchored by the gorget, which Rembrandt labored to get just right in his under-drawing and pentimenti, and which enhances the outward turn of his face toward us, dramatically lit from the left. The high position of his face within the composition, already evident in his drawing, reinforces the intensity of the portrait, which appears to stare down at us. Rembrandt portrays himself in the drawing with his mouth open, which has accordingly been interpreted as a "speaking portrait." In my view, Rembrandt records his complete absorption in the task of depiction, his slack jaw hanging down as he concentrates on capturing his own features, rather than assuming an artificial pose. The painting in the Hague is more composed, his lips slightly open, the corners of his mouth turned up in an barely perceptible smile.

Figure 1 - Rembrandt, Self-Portrait in a Gorget, (detail), ca. 1628, 37.9 x 28.9 cm (10 cm=1 in), The Hague	Figure 4- Rembrandt, Self-Portrait (detail), 1658, 131x102 cm, The Frick Collection	Figure 5 - Rembrandt, Self-Portrait (detail), 1660, 81x67.6 cm, The Metropolitan Museum of Art, New York

None of these elements are evident in the version in Nuremberg. There are no individual strands of hair or bunches of frizzy locks behind, but only vaguely described and mechanically repeated symmetrical, decorative curls. The transition to the shadow has been softened and does not form a contour at all, the chin is not pointed and the nose is not knobby. The shadow itself was too dark, so the copyist went over the right cheek and the neck below with thin strokes of white paint, which flatten the face even more. The figure now appears appear as if he had been in a fight, the right side of his face bruised, his right eye swollen. The right eye is in any case barely visible, and does not appear to stare at the viewer. The copyist was not recording himself looking at himself but merely repeating an existing composition, and with great difficulty, rather than "ease and spontaneity." It is not Hague painting, which is one of the few reliable records we possess regarding Rembrandt's appearance, but the Nuremberg version that diminishes the resemblance to Rembrandt's features by making the face more squat and widening the space between the eyes. Along with specific details, the plasticity of the face and costume have been sacrificed, as well as the distinctive pose. The figure does not appear to turn out to us or look down from above but is flattened out at the center of a two-dimensional composition. As the curators themselves admit,

The character is indeed expressed slightly better in the copy than in Rembrandt's original. By raising the head at a wider angle, closing the slightly parted lips of the original and replacing the wild curls with a well-groomed hairstyle, the copyist has given his portrait a distinguished air that is perfectly attuned to the aristocratic soldier.5

These observations are correct, only Rembrandt was the one who depicted himself as an aristocratic soldier, a young prince among painters. The student-copyist following his example was more of a foot-soldier, whose derivative "self-portrait" conveys uncertainty, a lack of distinction, and anonymity.

Self-Portrait with a Poodle

Figure 6 - Rembrandt, Self-Portrait as an Oriental, 1631, 66.5 x 52 cm (10 cm=.66 in), Paris, Musée de Petit Palais	Figure 7- Jouderville (?),Copy after Rembrandt, Self-Portrait as an Oriental, ca. 1631, 70.4 x 50.2 cm (10 cm=.66 in), Amsterdam, Museum Het Rembrandthuis

Further evidence for this conclusion is provided by another pair of paintings that were included in the exhibition, in which the relation of original and copy is not in doubt. Rembrandt's Self-Portrait as an Oriental in Paris was copied by a student in a painting now in the Rembrandthuis, Amsterdam (Figs.6 and 7). In this case as well, Rembrandt's original is executed with a meticulous painting technique, rendering his distinctive frizzy hair, cleft chin, and staring right eye, as well as the intricate detail and plasticity of the turban, velvet cloak with metal clasps, and embroidered silk robe. As in the case of the Nuremberg painting, the copyist summarizes, softens, or dulls these elements, in a largely two-dimensional composition.The copy in Amsterdam has long been attributed to Rembrandt's second pupil in Leiden, Isaac Jouderville, Unlike Dou, who quickly began to paint his own peculiarly static genre subjects, Jouderville continued to copy and vary Rembrandt's compositions for several years, focusing on single-figures and portraits, including what seem to be variations on Rembrandt's self-portraits even though thet depict women. in particular. Jouderville is therefore the most logical candidate for the author of the Nuremberg painting.
A particularly interesting detail is the absence of the poodle. X-rays reveal that this was not included at first in Rembrandt's original. The curators claim that "this is the only occasion on which Rembrandt painted a full-length portrait of himself" and propose that "he found it no easy task," although they assume the dog was "an afterthought" made several years later. Rather, Rembrandt already portrayed himself as a full-length figure without apparent difficulty in his Artist in the Studio in Boston, which was also included in the exhibition. I would propose another explanation for the addition. After Rembrandt saw his student's copy, which renders the feet in a dull, flat, awkward manner, and too close to the edge of the panel, undermining the spatial dimensions of the scene, he decided to underscore the depth of his own composition by employing the poodle as a space-creating repoussoir. The motif further underscores the distinction between his original and the student copy, specifically through an animal whose subtly rendered, complex hair was presumably beyond the technical capacities of the student to reproduce. The poodle is not simply an aristocratic attribute of an oriental potentate or prince but embodies Rembrandt's skill and originality. The dog even echoes Rembrandt's appearance, as is often said of pets and their owners, with its shaggy locks and top-heavy main of hair over its shaved lower body, paralleling Rembrandt's velvet cloak and leggings. Rembrandt's Self-Portrait as an Oriental not only emphatically identifies itself as "the original" but can also serve as a means to recognize Rembrandt's other originals. He conceivably had such a purpose in mind in relation to his prior self-portrait and its copy, which could have already been mistaken for the original in his time, as it recently was. Rembrandt's students copying his example thus stimulated his originality and forced him to define his artistic self or personality in increasingly explicit terms.