For an artist whose face is so well known through his numerous painted, drawn, and etched self-portraits, it is quite remarkable that Rembrandt’s image in this painting was not recognized by early nineteenth-century critics. While it was in the possession of Chevalier Érard and William Williams Hope, two important and discerning collectors, the sitter was thought to be the important Dutch admiral Maerten Harpertsz Tromp (1598–1653). One wonders what prompted this unexpected belief since Tromp’s known portraits look totally different.[1]  [1]
Since the provenance for this painting is not known prior to the mention in the Érard Collection, it is not known whether the identification was based on an even older tradition. For an image of Tromp from the early 1650s, see Jan Lievens’ Portrait of the Vice-Admiral, Maerten Harpertsz Tromp (Rijksmuseum, Amsterdam, inv. no. A 838). To judge from the commentary in the Érard catalog, the theory seems to have been based partially on the outmoded costume: the pleated white shirt, the dark overdress with its rich impastos bordering the front and slashed purple sleeves lined with yellow, and the brown beret worn at a jaunty angle over the elaborate yellow-and-red skullcap. The theory that the portrait depicted an admiral was reinforced by the gold-handled staff upon which the sitter rests his hand. But primarily, it seems, the depiction of the sitter’s character fit what was thought to be Tromp’s assured, noble, and philosophical nature.[2]  [2]
Sébastien Érard estate sale, Paris, August 7–14, 1832 (rescheduled from April 23 and following days), no. 119, 136–137, includes the following description: Des traits mâles, une contenance assurée, de la noblesse unie à beaucoup de simplicité, donnent une grande expression à ce beau portrait. Dans la demi-teinte qui l’enveloppe et qui va si bien à sa gravité, on pourrait voir une pensée philosophique, une allusion dont Rembrandt était bien capable. Martin Tromp, indifférent pour les titres, honorifiques, pour les choses d’apparat, modeste au plus haut point, ne dut trouver du plaisir à se montrer que quand il était en présence des ennemis de sa nation. Au surplus, quelqu’ait été l’intention du peintre, cette ombre répandue sur la figure d’un tel homme sied bien à son caractère.

While the sitter’s expression, created through subtle effects of lighting on the face as well as the unusual, sidelong glance, has continued to intrigue writers, entirely different interpretations of its character have been advanced by critics who recognized that the painting represented Rembrandt.[3]  [3]
John Smith, A Catalogue Raisonné of the Works of the Most Eminent Dutch, Flemish and French Painters, 9 vols. (London, 1829–1842), 7:86–87, no. 211, was the first to correctly identify the painting as a portrait of Rembrandt. Bode found little evidence of a philosophical mind in the image. He wrote that the “somewhat leering expression, half weary, half watchful, . . . is by no means favourable to his appearance.”[4]  [4]
Wilhelm von Bode assisted by Cornelis Hofstede de Groot, The Complete Work of Rembrandt, trans. Florence Simmonds, 8 vols. (Paris, 1897–1906), 5:15. Valentiner, on the other hand, saw in the face “a forehead already deeply lined, melancholy, almost despondent of aspect, indicating how early he had experienced the tragic side of life.”[5]  [5]
Wilhelm R. Valentiner, Rembrandt Paintings in America (New York, 1931), introduction. While Rosenberg described the same expression as “critical and deeply questioning,”[6]  [6]
Jakob Rosenberg, Rembrandt, 2 vols. (Cambridge, Mass., 1948), 1: 28. Pinder, reflecting more closely the sentiments of Valentiner, felt that the portrait expressed the cares and worries that were beginning to beset Rembrandt.[7]  [7]
Wilhelm Pinder, Rembrandts Selbstbildnisse (Munich, 1950), 81–82. Contrary to these interpretations of the image as representing a despondent and questioning individual was that of Goldscheider, who described the Rembrandt portrayed here as “a handsome, distinguished adventurer without the slightest resemblance to a humble Mennonite.”[8]  [8]
Ludwig Goldscheider, Rembrandt Paintings, Drawings and Etchings (London, 1960), 174, cat. 65.

While the reading of the emotional impact of the image may have varied markedly among these and other authors, they had in common a conviction that this painting was an exceptional work by the master.[9]  [9]
Ludwig Goldscheider, Rembrandt Paintings, Drawings and Etchings (London, 1960), 174, cat. 65, considered it “one of the finest portraits ever painted.” Thus the shock that greeted Horst Gerson’s 1969 pronouncement that the painting was an “18th- or 19th-century imitation, combining light effects typical of Rembrandt’s early work with a composition and mood characteristic of the later period” was felt throughout the world of Rembrandt scholarship.[10]  [10]
Abraham Bredius, Rembrandt: The Complete Edition of the Paintings, revised by Horst Gerson (London, 1969), 550, cat. 39. The reaction can be judged by the fact that Egbert Haverkamp-Begemann, “The Present State of Rembrandt Studies,” Art Bulletin 53 (March 1971): 93–94, listed this work first among what he considered Gerson’s “five or six spectacular ‘disattributions’ of well-known and admired paintings, in some cases never previously doubted.” Haverkamp-Begemann noted in 1993 that he continued to believe in the attribution to Rembrandt (personal communication). With that statement Gerson altered once again the way the painting has been viewed. For just as nineteenth-century critics had to reassess their interpretations of the figure’s mood when the identification of the sitter changed, it is necessary to raise the more fundamental question as to whether the expressive character of the painting is, after all, consistent with Rembrandt, and even with seventeenth-century sensibilities.

Technical analyses have shown that Gerson’s assertion that the painting is a later imitation is wrong. The character of the paint mixtures, the types of pigments used, and the presence of a double GroundGround
 The layer or layers used to prepare the support to hold the paint.—a red lower ground covered by a dark gray upper layer—are all totally consistent with Rembrandt’s workshop practices (see Technical Summary).[11]  [11]
I would like to thank Barbara A. Miller (former conservation scientist at the National Gallery of Art, who first analyzed the painting in 1981), Michael Palmer, and Melanie Gifford for their help in interpreting the technical data. There also appears to be nothing unusual in the type of canvas used or in the buildup of the image. The initial blocking-in of the form, which can be seen in the waist and unfinished right hand, is also consistent with Rembrandt’s manner of painting. Finally, X-radiographs [see X-radiographyX-radiography
 A photographic or digital image analysis method that visually records an object's ability to absorb or transmit x-rays. The differential absorption pattern is useful for examining an object's internal structure as well as for comparing the variation in pigment types.] reveal that modifications to the shape of the hat were made during the execution of the painting [fig. 1]   [fig. 1] X-radiograph composite, Rembrandt Workshop, Portrait of Rembrandt, 1650, oil on canvas, National Gallery of Art, Washington, Widener Collection, a phenomenon that is commonly found in Rembrandt’s own paintings (see his Self-Portrait). Originally, the plaid-patterned skullcap under the beret extended out behind the head more than it presently does.[12]  [12]
Its form can also be seen with the naked eye.

The conservation treatment of the painting in 1993, however, revealed that Gerson was correct in his intuition that the execution was somehow at variance with that found in Rembrandt’s own works around 1650.[13]  [13]
The conservation treatment was undertaken in 1992–1993. Although the general disposition of the figure, standing at an angle to the picture plane and looking quizzically at the viewer over his near shoulder, is consistent with Rembrandt self-portraits [fig. 2]   [fig. 2] Rembrandt van Rijn, Self-Portrait as the Apostle Paul, 1661, oil on canvas, Rijksmuseum, Amsterdam. Photo © Rijksmuseum, Amsterdam, the brushwork used to model the figure lacks both Rembrandt’s sensitivity and vigor. The most obvious instance where the modeling is at variance with Rembrandt’s style is the hand resting on the staff. Its superficially rendered form has nothing to do with his manner of modeling hands.

The modeling of the face, likewise, lacks firmness and conviction. While the play of light across the features is sensitively rendered, the restrained brushstrokes only vaguely suggest the underlying form, whether it be the shape of the eyelids or the contour of the nose. The weakness of character conveyed through Rembrandt’s questioning expression is also the result of the irresolute contours defining his features.

Another unusual aspect of this painting is that the costume is executed in a manner quite different from the face. Whereas the features are modeled with delicate nuance, the costume is indicated with a variety of bold techniques.[14]  [14]
For a careful analysis of the sitter’s historicizing costume, which “is reminiscent of German or Netherlandish attire of the first half of the 16th century,” see Stichting Foundation Rembrandt Research Project, A Corpus of Rembrandt Paintings, vol. 4, Self-Portraits, ed. Ernst van de Wetering  (Dordrecht, 2005), 399–400. Broad, flat planes of purple and yellow enliven the surface of the split sleeve; thick impastos, mixed with a variety of bright colors and then toned with a dark brown glaze, create the appearance of an embroidered frontispiece across the chest; and, finally, the plaid pattern of the skullcap is painted with vigorous, and quite specific, strokes of red and yellow. The attribution problems raised here are threefold. First, Rembrandt generally did not use such markedly different techniques in the face and costume of a portrait. Second, by the 1650s, he had developed a manner of painting that would allow him to suggest a wide range of textures in materials without significantly altering the techniques that he used to depict them. Finally, his paint always worked toward creating structure. In this costume, most of the accents of color, as for example on the sleeve or in the skullcap, do not work effectively to convey the nature of the material. This problem is also particularly evident in the superficial black strokes that define the collar of the costume.

Although these stylistic considerations are sufficient to remove the painting from Rembrandt’s own oeuvre, the identity of the artist who actually executed this portrait cannot be determined. The signature and date, while apparently not written by Rembrandt, appear to be integral to the surface and probably indicate that the painting was executed by a member of the workshop in about 1650 to be sold on the open market. It may well be that Rembrandt, after having posed for this painting, approved its concept and manner of execution before allowing its sale. To judge from the number of self-portraits Rembrandt painted and etched, and from the numerous portraits of him painted by members of his workshop, there must have been a ready market for images of the artist.

None of the painters known to have been in Rembrandt’s workshop around 1650, including Willem Drost (Dutch, c. 1630 - after 1680), Jacobus Leveck (1634–1675), Nicolaes Maes (Dutch, 1634 - 1693), and Constantijn van Renesse (Dutch, 1626 - 1680), can be convincingly associated with this work. The differences in the handling of the paint in the head and the costume are so pronounced in this work that I have wondered whether two artists might have executed the painting. No technical evidence, however, suggests that the painting was a collaborative effort. The stylistic discrepancies are probably the result of a workshop assistant basing his style for modeling the head on Rembrandt’s work of the mid-1630s and his manner of painting drapery on Rembrandt’s style of the early 1650s. Close stylistic comparison can be made to Man with a Gilded Helmet [fig. 3]   [fig. 3] Rembrandt School, Man with a Gilded Helmet, early 1650s, oil on canvas, Staatliche Museen zu Berlin, Preussischer Kulturbesitz, Gemäldegalerie. Photo: bpk, Berlin / Gemäldegalerie, Staatliche Museen, Berlin / Jörg P. Anders / Art Resource, NY, an unsigned and undated work from the Rembrandt workshop that is datable to the early 1650s.[15]  [15]
For an excellent discussion of this work, including information about its restoration, see Jan Kelch et al., Der Mann mit dem Goldhelm, Bilder im Blickpunkt (Berlin, 1986). In this work, too, the face and costume are rendered in strikingly different manners. In the Berlin painting a marked contrast exists between the relatively delicate modeling of the face and the thick impastos in the helmet, a contrast in techniques quite similar to that found in the Portrait of Rembrandt. While various attributions have been suggested for Man with a Gilded Helmet, including the artists Karel van der Pluym (1625–1672) and Heyman Dullaert (1636–1684), none is convincing.[16]  [16]
The attribution of this painting to Van der Pluym was made by Henry Adams, “If Not Rembrandt, Then His Cousin?” Art Bulletin 66 (September 1984): 427–441. Claus Grimm, “Handschrift, schildertechniek en beeldstructuur. Bijdrage tot het onderzoek naar toeschrijvingen, I, de helmen van Rembrandt,” Tableau 5 (1982–1983): 242–250, attributed the painting to Dullaert. As noted by the Rembrandt Research Project, an even closer correlation exists between Portrait of Rembrandt and two pendant portraits from 1650 by an unknown artist from Rembrandt’s workshop: Man in a Military Costume, Fitzwilliam Museum, Cambridge [fig. 4]   [fig. 4] Unkown follower of Rembrandt, Man in a Military Costume, 1650, oil on panel, Fitzwilliam Museum, Cambridge. Photo: The Bridgeman Art Library,[17]  [17]
See Br. 256, from the Fitzwilliam Museum, Cambridge. and Woman in a Fanciful Costume, John and Mable Ringling Museum of Art, Sarasota, Florida.[18]  [18]
See Br. 380, from the John and Mable Ringling Museum of Art, Sarasota. They reached their conclusions on the basis of technical analyses of the paint layers, and a comparable contrast between carefully modeled faces and freely rendered historicizing costumes as seen in Portrait of Rembrandt.[19]  [19]
Stichting Foundation Rembrandt Research Project, A Corpus of Rembrandt Paintings, vol. 4, Self-Portraits, ed. Ernst van de Wetering (Dordrecht, 2005), 400–401.   Whether or not the same unknown artist painted all of these historicizing portraits is a puzzle yet to be solved.

Arthur K. Wheelock Jr.

April 24, 2014